Pam Groenewald, Jason Thomas, Samuel J Clark, Diane Morof, Jané D Joubert, Chodziwadziwa Kabudula, Zehang Li, Debbie Bradshaw
{"title":"南非计算机编码死因推断方法与医生死因推断访谈编码之间的一致性。","authors":"Pam Groenewald, Jason Thomas, Samuel J Clark, Diane Morof, Jané D Joubert, Chodziwadziwa Kabudula, Zehang Li, Debbie Bradshaw","doi":"10.1080/16549716.2023.2285105","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>The South African national cause of death validation (NCODV 2017/18) project collected a national sample of verbal autopsies (VA) with cause of death (COD) assignment by physician-coded VA (PCVA) and computer-coded VA (CCVA).</p><p><strong>Objective: </strong>The performance of three CCVA algorithms (InterVA-5, InSilicoVA and Tariff 2.0) in assigning a COD was compared with PCVA (reference standard).</p><p><strong>Methods: </strong>Seven performance metrics assessed individual and population level agreement of COD assignment by age, sex and place of death subgroups. Positive predictive value (PPV), sensitivity, overall agreement, kappa, and chance corrected concordance (CCC) assessed individual level agreement. Cause-specific mortality fraction (CSMF) accuracy and Spearman's rank correlation assessed population level agreement.</p><p><strong>Results: </strong>A total of 5386 VA records were analysed. PCVA and CCVAs all identified HIV/AIDS as the leading COD. CCVA PPV and sensitivity, based on confidence intervals, were comparable except for HIV/AIDS, TB, maternal, diabetes mellitus, other cancers, and some injuries. CCVAs performed well for identifying perinatal deaths, road traffic accidents, suicide and homicide but poorly for pneumonia, other infectious diseases and renal failure. Overall agreement between CCVAs and PCVA for the top single cause (48.2-51.6) indicated comparable weak agreement between methods. Overall agreement, for the top three causes showed moderate agreement for InterVA (70.9) and InSilicoVA (73.8). Agreement based on kappa (-0.05-0.49)and CCC (0.06-0.43) was weak to none for all algorithms and groups. CCVAs had moderate to strong agreement for CSMF accuracy, with InterVA-5 highest for neonates (0.90), Tariff 2.0 highest for adults (0.89) and males (0.84), and InSilicoVA highest for females (0.88), elders (0.83) and out-of-facility deaths (0.85). Rank correlation indicated moderate agreement for adults (0.75-0.79).</p><p><strong>Conclusions: </strong>Whilst CCVAs identified HIV/AIDS as the leading COD, consistent with PCVA, there is scope for improving the algorithms for use in South Africa.</p>","PeriodicalId":49197,"journal":{"name":"Global Health Action","volume":null,"pages":null},"PeriodicalIF":2.2000,"publicationDate":"2023-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10795603/pdf/","citationCount":"0","resultStr":"{\"title\":\"Agreement between cause of death assignment by computer-coded verbal autopsy methods and physician coding of verbal autopsy interviews in South Africa.\",\"authors\":\"Pam Groenewald, Jason Thomas, Samuel J Clark, Diane Morof, Jané D Joubert, Chodziwadziwa Kabudula, Zehang Li, Debbie Bradshaw\",\"doi\":\"10.1080/16549716.2023.2285105\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>The South African national cause of death validation (NCODV 2017/18) project collected a national sample of verbal autopsies (VA) with cause of death (COD) assignment by physician-coded VA (PCVA) and computer-coded VA (CCVA).</p><p><strong>Objective: </strong>The performance of three CCVA algorithms (InterVA-5, InSilicoVA and Tariff 2.0) in assigning a COD was compared with PCVA (reference standard).</p><p><strong>Methods: </strong>Seven performance metrics assessed individual and population level agreement of COD assignment by age, sex and place of death subgroups. Positive predictive value (PPV), sensitivity, overall agreement, kappa, and chance corrected concordance (CCC) assessed individual level agreement. Cause-specific mortality fraction (CSMF) accuracy and Spearman's rank correlation assessed population level agreement.</p><p><strong>Results: </strong>A total of 5386 VA records were analysed. PCVA and CCVAs all identified HIV/AIDS as the leading COD. CCVA PPV and sensitivity, based on confidence intervals, were comparable except for HIV/AIDS, TB, maternal, diabetes mellitus, other cancers, and some injuries. CCVAs performed well for identifying perinatal deaths, road traffic accidents, suicide and homicide but poorly for pneumonia, other infectious diseases and renal failure. Overall agreement between CCVAs and PCVA for the top single cause (48.2-51.6) indicated comparable weak agreement between methods. Overall agreement, for the top three causes showed moderate agreement for InterVA (70.9) and InSilicoVA (73.8). Agreement based on kappa (-0.05-0.49)and CCC (0.06-0.43) was weak to none for all algorithms and groups. CCVAs had moderate to strong agreement for CSMF accuracy, with InterVA-5 highest for neonates (0.90), Tariff 2.0 highest for adults (0.89) and males (0.84), and InSilicoVA highest for females (0.88), elders (0.83) and out-of-facility deaths (0.85). Rank correlation indicated moderate agreement for adults (0.75-0.79).</p><p><strong>Conclusions: </strong>Whilst CCVAs identified HIV/AIDS as the leading COD, consistent with PCVA, there is scope for improving the algorithms for use in South Africa.</p>\",\"PeriodicalId\":49197,\"journal\":{\"name\":\"Global Health Action\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":2.2000,\"publicationDate\":\"2023-12-31\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10795603/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Global Health Action\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1080/16549716.2023.2285105\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2023/12/1 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q2\",\"JCRName\":\"PUBLIC, ENVIRONMENTAL & OCCUPATIONAL HEALTH\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Global Health Action","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1080/16549716.2023.2285105","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2023/12/1 0:00:00","PubModel":"Epub","JCR":"Q2","JCRName":"PUBLIC, ENVIRONMENTAL & OCCUPATIONAL HEALTH","Score":null,"Total":0}
Agreement between cause of death assignment by computer-coded verbal autopsy methods and physician coding of verbal autopsy interviews in South Africa.
Background: The South African national cause of death validation (NCODV 2017/18) project collected a national sample of verbal autopsies (VA) with cause of death (COD) assignment by physician-coded VA (PCVA) and computer-coded VA (CCVA).
Objective: The performance of three CCVA algorithms (InterVA-5, InSilicoVA and Tariff 2.0) in assigning a COD was compared with PCVA (reference standard).
Methods: Seven performance metrics assessed individual and population level agreement of COD assignment by age, sex and place of death subgroups. Positive predictive value (PPV), sensitivity, overall agreement, kappa, and chance corrected concordance (CCC) assessed individual level agreement. Cause-specific mortality fraction (CSMF) accuracy and Spearman's rank correlation assessed population level agreement.
Results: A total of 5386 VA records were analysed. PCVA and CCVAs all identified HIV/AIDS as the leading COD. CCVA PPV and sensitivity, based on confidence intervals, were comparable except for HIV/AIDS, TB, maternal, diabetes mellitus, other cancers, and some injuries. CCVAs performed well for identifying perinatal deaths, road traffic accidents, suicide and homicide but poorly for pneumonia, other infectious diseases and renal failure. Overall agreement between CCVAs and PCVA for the top single cause (48.2-51.6) indicated comparable weak agreement between methods. Overall agreement, for the top three causes showed moderate agreement for InterVA (70.9) and InSilicoVA (73.8). Agreement based on kappa (-0.05-0.49)and CCC (0.06-0.43) was weak to none for all algorithms and groups. CCVAs had moderate to strong agreement for CSMF accuracy, with InterVA-5 highest for neonates (0.90), Tariff 2.0 highest for adults (0.89) and males (0.84), and InSilicoVA highest for females (0.88), elders (0.83) and out-of-facility deaths (0.85). Rank correlation indicated moderate agreement for adults (0.75-0.79).
Conclusions: Whilst CCVAs identified HIV/AIDS as the leading COD, consistent with PCVA, there is scope for improving the algorithms for use in South Africa.
期刊介绍:
Global Health Action is an international peer-reviewed Open Access journal affiliated with the Unit of Epidemiology and Global Health, Department of Public Health and Clinical Medicine at Umeå University, Sweden. The Unit hosts the Umeå International School of Public Health and the Umeå Centre for Global Health Research.
Vision: Our vision is to be a leading journal in the global health field, narrowing health information gaps and contributing to the implementation of policies and actions that lead to improved global health.
Aim: The widening gap between the winners and losers of globalisation presents major public health challenges. To meet these challenges, it is crucial to generate new knowledge and evidence in the field and in settings where the evidence is lacking, as well as to bridge the gaps between existing knowledge and implementation of relevant findings. Thus, the aim of Global Health Action is to contribute to fuelling a more concrete, hands-on approach to addressing global health challenges. Manuscripts suggesting strategies for practical interventions and research implementations where none already exist are specifically welcomed. Further, the journal encourages articles from low- and middle-income countries, while also welcoming articles originated from South-South and South-North collaborations. All articles are expected to address a global agenda and include a strong implementation or policy component.