基础医学与临床 ›› 2025, Vol. 45 ›› Issue (7): 926-932.doi: 10.16352/j.issn.1001-6325.2025.07.0926

• 研究论文 • 上一篇    下一篇

应用贝叶斯网络识别2021—2022年国家基本公共卫生服务项目中的高血压控制关键因素

李丹颖1, 郭晓璟1, 朱晓磊2, 司向2, 张晓畅2*, 万霞1*   

  1. 1.中国医学科学院基础医学研究所 北京协和医学院基础学院 呼吸和共病全国重点实验室,北京 100005;
    2.中国疾病预防控制中心 慢病和老龄健康管理处,北京 100050
  • 收稿日期:2025-03-20 修回日期:2025-04-23 出版日期:2025-07-05 发布日期:2025-06-24
  • 通讯作者: *xiawan@ibms.pumc.edu.cn;zhangxc@chinacdc.cn
  • 基金资助:
    全国重点实验室专项经费(2060204);中国医学科学院医学与健康科技创新工程(2023-12M-2-001)

Identifying key factors of hypertension control using Bayesian networks in the 2021—2022 National Basic Public Health Service Project

LI Danying1, GUO Xiaojing1, ZHU Xiaolei2, SI Xiang2, ZHANG Xiaochang2*, WAN Xia1*   

  1. 1. State Key Laboratory of Respiratory Health and Multimorbidity, Institute of Basic Medical Sciences CAMS, School of Basic Medicine PUMC, Beijing 100005;
    2. Department of Chronic Disease and Aging Health Management,Chinese Center for Disease Control and Prevention, Beijing 100050, China
  • Received:2025-03-20 Revised:2025-04-23 Online:2025-07-05 Published:2025-06-24
  • Contact: *xiawan@ibms.pumc.edu.cn;zhangxc@chinacdc.cn

摘要: 目的 通过构建贝叶斯网络(BN)模型,探讨影响国家基本公共卫生服务慢性病患者管理评估项目中高血压患者血压控制情况的因素及之间的网络关系,为基本公共卫生服务中高血压的综合管理提供科学依据。方法 选取2021—2022年在中国国家基本公共卫生服务慢性病患者管理评估项目中基于中国东、中和西部地区的八省(自治区)收集的5 577例高血压患者,对患者个体情况及社区高血压管理情况等相关资料收集。采用Logistic回归分析模型进行血压控制影响因素的筛选,并使用BN描述各因素间的关系,展示在国家基本公共卫生项目中对血压控制的关键因素。结果 Logistic回归分析结果显示,在进行多因素筛选后,城乡、教育程度、饮酒、锻炼身体、超重/肥胖等个体因素以及社区医生建议减盐、戒烟等管理因素与患者血压控制情况有显著相关性(P<0.05)。构建的BN模型显示,共产生22条有向边,其中城市以及高血压知识知晓好的患者血压控制情况好,社区医生管理、服务则直接影响患者的行为习惯,均未直接指向血压控制。结论 一方面还应更加注重城乡之间的差异以及发挥对高血压患者的知识教育的积极作用,另一方面不管是患者个体行为习惯还是社区医生服务质量都需要更进一步的提高,才能提升血压控制情况。

关键词: 高血压, 贝叶斯网络, 国家基本公共卫生服务项目, 高血压影响因素

Abstract: Objective To explore factors affecting blood pressure control in chronic disease patients in China′s national basic public health service chronic disease patient management program and to find their relationships with Bayesian network(BN) model, in order to provide a scientific basis for comprehensive hypertension management. Methods 5 577 Hypertensive patients were selected from eight provinces(including autonomous regions) covering eastern, central and western parts of China during a survey from 2021 to 2022. Researchers collected individual and community-management data to screen influencing factors by Logistic regression, and to describe factor dependencies and to identify key determinants of blood pressure control with BN in. blood pressure control. Results Logistic regression revealed that urban/rural status, education, alcohol use, exercise, overweight/obesity and community-doctor advice on salt reduction, smoking cessation were significantly associated with blood pressure control(P<0.05). The BN model identified 22 directed edges showing that urban residence and good hypertension knowledge were more correlated with better control, while community-doctor management and services directly affected patient lifestyle habits but not blood pressure control. Conclusions Research should focus more on urban-rural disparities and hypertension education. Additionally, improving patient habits and community-doctor services is essential for better blood pressure control.

Key words: hypertension, Bayesian network, National Basic Public Health Service Project, hypertension influencing factor

中图分类号: