|
Activity Number:
|
94
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #309735 |
|
Title:
|
Modeling Response, Interaction, and Subgroup Analyses: Some Practical Issues
|
|
Author(s):
|
Mohammad Huque*+ and Mohamed Alosh
|
|
Companies:
|
Food and Drug Administration and Food and Drug Administration
|
|
Address:
|
Div of Biometrics IV Room 6232 Bldg 22, Silver Spring, MD, 20993,
|
|
Keywords:
|
multiplicity adujstment ; repeated measurements ; subgroup analyses ; testing for interaction
|
|
Abstract:
|
This presentation addresses some of the practical issues that arise in the assessment of treatment effect in clinical trials. In particular we consider: (i) repeated measurements analysis vs. final endpoint analysis, baseline adjusted vs. unadjusted analyses. (ii) lack of power and consequently under use of testing for interaction of treatment effect and (iii) susceptibility of subgroup efficacy findings to misinterpretation due to lack of specification of such analyses in the protocol and/or lack of adjustment for the multiple subgroup analyses which are frequently made. In addition, we present a flexible hierarchical approach for addressing multiplicity issue in protocol pre-specified subgroup analyses. Such an approach represents an improvement over the well-known traditional approaches which are usually so conservatives that significant results are rarely observed after adjustment.
|
