Innovations in the Use of External Data to Augment Controls in Clinical Trials

• Activity Number: 166 - Non-Clinical Statistics, Personalized Medicine, and Other Topics
• Type: Contributed
• Date/Time: Tuesday, August 10, 2021 : 10:00 AM to 11:50 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Frank Shen* and Rong Liu
• Companies: Bristol Myers Squib and Bristol Myers Squibb
• Keywords: External Data; Historical Data; Bayesian; Power Prior; Hierarchical Model; Sensitivity Analysis

Abstract

There is a wealth of data available for commonly used control treatments and, if properly integrated, the external control data could provide increased power, lowered costs, and other benefits to a study. We discuss a variety of methods to integrate historical control data to augment the current control arm of a study in a Bayesian and frequentist frameworks. We also discuss the potential of combining frequentist "all or nothing" methods like test-then-pool with Bayesian methods, and provide guidelines on how to compare these methods using simulations-based sensitivity analyses.