Abstract
In this article, we review the use of all-cause mortality and the composite endpoint as the primary endpoint and propose a recurrent composite endpoint as an alternative primary endpoint. Some general suggestions to trialists and practitioners are provided. In summary, the use of the composite endpoint is an issue that needs clarification.
Notes
a Mortality, nonfatal myocardial infarction, and hospital admission for acute coronary syndrome. The authors report number of events rather than subjects.
b The authors report number of events rather than subjects.
1Because we do not have access to the raw data, we estimate the hazard by −log(1 − ) where is the observed proportion of subjects in group τ experiencing the event within 1 year.