传染病疫情初期病死率及危险因素的估计模型

Estimation Model of the Case Fatality Risk and Risk Factors in the Early Stage of Infectious Disease Epidemic

目的研究传染性疾病致命性参数病死率的估计,特别是有多个风险因素存在时,不同分组下的病死率估计以及风险因素筛选问题。方法基于竞争风险理论建立病死率估计回归模型,通过轮廓似然等技术估计模型参数。结果将该回归模型应用于2013年H7N9禽流感数据分析,估计不同风险因素下的病死率,并通过BICcr准则发现性别间没有差异,而高年龄(60岁为界)组是一个有意义的风险因素。结论本文介绍的病死率估计回归模型除了可以直接估计病死率值外,也能进行相关风险因素的甄别。

Objective To study estimation of the case fatality risk of infectious diseases,especially,the case fatality risk under different groups and the screening of variables when there are multiple risk factors.Methods We introduced a regression model of case fatality risk based on the competition risks theory.To estimate regression coefficients and screen risk factors using profile likelihood technique and the BICcr criteria,respectively.Results The regression model was applied to analyze the H7 N9 avian influenza data in 2013 and to estimate the case fatality risk under different risk factors.The results show no difference between the genders,however,the age(cut-off:60 years old)is a significative risk factor.Conclusion The proposed regression model can estimate the case fatality risk and find the significative risk factors besides.