Bioequivalence and Bioavailability Forum

Hi John,

❝ This really bugs me (from EMEA guidance)...

❝

❝ "In studies with more than two treatment arms (e.g. a three period study including two references, one from EU and another from USA, or a four period study including test and reference in fed and fasted states), the analysis for each comparison should be conducted excluding the data from the treatments that are not relevant for the comparison in question."

Me too

❝ I have a dataset containing 4 treatments (from a 4-way 4-treatmens 4-sequence crossover study) and each treatment has different # of subjects (ie: A(T-Fast)=24, B(Ref Fast)=19, C(T-Fed)=22, D(Ref Fed)=24). I tried doing the stats as per the above statment (A vs D, C vs D, C vs A, D vs B) by removing the treatments that are not of interest but SAS gave me an error message for my SAS code

❝ estimat1 = Estimate "A - D" Treatment 1 0 0 -1

❝ "WARNING: More coefficients than levels specified for effect TREATMENT. Some coefficients will be ignored. NOTE: A vs D is not estimable." I got the same error message for each comparison.

❝

❝ Any reason why?

Have a look at the "Class levels" output of your Proc GLM evaluation.
Especially the levels of treatment .

You will notice that you have only 2 treatments in your EMA crippled evaluation (A and D in the data if the comparison A vs. D is concerned).
Thus you have to formulate:
Estimate "A - D" Treatment 1 -1
to get the 90% CIs of of A-D.