To the Editor: In their Commentary on the use of composite end points in clinical trials, Drs Tomlinson and Detsky1 asked why the clinical trial of rosiglitazone2 used the primary outcome of diabetes or death. Although the authors discussed various (often problematic) reasons for using composite end points, they did not consider an important legitimate reason for combining death with another end event: to focus on the long-term risk of the end point.
Sheehe PR. Composite End Points in Clinical Trials. JAMA. 2010;303(17):1698. doi:10.1001/jama.2010.522