ABSTRACT

Randomized clinical trials help us in evaluating the treatment effect of a therapy compared to a control treatment. In these trials, efficacy can be established by testing for superiority, non-inferiority, or equivalence. In a superiority trial, the aim is to demonstrate that the treatment effect of an experimental product (T) is superior to the treatment effect of a control where the control may be a placebo (P), best supportive care (BSC), or an active control (C). In a non-inferiority trial, the aim is to demonstrate that the treatment effect of an experimental product T is not inferior to an active control C by more than a prespecified quantity (δ) where δ is small enough so that it does not matter clinically. The aim in an equivalence trial is to demonstrate the treatment effect of T compared to an active control C is neither inferior nor superior by a prespecified margin of error. The null and alternative hypotheses in each of these scenarios are different and captured in the following formulations and Figure 9.1. Let Δ denote the difference in treatment effect between T and C.