Multiplicity Adjustment for Clinical Trials With Two Doses of an Active Treatment and Multiple Primary and Secondary Endpoints

作     者：Quan, Hui Capizzi, Thomas Zhang, Ji 

作者机构：Sanofi Aventis Biostat & Programming Bridgewater NJ 08807 USA Schering Plough Corp Biostat & Data Management Kenilworth NJ 07033 USA Sanofi Aventis Int Clin Operat Bridgewater NJ 08807 USA 

出 版 物：《STATISTICS IN BIOPHARMACEUTICAL RESEARCH》 (生物制药研究统计)

年 卷 期：2009年第1卷第3期

页      面：258-267页

核心收录：

学科分类：0710[理学-生物学] 1007[医学-药学(可授医学、理学学位)] 0714[理学-统计学（可授理学、经济学学位）] 10[医学] 

主　　题：Closed procedure Family-wise error rate Gatekeeping procedure Strong control 

摘      要：Frequently, two doses of an active treatment, multiple primary and secondary endpoints are simultaneously considered in Phase III confirmatory clinical trials. For these trials, many traditional multiplicity adjustment procedures do not take into account the possible dose effect on each endpoint and the priority order of the primary and secondary endpoints, and therefore may have lower power. To gain power, we consider the problem as a three-dimensional multiplicity problem: one dimension concerns the multiple doses, one dimension concerns the priority order of the primary and secondary endpoints, and another dimension concerns the multiple endpoints in each priority category. We propose procedures that consider the dose order and the priorities of the endpoints to form closed procedures and therefore control the family-wise error rate. Furthermore, we consider procedures that provide control of type I error rates in clinically relevant subfamilies of comparisons, providing a pragmatic approach to maintain study power as compared to the traditional approaches for the ultimate overall strong control. Simulation results and a real data example show that the procedures proposed in this article in general are easy to use and have improved power.