健康状况自评问卷信度和效度分析

Reliability and Validity Analysis for Questionnaire of Self-report Health Status

[目的]评价基于世界卫生组织的世界健康调查(WHS)问卷和儿童健康问卷(CHQ)的健康状况自评问卷信度和效度,为开展上海市人群健康水平综合评价奠定技术基础。[方法]在上海市城乡结合部一个居民户数大于400户的居民小区中通过随机抽样获得含有5～15岁儿童的上海户籍家庭80户(371人)作为调查对象。调查问卷以WHS问卷为基础,加入儿童健康问卷-家长卷28(CHQ-PF28)中的相关内容;被调查对象在经培训的调查员指导下完成其健康状况自评。应用重测相关系数和Cronbachα系数评价问卷信度;应用个体功能健康状况评分和总体健康状况的单因素方差评价区分效度;应用主成分法的因子分析评价问卷结构效度。[结果]反映个体健康状况的各维度评分重测相关系数大于0.5,大部分同一维度下各条目的Cronbach a系数均大于0.6,信度较好。自报患病或不良状态的评分明显高于未报患病或不良状态者(P<0.00),区分效度较好。测量成人健康条目反映出8个载荷较大的因子,测量儿童健康条目反映出5个载荷较大的因子,与问卷设计接近。[结论]基于WHS和CHQ问卷的健康状况自评问卷可应用于5岁及以上人群健康状况的综合评价。

[ Objective ] To evaluate the reliability and validity of seE-report health status questionnaire based on World Health Survey （ WHS ） questionnaire and Children Health Questionnaire （ CHQ ）. [ Methods ] A sample composed of 80 Shanghai permanent resident families owning at least one child aged 5-15 years old was randomly sampled from a residential area of more than 400 families near the suburb of Shanghai. A questionnaire was established based on WHS questionnaire with additional content about health status scale adapted from Children Health Questionnaire-Parents 28. Respondents finished health status self-report according to guidance from investigators. Retest correlation coefficient and Cronbach a coefficient were used to evaluate the reliability. One-way ANOVA between general health status and score for health status scaling was used to evaluate differential validity. Factor analysis of principal component was used to evaluate the validity. [ Results ] Retest correlation coefficients for scores of all dimensions were more than 0.5. Cronbach a coefficient for most of dimensions were over 0.6. Score for the group of self-reported prevalence or unhealthy status was significantly higher than that of not reported prevalence or unhealthy status （ P 〈 0.00 ）. Items for adult health status project showed 8 factors of bigger loading and those for child health status project showed 5 factors of bigger loading, which were in accordance with the design of the questionnaire to some extent. [ Conlusion ] The questionnaire may be applied to comprehensively evaluate the health of population aged over 5 years.