Chapter 9 210 proposed method compares favorably with traditional and novel methods in simulation in terms of RMSE and type I error, and may be a useful addition to the growing field of propensity score integrated dynamic borrowing approaches. The potential decrease in trial size and associated patient burden, the high unmet medical need in expanded access programs, together with the precautionary statistical set-up may favor the inclusion of expanded access with current trial data. Nonetheless, the inclusion of evidence sources remains a trade-off between bias due to including non-trial data and increased precision due to increased sample size.
RkJQdWJsaXNoZXIy MTk4NDMw