|
Activity Number:
|
408
|
|
Type:
|
Roundtables
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 12:30 PM to 1:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #301768 |
|
Title:
|
CANCELLED - Prespecifying Analyses: In the Lifecycle of a Clinical Trial, Does It Matter When, Where, and How You Prespecify?
|
|
Author(s):
|
Bruce Binkowitz
|
|
Companies:
|
Merck Research Laboratories
|
|
Address:
|
|
|
Keywords:
|
pre-specification ; post-hoc ; analysis plan
|
|
Abstract:
|
A critical criterion to the integrity of any statistical analysis from the regulatory and scientific point of view is whether it was "pre-specified," as opposed to "post-hoc." The distinction often lies in when the analysis was performed versus the analyst's state of knowledge of the data. This roundtable will discuss whether pre-specification is a simple yes/no issue, or more a continuum. Participants should be prepared to contribute with opinions and experiences on such related topics as the importance of how and where such pre-specification is done and whether the audience matters.
|
