北京市居民传染病健康素养综合指数临界点研究

目的应用受试者工作特征曲线法(ROC曲线法)探索北京市居民传染病健康素养判定标准的最佳界值点。方法分析数据来源于北京市居民传染病健康素养调查。以自报健康水平(健康状况是否为不好或很不好)为"金标准",应用ROC法确定判定标准的最佳界值点。应用非条件Logistic回归法分析是否具备传染病健康素养与自报健康水平之间的相关性。结果 ROC曲线分析结果显示,最佳临界点为得分等于1.55分(总分3分),ROC曲线下面积为0.664(95%CI:0.641-0.688),灵敏度、特异度分别为0.672和0.569。依据最佳临界点将居民划分为是否具备传染病健康素养(传染病健康素养得分达到1.55分)。北京市居民具备传染病健康素养的比例为65.9%。非条件Logistic回归分析结果显示,不具备传染病健康素养者自报健康较差(不好或很不好)的比例高于具备者(OR=1.958,95%CI:1.646-2.329)。结论 ROC曲线法是构建健康素养判定标准的科学方法。本研究确定的判定标准能够明显筛查出不良健康后果,具有一定的筛查价值。

Study on cut-off value for infectious diseases health literacy among residents in Beijing

Objective To explore the cut-off value for infectious diseases health literacy （HL） among residents in Beijing by using receiver operating characteristic （ROC） curve. Methods The data was from Survey on Infectious Dis- eases HL in Beijing. The best cut-off value was detected by using ROC curve with self-reported health status （bad or worse） as golden criterion. Logistic regression model was used to analyze the relationship between the level of HL and self- reported health status. Results The results of ROC curves demonstrated the best cut-off value was 1.55 points （ total points： 3）, and the area under ROC curve was 0. 664 （95% CI： 0. 641 -0. 688）. Sensitivity and specificity was 0. 672 and O. 569, respectively. Residents were categorized into satisfied infectious diseases HL group （scored above 1.55 points） and unsatisfied group （scored less than 1.55 points）. According to the best cutoff value （1.55 points）, 65. 9% of participants had satisfied infectious diseases HL in Beijing. Logistic regression model revealed that the percentage of par- ticipants who self-reported bad or worse health status in unsatisfied HL group was higher than that in satisfied HL group （ OR = 1. 958, 95 % CI： 1. 646 - 2. 329）. Conclusion ROC curve was an appropriate tool to explore the reference values for health literacy. The cut-off value of infectious disease HL found in this study provided a reference for sorting satisfied HL residents and screening their health status.