Bioequivalence and Bioavailability Forum

Dear Helmut, dear Priyanka,

❝ ❝ We are planning to go replicated design with two stage sequential approach.

❝

❝ Hhm - to my knowledge there's no such a method published right now. Agree that it would be desirable to have such a method.

Have a look into
Patterson et. al.
"Non-traditional study designs to demonstrate average bioequivalence for highly variable drug products"
European Journal of Clinical Pharmacology
Volume 57, Number 9 / November 2001, 663-670

There is on page 666 a study mentioned which was done as two-stage sequential with a replicate cross-over design. As far as I can see from the very short description the 95% CIs were used as decision criterion at the two stages (classical Bonferroni alpha spending). No model mentioned which incorporates the stages, mainly because the study didn't need a second stage. No hint if overall alpha=0.05 is preserved by this method.

Unfortunately I was not able to locate the mentioned original paper cited in that paper as
Patterson (2000),
"Approaches to meeting a nearly impossible regulatory hurdle:
demonstrating average bioequivalence for a highly variable drug product"
Clin. Pharmacol. Ther. 67: 115
to get more insight.
---------------- [edit] -----------------------;
Solved: Its an abstract of a speech or poster.
---------------- [/edit] ----------------------;

BTW: I would never consider or plan a study with a design I'm not able to evaluate, whatever advantages this design else had! But may be others have more foolhardy guts.