摘要：In this paper we present the standard mixture model with a frailty term. This model has the advantage in relation the usual survival models, in the sense of incorporate the heterogeneity of two subpopulations (immune and susceptible) to the event of interest. We consider the model with a frailty term as alternative to modelling long-term survival data. In this mo del a random e.ect, also called frailty, is introduced on the baseline hazard function, in order to control the unobservable heterogeneity of the units under study. A simulation study is present to analyze the frequentists properties of estimation procedures. A cost study in the estimation of the cured rate is discussed. We explore the use of the gamlss package in R as a tool for inference in cure rate models. The developed procedures are illustrated on a real data set.
关键词：Frailty model ; GAMLSS ; Long-term survival models ; Survival analysis ; ; Variances ratio
