论文部分内容阅读
目的使用不同方法分析浙江省城市居民2周患病率的影响因素。方法分别使用单水平Logistic回归和两水平Logistic回归,对浙江省第四次卫生服务调查的资料进行分析。结果由单水平Logistic回归得出的差异有统计学意义的一些影响因素在两水平Logistic回归模型中变得不明显。2种方法均得出的影响2周患病率的因素(P<0.05)有:户主受教育程度、个人受教育程度、就业状况、是否患有慢性病。结论卫生服务调查数据存在结构层次,两水平模型考虑了数据的层次结构,得出的影响因素比传统回归模型更可靠,因此认为使用两水平Logistic回归分析更为恰当。
Objective To analyze the influential factors of prevalence of 2-week urban residents in Zhejiang Province using different methods. Methods The single-level Logistic regression and two-level Logistic regression were used to analyze the data of the fourth health service survey in Zhejiang Province. Results Some of the significant differences between the two logistic regression models were statistically insignificant. The factors influencing 2-week prevalence (P <0.05) of the two methods were: education level of the head of household, education level of individual, employment status and whether or not suffering from chronic diseases. Conclusions There is a hierarchy of health services survey data. The two-level model considers the hierarchy of data, and the influencing factors are more reliable than traditional regression models. Therefore, it is more appropriate to use two-level Logistic regression analysis.