Bioequivalence and Bioavailability Forum

Hi Lucas,

confusing that they ask for means and not medians. I am nevertheless inclined to say you should use non-parametrics (Wilcoxon signed ranks / Hodges-Lehman) for requirement #1.

A potential way forward to requirement #2 is the construction of confidence intervals based on proportions of score 1 and above from binomial testing.
It worries me a little with several time points. As I see it this means several tests must be done which can be rather nasty. It makes power calculations quite difficult prior to planning the pivotal trial.

Those are guesses from my side. I have not been involved in Rivastigmine patch developments.
Best of luck.