Sequential Monitoring of Covariate Adaptive Designs

• Activity Number: 486
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Hongjian Zhu*+ and Feifang Hu
• Companies: The University of Texas Health Science Center at Houston and University of Virginia
• Keywords: Sequential monitoring ; covariate ; adaptive designs ; Type I error

Abstract

Covariate adaptive randomization (CAR) designs such as the Pocock-Simon design and the stratified permuted block design have been used to balance treatment assignments within covariate profiles and reduce bias. However, due to its mechanism of allocating subjects, few theoretical results have been proposed. More importantly, sequential monitoring becomes a standard in modern clinical trials. As a result, the protection of Type I error rate when sequentially monitoring a CAR design is of special importance for the development and application of CAR designs. In the talk, we propose to sequentially monitor a general family of CAR designs and offer some important asymptotic results.