I:M配对病例一对照研究中条件Logistic回归分析——启东县肝病高危人群肝癌危险因素的相对危险度估计

Analysis of Conditional Logistic Regression Model for 1: M Matched Pair Case—Control. Study: The Estimation of Relative Risks of: Liver Cancer Facters in Hepatic. High Risk Population in Qidong County.

应用1:M匹配病例一对照条件Logistic回归模型对启东肝病高危人群14年前瞻资料进行再抽样研究,所筛选出的HBsAg、肝癌家族史、原肝病类型等三个危险因素的相对危险度作了估计,其结果与Poisson回归模型的估计一致,并显示以1:4匹配效果为佳。

This paper deseribes an approach for analyzing 1: M matched (?)air case-control data sampling from a fourteen-years prospective study of a hepatic high risk population in Qidong County. By using conditional togistic regression, the odds ratios of three liver cancer risk factors, namely, HBsAg, family history of liver cancer, and the clinical category of the original hepatopathy are estimated, The result showed that parameter estimations of conditional logistic model are agreeable with those of poisson regression model and it is also revealed that the effect of 1: 4 matched pair case-control design is the best.