|
Activity Number:
|
258
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305094 |
|
Title:
|
Conditional Estimation of Treatment Effects for Clinical Trials with Group Sequential Designs
|
|
Author(s):
|
Michael W. Lee*+ and Hui Quan
|
|
Companies:
|
Johnson & Johnson Pharmaceutical R&D and sanofi-aventis
|
|
Address:
|
920 Route 202, S., Raritan, NJ, 08869,
|
|
Keywords:
|
interim analysis ; early stopping ; multiple endpoints ; estimate bias
|
|
Abstract:
|
While preserving type I error rate, a group sequential design allows a trial to be stopped earlier if the treatment effect can be demonstrated through interim analyses based on the prespecified stopping rule. Thus, group sequential designs have been used widely, especially for long-term trials. However, researchers have demonstrated the biasness of the unadjusted estimate of the treatment effect. They proposed adjusted estimate to address the issue. In this presentation, we will propose a conditional approach for obtaining the adjusted estimate of treatment effect when the stopping rule relies on the outcomes of multiple primary endpoints. The proposed approach provides estimate with reduced bias and confidence interval with approximately the nominal coverage rate across stopping time. An example will be used to illustrate the application of our approach.
|
