An Improved Method for Testing Two Families of Endpoints

• Activity Number: 409
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Haihong Li*+ and Abdul J. Sankoh
• Companies: Vertex Pharmaceuticals and Vertex Pharmaceuticals
• Keywords: Clinical trials ; Multiple comparisons ; Multiple primary endpoints ; Multiple secondary endpoints

Abstract

One of the recent advancements in multiple testing techniques in clinical trials is the adaptive alpha allocation approach proposed by Li & Mehrotra (SiM 2008), which groups the hypotheses into 2 families based on perceived powers and allows the significance level for the 2nd family to be set adaptively based on the largest observed p-value in the 1st family. While this is a major improvement over the fixed allocation scheme, we show that this method maintains nominal size and adequate power only when there is a single endpoint in the 1st family. We introduce improvement that is substantially more powerful in the general case of multiple hypotheses in the 1st family. In the case when the test statistics are correlated, we introduce well defined functions to calculate the significance level for the 2nd family. The numerical computation for our method is straightforward.