Estimating Treatment Effects with Treatment Switching via Accelerated Failure Time Model with Frailty Term

• Activity Number: 454
• Type: Contributed
• Date/Time: Wednesday, August 6, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Fang-I Chu*+ and Yuedong Wang
• Companies: and University of California, Santa Barbara
• Keywords: Treatment Switching ; Noncompliance ; Survival Analysis ; Clinical Trial ; Survival Analysis ; Accelerated Failure Time model

Abstract

Treatment switching occurs frequently in clinical trials, where, during the course of trial, patients fail to comply with the original assigned regime. Fail to account for treatment switching as we attempt to evaluate the treatment effect would lead to potentially biased conclusions. In this study, we propose the Accelerated Failure Time model (AFT) with frailty term to accommodate treatment switches based on Zeng's semi-parametric model(2012). The frailty term is introduced in order to model the relation between the event time of progression/switching and failure. Simulation studies are conducted to demonstrate the improved performance of proposed model compared with existing methods in literature.