Extending Fine and Gray Model: New Approach for Competing Risks Analysis

• Activity Number: 292
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Anna Bellach*+ and Jason Fine and Ludger Rüschendorf and Michael Kosorok
• Companies: University of Copenhagen and University of North Carolina at Chapel Hill and Albert Ludwigs University and University of North Carolina at Chapel Hill
• Keywords: competing risks ; Fine and Gray model ; semiparametric transformation models ; pseudo likelihood function ; NPMLE ; time varying covariates

Abstract

Fine and Gray have established a version of the Cox's model for the hazard rate of the subdistribution in a competing risks setting. As the proportional hazards assumption does not hold in general, it is of interest also to consider other regression models for competing risks settings. We introduce a new pseudo likelihood function that can be used to derive estimators for a general class of models. For the class of semiparametric transformation models proposed by Zeng and Lin we prove the consistency and asymptotic normality of the estimators. In simulation studies and real data examples our method works well. The Fine and Gray model is a special case of our general approach.