gms | German Medical Science

Choosing an appropriate sample size for a clinical trial is a crucial and challenging task, as it requires balancing the competing goals of minimizing patient risk and trial cost while ensuring sufficient statistical power. The required sample size depends on the effect size of the medical treatment and the variance of the endpoint. In practice, these values are unknown, leading to planning uncertainties and the potential problem of oversized or underpowered clinical trials. To address this issue, multi-stage trials with sample size recalculation have been proposed, where the sample size is adjusted based on interim analyses. For two-stage trials, there exists research about the benefits of sample size recalculation and the according approaches are applied in practice [1], [2]. For three-stage trials, previous literature merely examines designs with efficacy and futility stopping [3], [4]. The potential of sample size recalculation in three-stage trials has not yet been investigated. In this study, we analyse the use of sample size recalculation in three-stage trials with futility and efficacy stopping after the first two stages and sample size recalculation in the final stage. We investigate the extent to which sample size recalculation at the final stage can mitigate the risk of underpowered or oversized trials when the assumed effect size deviates from the true effect size. To see the practical implications, we perform simulations based on real-world clinical trials and examine, in addition to the power and expected sample size, other measures like the variance in sample size, which would be meaningful for the applicability.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.