Optimizing Utilization of Composite and Repeat Endpoints in RCT

It could enable trials to be powered with smaller sample sizes. It can also be used to combine diverse information in a way to better guide clinical decisions.

The methodology exists and the need is growing as we try to sort out smaller treatment effects and our focus shifts from MI (which is fairly easy to define) to heart failure (which is more nebulous).
Traditional time-to-first endpoint analyses of clinical trials fail to capture the full impact of treatment in diseases with recurring endpoints (like heart failure hospitalization) or to capture the net benefit/risk of treatments with significant opposing effects (like the anti-thrombotic pro-bleeding effects of drugs like warfarin and dabigatran in atrial fibrillation or dual anti-platelet therapy after cardiovascular stenting), New statistical methods like the (win-ratio) when used together with patient- and clinician-based rankings of the importance of possible outcomes might provide a quantitative way to optimize the relevance of trials with multiple diverse endpoints to practical clinical decision making. I propose to encourage the use of these and other similar methodologies to analyze endpoints in new and existing trials. It would require a paradigm shift in how trialists look at endpoint data and in how regulators interpret trials.