Statistics and Its Interface

Volume 2 (2009)

Number 4

Bayesian cure rate model accommodating multiplicative and additive covariates

Pages: 513 – 521



Luis E. Nieto-Barajas (Department of Statistics, ITAM, Progreso Tizapan, Mexico)

Guosheng Yin (Department of Statistics and Actuarial Science, The University of Hong Kong, Hong Kong)


We propose a class of Bayesian cure rate models by incorporating a baseline density function as well as multiplicative and additive covariate structures. Our model naturally accommodates zero and non-zero cure rates, which provides an objective way to examine the existence of a survival fraction in the failure time data. An inherent parameter constraint needs to be incorporated into the model formulation due to the additive covariates. Within the Bayesian paradigm, we take a Markov gamma process prior to model the baseline hazard rate, and mixture prior distributions for the parameters in the additive component of the model. We implement a Markov chain Monte Carlo computational scheme to sample from the full conditional distributions of the posterior. We conduct simulation studies to assess the estimation and inference properties of the proposed model, and illustrate it with data from a bone marrow transplant study.


additive hazards model, cure rate model, Markov gamma process, mixture prior, proportional hazards model, semiparametric method, survival fraction

2010 Mathematics Subject Classification

Primary 62N01. Secondary 62N03.

Full Text (PDF format)