Development of a risk assessment model for multimorbidity of diabetes, hypertension, and coronary heart disease with XGBoost in Primary Care in Shanghai, China: Retrospective Study (Preprint)

Author:

Chen Ning,Ding Yan,Zhu Wenqing,Fan Feng,Jin Hua,Chen Yanying,Liu Jing,Chen Lin,Wan Xiaoguang,Guo Jing,Yang Zaijia,Sun Lei,Shan Houqian,Yu Dehua,Shi Jianwei,Wang Zhaoxin

Abstract

BACKGROUND

Multimorbidity has become an escalating health issue worldwide. Among various heterogeneous patterns of common multimorbidity, the cardiometabolic cluster is one of the most prevalent ones. Evaluating the risk of multimorbidity of the cardiometabolic pattern from the perspective of general practice has become a crucial agenda in primary care.

OBJECTIVE

The aim of this study was to develop a comprehensive risk assessment model for multimorbidities of diabetes, hypertension, and coronary heart disease of the elderly in the community based on big data in Shanghai, China.

METHODS

Retrospective data of 40,261 residents from 47 community health centers from 2017 to 2019 were extracted, including the residents' health records and health examination data, the hospital management information system data (HIS), and the imaging examination database. The machine learning algorithm, XGBoost was utilized for the construction of the comprehensive risk assessment model for multimorbidities of diabetes, hypertension, and coronary heart disease. Area under the receiver operating characteristic curve, accuracy, precision, recall, F1 Score and other indicators were used for model evaluation.

RESULTS

A total of 46 features was incorporated into the final comprehensive risk assessment model for multimorbidities of diabetes, hypertension, and coronary heart disease. The micro-average AUC value of the optimal model was 0.822. The macro average AUC value was 0.795. The weighted average AUC value was 0.784. These parameters showed a high superiority of the constructed model.

CONCLUSIONS

The comprehensive risk assessment model for multimorbidities of diabetes, hypertension, and coronary heart disease based on XGBoost integrated medical and public health data of residents in community and uncovered multidimensional risk factors from four dimensions. It is of high directive value to implement the comprehensive risk assessment in improving the corresponding health management strategies for residents with multimorbidities.

Publisher

JMIR Publications Inc.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3