Event History Models with Risk-Free Fractions for Interval Censored Data

Abstract

In the analysis of survival time, its underlying assumption is that All the study subjects are susceptible to the event. In general, this assumption does not adequately hold in event history analysis for studying time to an event other than death. In the analysis of non-susceptibility as a general phenomenon in biomedical, epidemiological and sociological research, recent studies proposed mixture regression models to formulate both the probability of occurrence of an event and the time distribution of occurrence of the event. For right censored data, Kuk and Chen (1992) generalized Farewell's (1982) logistic-Weibull model to a logistic-Cox model, and Yamaguchi (1992) presented a logistic and generalized gamma mixture model. Exact event times with various combinations of right, left and Interval censored data frequently arise in studies of longitudinal or current-status data. In this talk, we present an extension of the semiparametric logistic-Cox mixture regression model to this more complicated censoring mechanism. A parametric logistic-accelerated failure time mixture regression model for general interval censorship is given in comparison. Both models are illustrated in the analyses of local epidemiological data. Further related works on multi-path mover-stayer models for chronic disease processes are also discussed.