A Group Sequential Method Using Hochberg Procedure for Clinical Trials with Multiple Primary Endpoints

• Activity Number: 599
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Kentaro Sakamaki*+
• Companies: Yokohama City University
• Keywords: multiple endpoints ; group sequential method ; Hochberg procedure

Abstract

In some clinical trials with multiple primary endpoints, inferences are made by group sequential methods, such as a group sequential Holm procedure (Ye, Li, Liu and Yao published online). Since existing methods employ Bonferroni inequality, they are underpowered when endpoints are correlated. In this talk, I propose a group sequential method with Hochberg procedure that is based on Simes inequality. An advantage of this method is the potential for a gain in power without specifying a correlation structure between endpoints. It can be been shown that the proposed method preserves the familywise error rate (FWER) in the strong sense using the closed testing procedure. It was confirmed that the proposed method controls the FWER and increases power compared with other group sequential procedures through simulation studies.