Bioequivalence and Bioavailability Forum

Hi John,

❝ I came across this kel determination by Phoenix Winnonlin. Do you think Kel should be labelled as "Cannot be determined"? I think it should be since the last 3 points of the elmination phase show a +ve trend.

Difficult remote diagnosis. It seems that you already tried to exclude the data point (“red cross”). Would you mind exporting the plot from PHX as a PNG; width 640px and upload it again?
It’s the responsibility of the user to check all estimations (even mentioned somewhere in the User’s Guide). Here we have again an example of the poor performance of the max. R²_adj algorithm. I bet In ‘classical’ WinNonlin even C_max would have been included.
Personally I would try starting at two time points after the inflection of the profile – but this would give you an even longer half-life. Without knowing the drug, the analytical method, etc. I don’t know whether this makes sense. Do you see similar long half-lives in other subjects as well?