Term
Non-inferiority trial
A trial designed to show a drug is not meaningfully worse than an existing treatment, within a pre-specified margin.
When an effective treatment already exists, beating placebo is not the relevant question. A non-inferiority trial pre-specifies a margin, the largest disadvantage that would still be clinically acceptable, and succeeds if the data rule out being worse than that margin.
A successful pre-specified non-inferiority or equivalence test counts as a met primary endpoint under this site's rubric. The design's weak point is the margin itself: a generous margin can make 'not much worse' easy to claim, which is why the margin belongs in the protocol before the data exist.
Related terms