|
Activity Number:
|
521
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Statisticians in the Pharmaceutical Industry
|
| Abstract - #307813 |
|
Title:
|
Performance of Multiple Testing Procedures To Compare Three Doses of a Test Drug and Placebo
|
|
Author(s):
|
Lan Kong*+ and Gary Koch and Thomas Liu and Huei Wang
|
|
Companies:
|
University of Pittsburgh and The University of North Carolina at Chapel Hill and Amgen Inc. and Amgen Inc.
|
|
Address:
|
313 Parran Hall, 130 Desoto Street, Pittsburgh, PA, 15261,
|
|
Keywords:
|
closed testing procedure ; familywise Type I error ; multiple comparison ; power
|
|
Abstract:
|
A study design with two or more doses of a test drug and placebo is frequently used in clinical drug development. Multiplicity issues arise when there are multiple comparisons between doses of test drug and placebo, and also when there are comparisons of doses with one another. An appropriate analysis strategy needs to be specified in advance to avoid spurious results through insufficient control of Type I error, as well as to avoid the loss of power due to excessively conservative adjustments for multiplicity. For evaluation of alternative strategies with possibly complex management of multiplicity, we compare the performance of several testing procedures through the simulated data that represent various patterns of treatment differences. The purpose is to identify which methods perform better or more robustly than the others and under what conditions.
|
