Objective To investigate the effect of proxy-report in the self-report heath questionnaire (EQ-5D) in the health service survey.Methods Chi-square test and t-test were used to Compare the demographic characteristic and heath behavior between proxy-report and self-report groups.Multivariate logistic models were used to explore the effect of proxy-report on the risk of reporting health problems in EQ-5D dimensions,adjusting other covariates.Results The proxy-report group were mostly young,unmarried and migrant workers.There are significant differences between proxy-report and the self-report groups in the fields of smoking,drinking,sick,hospital and other lifestyle and healthy status.The results of the multivariate logistic models showed that the OR(95％ CI) of proxy-report factor in the Mobility、Self-Care、Usual Activities、Pain/Discomfort、Anxiety/Depression in EQ-5D were respectively I.6(1.5,1.8) 、2.3 (2.0,2.6) 、2.0 (1.8,2.2)、1.2 (1.1,1.3)、1.4 (1.3,1.6).Conclusion The proxy-report group is apt to rated the health status of the subjects who are reported more serious.Therefore,it is necessary to control the selection bias bought about from proxy-report both at the stages of survey implementation and data analysis.%目的 探讨国家卫生服务调查中代答对自报健康问卷(EQ-5D)调查结果的影响.方法 比较自答人群与代答人群的人口学特征、健康行为习惯并进行单变量差异性分析.应用logistic多因素模型研究代答对EQ-5D量表五个维度调查结果的影响.结果 代答人群多为年轻、未婚或外出务工人员.在吸烟、饮酒、患病、住院等生活方式和健康状况方面,代答人群与自答人群存在差异.在EQ-5D量表的行动、自我照顾、日常活动、疼痛/不适、焦虑/抑郁五个维度上,多因素logistic回归模型结果显示,代答的OR值及95％可信区间分别为1.64(1.48,1.81)、2.29(2.03,2.59)、1.96(1.77,2.18)、1.21(1.11,1.32)、1.42(1.29,1.56).结论 在自报健康调查中,代答者明显地存在将被代答者的健康问题严重化的倾向,应该从调查实施和数据分析两个阶段采取有效措施降低代答带来的偏倚.
展开▼
