职业与健康 ›› 2025, Vol. 41 ›› Issue (7): 976-980.

• 论著 • 上一篇    下一篇

基于logistic回归和决策树的社区居民传染病健康素养影响因素分析

岳淑玲, 张微, 马建颖, 张晓松, 路瑜臻, 张卫红   

  1. 华北理工大学附属医院心内科,河北 唐山 063000
  • 收稿日期:2024-07-11 修回日期:2024-07-22 出版日期:2025-04-01 发布日期:2025-12-17
  • 作者简介:岳淑玲,女,副主任护师,主要从事临床护理与护理管理工作。

Analysis of factors influencing infectious disease health literacy among community residents based on logistic regression and decision tree

YUE Shuling, ZHANG Wei, MA Jianying, ZHANG Xiaosong, LU Yuzhen, ZHANG Weihong   

  1. Department of Cardiology,North China University of Science and Technology Affiliated Hospital,Tangshan,Hebei 063000,China
  • Received:2024-07-11 Revised:2024-07-22 Online:2025-04-01 Published:2025-12-17

摘要: 目的 运用决策树和logistic回归模型对社区居民传染病健康素养影响因素进行分析,为社区护士采取措施提供参考。方法 2022年3—6月,采用便利抽样方法选取1 795名社区居民为研究对象,进行问卷调查,通过构建决策树和logistic回归模型分析社区居民传染病健康素养的影响因素,比较2种模型的预测效果。结果 Logistic回归模型结果显示,文化程度、月收入、是否参加健康教育、互联网使用频率、健康档案建立情况及年龄是社区居民传染病健康素养的影响因素(均P<0.05)。决策树模型分析结果显示,文化程度、领悟社会支持、互联网使用频率、是否参加健康教育、家庭功能、自评身体健康状况及年龄为社区居民传染病健康素养的影响因素(均P<0.05)。Logistic回归模型的灵敏度为75.5%,特异度为82.0%,受试者工作特征曲线(ROC曲线)下面积为0.859;决策树模型灵敏度为78.9%,特异度为76.4%,ROC曲线下面积为0.851,2个模型的ROC曲线下面积差异无统计学意义(Z=1.640,P>0.05)。结论 Logistic回归模型与决策树模型的结合,可以对社区居民传染病健康素养影响因素提供较为全面的解释且具有较高的应用价值,为社区护士采取针对性预防和护理干预提供依据。

关键词: Logistic回归, 决策树, 社区居民, 传染病健康素养

Abstract: Objective To conduct an in-depth analysis of the influencing factors of infectious disease health literacy among community residents using decision trees and logistic regression models,and provide reference for community nurses to take measures. Methods From March to June 2022,a total of 1 795 community residents were selected by convenience sampling for questionnaire survey. The influencing factors of infectious disease health literacy of community residents were analyzed by constructing decision tree and logistic regression models,and the prediction effects of the two models were compared. Results The logistic regression model results showed that education level,monthly income,whether to participate in health education,frequency of Internet use,establishment of health file,and age were the factors affecting the infectious disease health literacy among community residents(all P<0.05). The results of decision tree model analysis showed that education level,perceived social support, frequency of Internet use,whether to participate in health education,family function,self-rated physical health status,and age were the factors influencing the health literacy of community residents(all P<0.05). The sensitivity of the logistic regression model was 75.5%,the specificity was 82.0%,and the area under the ROC curve was 0.859. The sensitivity of the decision tree model was 78.9%,the specificity was 76.4%,and the area under the ROC curve was 0.851. There was no statistically significant difference in the area under the ROC curve between the two models(Z=1.640,P>0.05). Conclusion The combination of logistic regression model and decision tree model can provide a more comprehensive explanation for the influencing factors of infectious disease health literacy in community residents and has a high application value,and provide the basis for community nurses to take targeted prevention and nursing intervention.

Key words: Logistic regression, Decision tree, Community residents, Infectious disease health literacy

中图分类号: