|
Activity Number:
|
119
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 9, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #301205 |
|
Title:
|
Testing Superiority and Noninferiority Hypotheses in Active Controlled Clinical Trails E
|
|
Author(s):
|
Yi Tsong*+ and Juan (Joanne) Zhang
|
|
Companies:
|
U.S. Food and Drug Administration and U.S. Food and Drug Administration
|
|
Address:
|
HFD-705, 5600 Fishers Lane, Rockville, MD, 20857,
|
|
Keywords:
|
active control ; cross-trial comparison ; noninferiority test ; superiority test ; simultaneous test
|
|
Abstract:
|
Switching between testing for the hypothesis of superiority and the hypothesis of noninferiority has been an important statistical issue in the design and analysis of active controlled clinical trials. In practice, it is often conducted with a two-stage testing procedure. It has been shown that there is no Type I error rate adjustment required to switch to testing for the hypothesis of noninferiority once the data fail to support the superiority claim. Neither is the adjustment required to switch to testing the hypothesis of superiority once the null hypothesis of noninferiority is rejected. It has been shown to be true in trials with properly pre-specified noninferiority margin in a generalized historical control approach. However, when using a cross-trial comparison approach for noninferiority testing, controlling the Type I error rate may become an issue with this conventional two-stage procedure. We proposed to adopt a single-stage simultaneous testing concept as proposed by Ng (2003) for testing both hypotheses of noninferiority and superiority simultaneously. The Fieller's confidence interval procedure as proposed by Hauschke et al. (1999) is used in the proposed approach.
|
