多水平Poisson模型在公共卫生领域中的应用  

The Application of Multilevel Poisson Models in Public Health

作　　者：丁丞[1,2] 徐凯进[2] 卢汉体 初菁菁 沈毅[1] 

机构地区：[1]浙江大学医学院,杭州浙江310058 [2]浙江大学医学院附属第一医院传染病诊治国家重点实验室,杭州浙江310003

出　　处：《生物数学学报》2016年第2期195-202,共8页Journal of Biomathematics

基　　金："十二五"国家科技重大专项(2013ZX10004904)

摘　　要：本文结合实例探讨多水平Poisson模型在公共卫生领域的应用,根据结核病患病数据资料的层次结构特征,采用多水平Possion模型拟合不同模型并分析比较.得出性别和年龄因素对于结核患病率有影响,不同地区间结核患病率也有差别.具有层次结构特征的数据适宜采用多水平模型进行分析,多水平Possion模型适合用于以人群为基础的公共卫生数据资料的分析.To discuss the applying of multilevel Poisson models in public health with practical cases. Analyse and constrast the different fitting models with tuberculosis data, which had a characteristic of hierarchy structure. Gender and age had an influence on the prevalence of tuberculosis, and the prevalences of tuberculosis in different regions were not the same. Multilevel models are suitable methods when dealing with the data which has a structure of hierarchy, the multilevel Poisson models are applicable to the data of public heath based on population.

关 键 词：多水平 Poisson模型 结核病 

分 类 号：R183.3[医药卫生—流行病学]