Hypothesis Testing of Covariate-Adaptive Randomized Clinical Trials with Survival Outcomes

• Activity Number: 372
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Lu Wang* and Hongjian Zhu and Jing Ning
• Companies: The University of Texas Health Science Center and The University of Texas School of Public Health and MD Anderson Cancer Center
• Keywords: Adaptive Design ; Survival outcomes ; Pocock and Simon's marginal procedures ; stratified permuted block design

Abstract

Covariate-adaptive designs are often implemented to balance important covariates in clinical trials. For clinical trials with survival outcomes, the theoretical properties of conventional estimates and testing hypotheses are usually unknown under covariate adaptive randomization. We provide theoretical foundation of hypothesis testing under covariate-adaptive designs based on the semiparametric AFT model with an unspecified error distribution. Weighted least square estimator with right censoring (stute, 1994) is used in our discussion. A large class of covariate-adaptive designs including Pocock and Simon's marginal procedure (Pocock and Simon, 1975) stratified permuted block design, etc. meet the conditions we put in the paper.