Bioequivalence and Bioavailability Forum

Dear beman!

❝ In Literature and FDA-Guideline (1999) only following sequences are recommended:

❝ RTR (2 Times R)

❝ TRT (2 Times T)

… and some of others: TRR|RTT, TRR|RTR|RRT, TRT|RTR|TRR|RTT

❝ This design was needed for individual bioequivalence (because you need the CVintra of the Reference and the Test Formulation.

Right.

❝ Proposal 1:

❝ TRR

❝ RRT

❝

❝ or

❝

❝ Proposal 2:

❝ TRT

❝ RTR

❝ TRR

Don’t forget the partially replicated design TRR|RTR|RRT quoted in FDA's progesterone guidance.

❝ Proposal 2 is the better one. But it is possible to use proposal 1, or are the disadvantages to big ?

I would also say that #2 is better, because in #1 you never dose T in period 2. Imagine a (rather hypothetical) situation where you have a large period effect in period 2. Both designs are not balanced, but the bias on the treatment effect would be larger in #1 as compared to #2.