Proportional Hazards Changepoint Models in Survival Analysis
提出一种通过分段分布参数化基线风险的比例风险分析方法，用GLIM和迭代法做最大似然估计，两段威布尔拟合优于简单威布尔模型，适用于存在变点的生存数据。
This paper describes an approach to proportional hazards analysis of survival data with covariates by parametric modelling of the base-line hazard in terms of piecewise distributions. Maximum likelihood estimation using GLIM and an iterative method is straightforward. Applications of the method and its use with competing risks are given, in which a two-piece Weibull fit is clearly superior to the simple Weibull model. Wide generality results from the fact that any given montonic increasing transformation may be applied to the base-line hazard parameter. It can be expected that piecewise models of this kind will usefully describe many proportional hazards survival processes involving changepoints at which the ruling conditions suddenly alter.