Weighted Estimation of the Accelerated Failure Time Model in the Presence of Dependent Censoring

• Activity Number: 531
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Youngjoo Cho*+ and Debashis Ghosh
• Companies: The Pennsylvania State University and Penn State University
• Keywords: Bivariate Data ; Failure Times ; Perturbation Method ; Resampling ; U-statistic

Abstract

Independent censoring is one of the crucial assumptions in models of survival analysis. However, this is impractical in many medical studies, where the presence of dependent censoring leads to difficulty in analyzing covariate effects on disease outcomes. The semicompeting risks framework proposed by Lin et al. (1996, Biometrika) and Peng and Fine (2006, Journal of the American Statistical Association) is a suitable approach to handling dependent censoring. These authors proposed estimators based on an artificial censoring technique. However, they did not consider efficiency of their estimators in detail. In this paper, we propose a new weighted estimator for the accelerated failure time (AFT) model under dependent censoring. One of the advantages in our approach is that these weights are optimal among all the linear combinations of these two estimators previously referenced. Moreover, to calculate these weights, a novel resampling-based scheme is employed. Attendant asymptotic statistical results for the estimator are established. In addition, simulation studies, as well as application to real data, show the gains in efficiency for our estimator.