454 – Innovative Designs in Medical Device Trials
An Application of Adaptive Enrichment Design to CRT Trial Data
Yonghong Gao
FDA
Rosenblum and Van der Laan (2011) proposed a new adaptive enrichment design for a two-stage randomized trial where the enrollment decision at the second stage and the hypothesis testing at the final are based on the first stage data. The proposed methodology provides a strong control of the family-wise type I error rate under a wide range of interim decision rules. In this paper we apply this new methodology to a CRT trial data under two interim decision rules. We also look at the effect of the decision rules to the overall power of this adaptive enrichment design through simulation studies. We conclude that decision rule at interim analysis is critical, and clinical input as well as statistical considerations should be taken into account when utilizing this new adaptive enrichment method for the study.
