Bioequivalence and Bioavailability Forum

Dear Dr Gunasakar,

I disagree. Add-on studies (with α_adj 0.025) were mentioned in the Jan 2010 draft, but dropped in the final (May 2012) guidance. Only group sequential designs (Pocock 1977; α_adj 0.0294) and adaptive designs (Potvin et al. 2008; Method C: α_adj 0.05 or 0.0294) are acceptable. Since Pocock’s method was never validated for BE studies (only: superiority testing, parallel design, n₁=n₂, known variance, Z instead of t) I would only go with Method C (or alternatively with Method D by Montague et al. 2011 if T/R is expected to be 0.9; α_adj 0.028).