When investigators measure the effect of treatment on an aggregate of endpoints of various importance. Inferences from composite endpoints are...
Continue reading

When investigators measure the effect of treatment on an aggregate of endpoints of various importance. Inferences from composite endpoints are strongest in the rare situations in which (1) the component endpoints are of similar patient-importance, (2) the endpoints that are more important occur with at least similar frequency to those that are less important, and (3) strong biologic rationale supports results that, across component endpoints, show similar relative risks with sufficiently narrow confidence intervals.