|
Activity Number:
|
418
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #308586 |
|
Title:
|
Sample Size Re-estimation for Adaptive Sequential Design in Clinical Trials
|
|
Author(s):
|
Ping Gao*+ and James H. Ware
|
|
Companies:
|
The Medicines Company and Harvard University
|
|
Address:
|
1019 Chambers Ct, Bridgewater, NJ, 08807,
|
|
Keywords:
|
sequential design ; sample size re-estimation ; conditional power ; conditional type I error ; Brownian motion
|
|
Abstract:
|
In designing clinical trials, sample size is often estimated based on prior information of the effect of the test drug or test treatment regimen, in the form of expected size of the treatment effect and its variance. It is therefore often desirable to perform an interim analysis or interim analyses to reassess the treatment effect to either stop the trial early due to overwhelming evidence of efficacy or increase sample size to assure adequate power. We propose a conditional power based sample size calculation for sequential designs. Our method extends Fisher's (1998) result, which does not allow for early termination.
|
