Bayesian Response Adaptive Randomization for Delayed Response with a Short-Term Outcome

• Activity Number: 350
• Type: Contributed
• Date/Time: Tuesday, August 2, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Mi-Ok Kim*+ and Chunyan Liu and J. Jack Lee
• Companies: Cincinnati Children's Hospital Medical Center and Cincinnati Children's Hospital Medical Center and The University of Texas MD Anderson Cancer Center
• Keywords: Bayesian ; Response Adaptive Design ; clinical trial

Abstract

We consider an application of response-adaptive randomization (RAR) design in a clinical trial where the primary endpoint takes a long time to observe but a short-term "surrogate" outcome is available. The asymptotic properties of the design have been shown to be little affected when the delay is not very long relative to the accrual process (e.g. Bai et al, 2003; Hu et al, 2008). These theoretical results, however, are not useful when the delay is long, as in many survival outcome trials, or the sample size is small. Huang et al (2009) proposed a Bayesian approach of utilizing the short-term outcome for a RAR for the long-term response. We use Huang et al's approach to study the effect of the delay on a RAR design with a small n and to investigate when Huang et al (2009)'s approach is beneficial compared to an approach that does not utilize the short-term outcome. Simulation results show that the more complex Huang et al's approach performs better as the priors becomes more informative and/or if the treat effect differs by the short-term outcome.