A0896
Title: Platform designs with added new arms
Authors: Haitao Pan - St. Jude Children's Research Hospital (United States) [presenting]
Abstract: Platform trials defined as multi-arm multi-stage trials with adding and removing experimental arms have gained attention recently. Statistical issues with respect to ``adding new arms'', however, have not been well discussed. In the setting of pre-planned deferred arms, with goals of controlling the familywise error rate (FWER) and achieving a targeted experimental-wise power, we provide a principled approach, including, how to modify the critical boundaries to control the FWER when new arms are added, how to re-estimate the sample size and provide the optimal allocation ratio for the control to experimental arm to maintain the experimental-wise power. The influence of timing of adding new arms to the design's operating characteristics has also been examined. We developed an R package for practitioners to implement this method.