A1164
Title: External control designs to incorporate real-world evidence with adaptive information borrowing
Authors: Jun Yin - Moffitt Cancer Center (United States) [presenting]
Peter Noseworthy - Mayo Clinic (United States)
Xiaoxi Yao - Mayo Clinic (United States)
Abstract: Randomized clinical trials (RCTs) are considered the standard approach for assessing drug efficacy; however, patient recruitment in RCTs can be challenging for rare diseases. Innovative approaches can enhance clinical trials' efficiency, such as information borrowing, which involves leveraging information from external data (i.e., previously completed trials or contemporaneous practice data). A novel Bayesian method of information borrowing is proposed from external data in clinical trials, motivated by a Mayo Clinic pragmatic trial with real-world controls. The proposed method allows borrowing when appropriate but also accommodates heterogeneous scenarios. Performance is evaluated using simulation studies.
