Abstract:
|
We consider optimized, adaptive, confirmatory, enrichment trials to determine subgroups of patients where a new treatment is effective. Especially, we consider trials with an adaptive interim analysis where the size of subgroups can be adapted or a subgroup can be dropped. The adaptations are performed based on unblinded interim data. Applying a Bayesian decision-theoretic framework, we determine optimal adaptation rules. The optimal design maximizes a utility function that takes the statistical power as well as the population prevalence of the subgroups into account. To control the probability for false positive decisions, we apply the conditional error rate approach. Optimal adaptive enrichment designs are derived for trials controlling the familywise error rate (based on the closed testing procedure) as well as for umbrella trials in which only the per-comparison type 1 error rate is controlled. We illustrate the optimal designs for specific numerical examples and discuss their efficiency.
|