Clinical trials of superiority, non-inferiority or equivalence?

Abstract

Efficacy and effectivity of new interventions are generally established through randomized clinical trials (RCTs). However, among many other methodological challenges, specifying the hypothesis of a RCT remains complex problem for clinical researchers.In this manuscript we discuss the characteristics of three variants of RCTs: superiority RCT (SRCT), non-inferiority RCT (NIRCT), and equivalence RCT (ERCT). These three types of RCT have different assumptions about the effects of an intervention, so setting hypotheses and defining objectives requires knowing some assumptions underlying these RCTs, including elements related to the estimation of the sample size for each one. The aim of this manuscript was to describe methodological differences between SRCT, NIRCT and ERCT.