Using the geometric average hazard ratio in sample size calculation for time-to-event data with composite endpoints

摘要

Composite endpoints (CEs), defined as the union of several outcomes, are extensively used as a primary endpoint when designing a clinical trial. In time-to-event studies, CE refers to the elapsed time from randomization to the earliest observation among its components. It is common in oncology trials to use progression-free survival (PFS) as a primary endpoint: this outcome is defined as the time elapsed from randomization to tumor progression or death from any cause, whichever occurs first [1]. In cardiovascular trials, major adverse cardiac event (MACE) is generally defined as a composite endpoint of time to cardiovascular death, myocardial infarction, stroke and target vessel revascularization [2]. Composite endpoints are as well often used for infectious diseases. In the ARREST trial [3], the primary endpoint was the time to bacteriologically confirmed treatment failure, disease recurrence or death.