Bayesian Modeling and Inference of Survival Data with Semi-Competing and Competing Risks

• Activity Number: 7
• Type: Invited
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Ming-Hui Chen*+ and Mario de Castro and Yuanye Zhang and Anthony V. D'Amico
• Companies: University of Connecticut and Universidade de São Paulo and Novartis and Brigham & Women's Hospital
• Keywords: Cure rate model ; Fully specified subdistribution ; Identifiability ; Markov chain Monte Carlo ; Prostate cancer ; Semi-Markov model

Abstract

Semi-competing risks data include the time to a nonterminating event and the time to a terminating event while competing risks data include the time to more than one terminating events. Our study is motivated from a prostate cancer study, which has one nonterminating event and two terminating events with both semi-competing risks and competing risks present. Due to the complication of two non-informative censoring times for the nonterminating event and the terminating events, the existing semi-competing risks models may not be identifiable. In this paper, we propose a new model based on the cure rate model and the fully specified subdistribution (FS) competing risks model along with a time dependent covariate for this type of data. The proposed model is more parsimonious than the conventional semi-competing risks models and, hence, it is more identifiable for this type of data. The properties of the proposed model are examined and an efficient Markov chain Monte Carlo sampling algorithm is also developed. We apply the proposed methodology to a detailed case study in prostate cancer.