Bioequivalence and Bioavailability Forum

Dear Dan!

❝ Is it possible to conduct in case of a highly variable long half life drug a parallel group study with an add on design?

Theoretically yes. Actually group sequential methods were developed for parallel designs (superiority case). It should be possible to adapt these methods even more easily to the equivalence case than in the recent work for cross-overs (Potvin et al. 2008, Montague et al. 2011). But since the Null-hypothesis is formulated differently we would need simulations to demonstrate no α-inflation. Maybe one of the forum’s members is already working on it?

❝ Rationale: widening of acceptance range not possible due to the fact that the replicate design is out of question but intention to be more on the safe side with the possibility to adapt sample size.

Right.