In clinical trials, why do we not always use Equivalence trials instead of superiority and non-inferiorty? So I understand what these three types are aimed for, I know they answer different questions. However, why wouldn't we always use equivalence trials? I mean if for example were are comparing two drugs and our null hypothesis is that the two drugs are different, if we found out that the null hypothesis is true (we found their efficacy to be different), then we would also know which one is superior and which is inferior (after all how else did we know there is a difference at all?), so why would we bother with designing studies specifically to address these questions?
I assume that one reason for that is that equivalence design would need a larger sample size than the other two designs?
 A: *

*In a group sequential design, non-inferiority is an interim stopping rule to declare early that we fail to reject the grand null hypothesis of superiority. We don't power for this kind of interim stopping rule. We pay the penalty of having a higher type II error rate and stop the trial early when the interim results are underwhelming. It is not a formal hypothesis test.


*Powering a study to demonstrate equivalence usually takes a stupendous sample size, so much so that it's not possibly cost effective as a research hypothesis, especially since the rationale for testing equivalence in most settings is to gain support for less harmful or less expensive alternatives to accepted therapies.


*Non-inferiority is an often used hypothesis in the medical device world for the above mentioned reasons (better safety, better cost), but it's not a guarantee for approval, and leaves many more areas to be addressed than a simple superiority study.
By and large, the reason why we don't see these hypotheses used more is more for political reasons, for instance, an insurance will only reimburse for the therapy that is proven to be better, or the shareholders or board of a company will only greenlight trials that are designed for superiority, since one can charge lucrative prices for new products under patent.
