|
Activity Number:
|
366
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #300865 |
|
Title:
|
Prior Estimation for Empirical Bayes Binomial Models
|
|
Author(s):
|
Boris G. Zaslavsky*+
|
|
Companies:
|
U.S. Food and Drug Administration
|
|
Address:
|
1401 Rockville Pike, Rockville, MD, 20879,
|
|
Keywords:
|
beta distribution ; beta - binomial distribution ; maximum likelihood ; sample size
|
|
Abstract:
|
With the high cost of clinical trials, traditional methods of sample size determination may be too conservative. Bayesian methods are often used in the clinical trial environment to reduce required sample sizes and/or increase power. Bayesian methods are vital in trials where the observed results have to be extended on more extreme values which are unavailable for ethical reasons. The Beta distribution is a natural prior for binomial models. Under the empirical Bayes approach, the parameters of this distribution are the maximum likelihood estimator of the marginal beta-binomial distribution. For the sample size calculations, the maximum likelihood solution should be adjusted by a discount factor to reflect a partial exchangeability of historical trials as opposed to current studies. We suggest to measure the exchangeability.
|
