The Identifiability of Dependent Competing Risks Models Induced by Bivariate Frailty Models

• Activity Number: 489
• Type: Topic Contributed
• Date/Time: Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
• Sponsor: International Chinese Statistical Association
• Author(s): Antai Wang*+ and Krishnendu Chandra and Ruihua Xu and Junfeng Sun
• Companies: New Jersey Institute of Technology and Columbia University and NIH and NIH
• Keywords: Competing risks models ; Dependent Censoring ; Archimedean Copula Models ; Identifiability Conditions ; Discrete Covariates ; Bivariate Frailty Models

Abstract

In this talk, we propose to use a special class of bivariate frailty models to study dependent censored data. The proposed models are closely linked to Archimedean copula models. We give sufficient conditions for the identifiability of this type of competing risks models. The proposed conditions are derived based on a property shared by Archimedean copula models and satisfied by several well known bivariate frailty models. It turns out that our conditions can be applied to competing risks models with discrete covariates. Under the proposed identifiability conditions, EM algorithm provides us with consistent estimates of the unknown parameters. Simulation studies have shown that our estimation procedure works quite well. We fit a dependent censored leukemia data set using the Clayton copula model and end our paper with some discussions.