{"title":"预测印度低社会人口指数邦五岁以下儿童死亡率的生存机器学习方法。","authors":"Mukesh Vishwakarma, Gargi Tyagi, Rehana Vanaja Radhakrishnan","doi":"10.34172/jrhs.9033","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Each year, millions of children under five die globally, with many of these deaths being preventable. The situation is particularly concerning in low sociodemographic index (LSDI) states of India, where the under-five mortality rate is 45 children per 1000 live births. This study aimed to predict under-five mortality and determine related key factors. <b>Study Design:</b> A cross-sectional study.</p><p><strong>Methods: </strong>This study analyzed National Family Health Survey-5 (NFHS-5) data related to 94,202 children from the LSDI states of India. Several survival models were tested, including Cox proportional hazards, random survival forest, and gradient-boosted survival, to identify factors linked to child mortality. Model performance was evaluated using metrics such as the concordance index, integrated Brier score, and time-dependent receiver operating characteristic (ROC) curves.</p><p><strong>Results: </strong>Among the studied children, 4.5% (4,284) died before their fifth birthday. The risk of death was higher in children born to younger (15-25 years) mothers (hazard ratio [HR] = 1.113, 95% confidence interval (CI): 1.034, 1.198; <i>P</i> < 0.001), uneducated mothers (HR = 1.263, 95% CI: 1.098-1.454; <i>P</i> < 0.0001), mothers with a poorer wealth index (HR = 1.719, 95% CI: 1.475-2.003; <i>P</i> < 0.0001), and children with low birth weight (HR = 2.091, 95% CI: 1.934-2.26; <i>P</i> < 0.001). The random survival forest model outperformed in identifying these risk factors.</p><p><strong>Conclusion: </strong>This study highlights the importance of empowering women through education, improving family planning, addressing poverty, and providing equitable healthcare to reduce child mortality. These insights can help shape policies and initiatives to improve the survival and health of children in vulnerable communities.</p>","PeriodicalId":17164,"journal":{"name":"Journal of research in health sciences","volume":"25 3","pages":"e00653"},"PeriodicalIF":1.4000,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12445885/pdf/","citationCount":"0","resultStr":"{\"title\":\"Survival Machine-Learning Approach for Predicting Under-Five Mortality in Low Sociodemographic Index States of India.\",\"authors\":\"Mukesh Vishwakarma, Gargi Tyagi, Rehana Vanaja Radhakrishnan\",\"doi\":\"10.34172/jrhs.9033\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Each year, millions of children under five die globally, with many of these deaths being preventable. The situation is particularly concerning in low sociodemographic index (LSDI) states of India, where the under-five mortality rate is 45 children per 1000 live births. This study aimed to predict under-five mortality and determine related key factors. <b>Study Design:</b> A cross-sectional study.</p><p><strong>Methods: </strong>This study analyzed National Family Health Survey-5 (NFHS-5) data related to 94,202 children from the LSDI states of India. Several survival models were tested, including Cox proportional hazards, random survival forest, and gradient-boosted survival, to identify factors linked to child mortality. Model performance was evaluated using metrics such as the concordance index, integrated Brier score, and time-dependent receiver operating characteristic (ROC) curves.</p><p><strong>Results: </strong>Among the studied children, 4.5% (4,284) died before their fifth birthday. The risk of death was higher in children born to younger (15-25 years) mothers (hazard ratio [HR] = 1.113, 95% confidence interval (CI): 1.034, 1.198; <i>P</i> < 0.001), uneducated mothers (HR = 1.263, 95% CI: 1.098-1.454; <i>P</i> < 0.0001), mothers with a poorer wealth index (HR = 1.719, 95% CI: 1.475-2.003; <i>P</i> < 0.0001), and children with low birth weight (HR = 2.091, 95% CI: 1.934-2.26; <i>P</i> < 0.001). The random survival forest model outperformed in identifying these risk factors.</p><p><strong>Conclusion: </strong>This study highlights the importance of empowering women through education, improving family planning, addressing poverty, and providing equitable healthcare to reduce child mortality. These insights can help shape policies and initiatives to improve the survival and health of children in vulnerable communities.</p>\",\"PeriodicalId\":17164,\"journal\":{\"name\":\"Journal of research in health sciences\",\"volume\":\"25 3\",\"pages\":\"e00653\"},\"PeriodicalIF\":1.4000,\"publicationDate\":\"2025-06-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12445885/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of research in health sciences\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.34172/jrhs.9033\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"PUBLIC, ENVIRONMENTAL & OCCUPATIONAL HEALTH\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of research in health sciences","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.34172/jrhs.9033","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"PUBLIC, ENVIRONMENTAL & OCCUPATIONAL HEALTH","Score":null,"Total":0}
Survival Machine-Learning Approach for Predicting Under-Five Mortality in Low Sociodemographic Index States of India.
Background: Each year, millions of children under five die globally, with many of these deaths being preventable. The situation is particularly concerning in low sociodemographic index (LSDI) states of India, where the under-five mortality rate is 45 children per 1000 live births. This study aimed to predict under-five mortality and determine related key factors. Study Design: A cross-sectional study.
Methods: This study analyzed National Family Health Survey-5 (NFHS-5) data related to 94,202 children from the LSDI states of India. Several survival models were tested, including Cox proportional hazards, random survival forest, and gradient-boosted survival, to identify factors linked to child mortality. Model performance was evaluated using metrics such as the concordance index, integrated Brier score, and time-dependent receiver operating characteristic (ROC) curves.
Results: Among the studied children, 4.5% (4,284) died before their fifth birthday. The risk of death was higher in children born to younger (15-25 years) mothers (hazard ratio [HR] = 1.113, 95% confidence interval (CI): 1.034, 1.198; P < 0.001), uneducated mothers (HR = 1.263, 95% CI: 1.098-1.454; P < 0.0001), mothers with a poorer wealth index (HR = 1.719, 95% CI: 1.475-2.003; P < 0.0001), and children with low birth weight (HR = 2.091, 95% CI: 1.934-2.26; P < 0.001). The random survival forest model outperformed in identifying these risk factors.
Conclusion: This study highlights the importance of empowering women through education, improving family planning, addressing poverty, and providing equitable healthcare to reduce child mortality. These insights can help shape policies and initiatives to improve the survival and health of children in vulnerable communities.
期刊介绍:
The Journal of Research in Health Sciences (JRHS) is the official journal of the School of Public Health; Hamadan University of Medical Sciences, which is published quarterly. Since 2017, JRHS is published electronically. JRHS is a peer-reviewed, scientific publication which is produced quarterly and is a multidisciplinary journal in the field of public health, publishing contributions from Epidemiology, Biostatistics, Public Health, Occupational Health, Environmental Health, Health Education, and Preventive and Social Medicine. We do not publish clinical trials, nursing studies, animal studies, qualitative studies, nutritional studies, health insurance, and hospital management. In addition, we do not publish the results of laboratory and chemical studies in the field of ergonomics, occupational health, and environmental health