Bioequivalence Workflow Execution Error 11016

Phoenix Certara Forums Phoenix WNL basics
1

Hi all,

I’m working on a 3 period, 3 treatment, crossover PK study; using the bioequivalence module to determine BE (Test/Reference) (Test/Comparator) I’m able to execute the run where I have an equal number of observations per treatment. Where I have an uneven number, LSMs are not estimable and I receive the error message: “ERROR 11016: Numerical problem when computing satterthwaite df.”

I suspect the issue/answer is within the information above but if anyone could shed any light on this, I’d be very grateful.

Thanks in advance,

Danielle

2

Is it possible you can share your project ?

I’m not sure if Phoenix is defaulted to the replicated model (probably is), or if you made other changes. It could be the expected instability with a partial replicate model.

3

Hi Danielle,

from your description

… 3 period, 3 treatment, crossover PK study […] to determine BE (Test/Reference) (Test/Comparator)

your design is not clear to me. Do you have a higher-order crossover (e.g., 3×3 Latin Squares like TCR|CRT|RTC or a 6-sequence Williams’ design)?

If yes, you should set up data wizards to obtain two data sets in Incomplete Block Designs (i.e., in the first you exclude R and in the second C). Then perform two separate analyses, one T vs. C and one T vs. R. See also the Vignette of the R-package PowerTOST.