摘要
本文建立了追踪研究中致癌因素残余效应及其累积的数学模型,讨论了个体发病概率随多个具有残余效应的因素变化的加法模型和乘法模型,提出了允许暴露水平以等级形式表示的参数估计和检验,以及模型拟合优度检验的方法,并推广到1:R 匹配的病例-对照研究。通过所建模型讨论了现有文献中难以解决的一些时效问题,并分析了职业膀胱癌病例-对照研究的一份实际资料。
Mathematical models for time-dependent residual effect of risk factorand its accumulation in follow-up cancer study are developed;both of additivemodel and multiplicative model describing the relationship between incidencerate and multiple covariates with residual effect are discussed and the relevantstatistical methods for parameter estimation,hypothesis test andgoodness-of-fit test are provided where the exposure levels are allowed to begiven in an ordinal form.Especially,these models are generalized to 1:Rmatched case-control study and executed on the IBM PC/XT system.Someproblems relating to time effect that could be hardly dealt with in the past arediscussed under these models.All procedures are applied to a case-controlstudy of occupational bladder cancer.
出处
《中华劳动卫生职业病杂志》
CAS
CSCD
1990年第6期321-324,共4页
Chinese Journal of Industrial Hygiene and Occupational Diseases
基金
国家七五攻关项目"适应于肿瘤预防研究的生物统计学理论与技术"
关键词
膀胱癌
时效分析
致癌因素
Time-effect analysis
Residual effect
Case-control study
Occupational cancer
