Joseph Ibrahim, Harvard University

A New Bayesian Model For Survival Data With a Surviving Fraction

We consider Bayesian methods for right censored survival data for populations with a surviving (cure) fraction. We propose a model which is quite different from the standard mixture model for cure rates. We provide a natural motivation and interpretation of the model and derive several novel properties of it. First, we show that the model has a proportional hazards structure, with the covariates depending naturally on the cure rate. Secondly, we derive several properties of the hazard function for the proposed model, and establish mathematical relationships with the mixture model for cure rates. By introducing latent variables, we develop efficient Markov chain Monte Carlo algorithms for sampling from the posterior distribution of the parameters. Prior elicitation is discussed in detail, and classes of noninformative and informative prior distributions are proposed. Several theoretical properties of the proposed priors and resulting posteriors are derived, and comparisons are made to the standard mixture model. A real dataset from a melanoma clinical trial is discussed in detail.