Online Program Home
My Program

Abstract Details

Activity Number: 343 - Innovative Trial Designs and Analytics
Type: Contributed
Date/Time: Tuesday, July 30, 2019 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #304971 Presentation
Title: A Bayesian Design with Conditional Borrowing of Historical Data in a Rare Disease Setting
Author(s): Peng Sun* and Ming-Hui Chen and Yiwei Zhang and John Zhong and Charlie Cao and Guochen Song and Zhenxun Wang
Companies: and University of Connecticut and Biogen and Biogen and Biogen and Biogen and University of Minnesota,
Keywords: Bayesian design ; power prior; conditional borrowing; randomized controlled trial; hierachical modeling; Type I error
Abstract:

In this presentation, a Bayesian design that enables conditional borrowing of historical data will be proposed in a randomized controlled trial (RCT) setting for the treatment of a rare disease. Conditional borrowing means that borrowing of historical data using power prior occurs only if the difference in sample means between the concurrent control and the historical control falls within a pre-specified range. The key features of the proposed Bayesian design are that, as the distribution of the concurrent control shifts away from that of the historical control, the resulting inflation of Type I error is bounded and the Bayesian estimator of treatment difference remains unbiased. Details of the operational characteristics of the proposed design will be presented and advantage of the proposed design over the Bayesian hierarchical modeling approach will be discussed.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program