|
Activity Number:
|
446
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Statistics in Biopharmaceutical Research Journal
|
| Abstract - #302895 |
|
Title:
|
Variance-Based Sample Size Reassessment Increases Relative Efficiency of Studies
|
|
Author(s):
|
Deepak B. Khatry*+
|
|
Companies:
|
MedImmune, LLC
|
|
Address:
|
One Medimmune Way, Gaithersburg, MD, 20878,
|
|
Keywords:
|
Error rates ; Information-based design ; Study power ; Variable-sample design ; Variance misspecification ; Phase III clinical trial
|
|
Abstract:
|
Variance misspecification in sample size calculations can cause studies to be under- or over-powered. Fixed-sample study designs are not protected against variance misspecifications. Variable-sample designs with variance assessment and sample size re-estimation (SSR) at interim time can be effective in minimizing discrepancy between desired and end-of-study power levels. Combination of SSR with a group sequential trial (GST) can be implemented within an information-based study design framework. In this paper, simulation is used to compare consequences of variance misspecification in sample size calculations among different fixed- and variable-sample study designs. Results indicate that SSR designs with variance assessment and sample size re-estimation conducted at 50% of initially-calculated sample size are, on average, more robust than other fixed-sample study designs including GSTs.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |
