Group-Sequential Randomized Trial Design Utilizing the MEM Framework for Incorporating Historical Data

• Activity Number: 73 - Alternative Designs and Related Topics
• Type: Contributed
• Date/Time: Sunday, July 28, 2019 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Ales Kotalik* and David Michael Vock and Joseph Koopmeiners
• Companies: University of Minnesota and University of Minnesota and University of Minnesota
• Keywords: adaptive designs; group-sequential design; historical data; supplementary data; data borrowing

Abstract

Definitive clinical trials are resource intensive, often requiring a large number of participants. One approach to improving the efficiency is to incorporate historical information into the primary trial analysis. This approach has potential in the areas of pediatric or rare disease trials, where achieving reasonable power is difficult.

We introduce a novel Bayesian group-sequential design based on the Multisource Exchangeability Models, which allows for dynamic borrowing of historical information at interim analyses. Our approach achieves synergy between group sequential and adaptive borrowing methods to attain improved power compared to using either method alone. We discuss frequentist operating characteristics of our design, such as power in the current trial, or a possible Type I error rate inflation. Our method achieves earlier stopping of the primary study while increasing power. We discuss the power and sample size calculations necessary for design of such a trial and also address the issue of potential non-monotone information accrual. We present our method for a continuous and binary outcome, as well as in a linear regression setting.