论文部分内容阅读
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。
Application 1: M-matched case-control logistic regression model Re-sampling 14-year-old prospective data of high-risk groups of Qidong liver disease, the relative risk of three risk factors such as HBsAg, family history of hepatocellular carcinoma and type of hepatic disease were screened out. It is estimated that the result is consistent with the estimation of the Poisson regression model and shows that the 1:4 matching effect is better.