Issues and Some Analysis Methods on Composite Endpoints

• Activity Number: 565
• Type: Topic Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Nelson Lu*+ and Rajesh Nair and Chenguang Wang and Yunling Xu
• Companies: U.S. Food and Drug Administration/CDRH and U.S. Food and Drug Administration/CDRH and U.S. Food and Drug Administration/CDRH and U.S. Food and Drug Administration/CDRH
• Keywords: composite endpoint

Abstract

A composite endpoint consists of multiple single endpoints that are combined in order to obtain a larger number of events during a clinical study. By setting up a composite endpoint, less sample size is required and thus the study is less costly. However, issues arise and are well recognized for such an endpoint. Oftentimes, the importance to the patient of the individual components may not be similar. Results may be dominated or driven by less important endpoint(s). The situation becomes even more challenging when there are competing risks between endpoints.

Analyzing all individual components separately may partly address this issue. Yet it may not be able to provide convincing evidence due to insufficient power and maybe multiplicity issues. In this presentation, several methods to analyze a composite endpoint are proposed. Examples will be used to illustrate and assess these methods.