Dynamic Borrowing from a Single Prior Data Source Using the Conditional Power Prior

• Activity Number: 203 - Recent Advances in Bayesian Adaptive Designs, from Early Phase to Confirmatory Trials
• Type: Topic-Contributed
• Date/Time: Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Caucus for Women in Statistics
• Author(s): Jianjin Xu* and Laura Thompson
• Companies: Food and Drug Administration and Food and Drug Administration
• Keywords: conditional power prior; Bayesian ; adaptive design; similarity margin

Abstract

The conditional power prior (CPP) is a popular method to borrow information from a single prior data source. The amount of borrowing is controlled by the power parameter which is fixed before running the new study. However, fixing this parameter before running a new study is often difficult and may be unwise because if the outcomes in the current study are much different from the prior data outcomes, the power parameter cannot be changed to reflect a more appropriate degree of borrowing. In this talk, we introduce a statistical method with power parameter dynamically determined by the similarity between the interim data of the current study and the prior data, with the similarity restricted by a pre-specified clinically justified margin. The proposed approach allows for adaptive borrowing of the prior data with the degree of borrowing readily understood by the clinicians. Through simulations, we show that our approach has good operating characteristics while reducing type I error rate in areas outside of the clinical “similarity region”.