From non-inferiority to superiority: the shift towards patient-centric outcomes

Abstract

The recently published REC-CAGEFREEI trial 1 provides an interesting example of non-inferiority vs. superiority design of clinical trials. The trial failed to show non-inferiority (P = 0.65) of drug-coated balloon angioplasty (with the option of rescue stenting) to the intended deployment of drug-eluting stents on the primary composite endpoint of car-diovascular death, target vessel myocardial infarction, and target lesion revascularization. However, not the result, but the motivation and reporting on the trial draw our attention (Figure 1). Non-inferiority designs aim to show that a novel therapy is not worse than a standard of care by more than a pre-specified non-inferiority margin on an efficacy outcome. This margin represents an acceptable loss on efficacy, which is justified by a putative advantage of the novel therapy on patient outcomes other than efficacy, such as improved safety, better quality of life, more convenient administration, or lower cost. A non-inferiority design should thus be motivated by a clear advantage. In stent trials, for example, a short-term reduction in