During the course of a clinical trial, the procedure may need to be modified based on medical or practical results. For example, in trials with low recruitment rates, the data monitoring committee often changes the inclusion/exclusion criteria, the number of sites, the endpoint, or the protocol itself. When such changes occur, protocol amendments are needed. However, the actual patient population after protocol amendments may be slightly different from the target population. In this study, we focus on the scenario in which the populations have heterogeneous variances across the modifications of trial procedures due to protocol amendments. A statistical method based on the Satterthwaite approximation with corresponding sample size adjustments is developed to evaluate the treatment effect. Our simulation demonstrates that the proposed method controls the Type I error rate and provides sufficient power.
Date:
2024-08-28
Relation:
Statistics in Biopharmaceutical Research. 2024 Aug 28;Article in Press.
