Bioequivalence and Bioavailability Forum

Dear Sinhmar,

❝ [...] So in SAS and Winnonlin outputs we are getting variance for both test and reference while we are repeating only reference as per our design.

❝ So my question is how are we getting the variance for Test [...]

At my present knowledge it is not totally clear for me, how SAS (with the SAS code given in the mentioned thread) and WinNonlin are calculating the variance components of the intra-individual variabilities for the test and reference formulations.
My feeling is that there is a problem with the model formulation within both software. An indication for that is that both came out with different values for the model parameters (especially intra- and inter-subject variabilities).

Thus I would go for the moment with the so called "Method of moments" in evaluating scaled ABE. The necessary formulas for estimating the intra-individual variability of the reference and the estimator of the difference between test and reference for the extra-reference design can be found in
R.J. McNally
Tests for Individual and Population Bioequivalence Using 3-Period Crossover Designs
which can be found online here.

This paper deals indeed with individual BE but can puzzled together with the paper of Haidar et.al. to evaluate scaled ABE.

Haidar et.al.
Bioequivalence Approaches for Highly Variable Drugs and Drug Products
Pharmaceutical Research 25/1, 237-241 (2008)
free online resource here

Hope this helps.
See also this thread.