Keywords: real-world data, propensity score, Bayesian power prior
The amount of real-world data (RWD) collected from sources in real-world settings are increasing ultrarapidly. With the wide availability of various RWD, there has been a great level of interest in using real-world data in regulatory decision makings. Based on propensity score adjustment, we propose a modified Bayesian power prior approach for incorporating RWD in the design and analysis of single group clinical trials. Numerical studies are conducted to evaluate the performance of the proposed method in different settings. In addition, a hypothetical single group medical device study is provided to illustrate the proposed method in the context of regulatory review.