Bioequivalence and Bioavailability Forum

Hei Pasi,

please see the Forum’s Policy.
Since you are affiliated with STAFINN and an adjunct professor of biostatistics are you up to answer some questions?

❝ […] consistency etc checking is performed …

What’s this exactly? Are you essentially following these routines? Are you aware that for EMA Proc MIXED is of historical interest only (GL p15)? BTW, CPMP/EWP/QWP/1401/98 ≠ ICH.

❝ … data is derived ready for PK calculation which is performed automatically.

Automatically…
According to the website you are using ‘classical’ WinNonlin (BTW, which version?). Are you aware that in WinNonlin ≤v5.3 in the automatic method for estimation of λ_z C_max/t_max may be included (based on max. R²_adj)? This flaw was corrected in Phoenix/WinNonlin (all versions). Unbiased esti­mates of the elimination in a one-compartment model are obtained for timepoints ≥2× t_max. Which trapezoidal rule are you using? Hopefully lin-up/log-down – otherwise AUC will be positively biased if data points are missing in the profile. And so on and so forth…

❝ When the PK parameters are calculated (non compartmental calculation),…

Only IR formulations (you refer to CPMP/EWP/QWP/1401/98 on your site)? What about MR pro­ducts (CPMP/EWP/280/96) and the recent Q&A (see here)?

❝ the statistical analysis is performed

Which code are you using for replicate designs? Which method are you using in assessing outliers?

❝ One day, not much other manpower than the expert review of the pk calculations. Follows the EMA BE-study guidelines.

Don’t get me wrong, but what you call “the expert review of the pk calculations” in many cases would lead to doing almost the entire stuff manually anyhow. By ‘manually’ I mean setting the limits for λ_z and getting PK metrics which are not part of WinNonlin’s default output. Phoenix/WinNonlin’s (global) C_min is definitely not what EMA wants to see (the concentration at τ). HVD and t_75% are funny to setup in classical WinNonlin.

You make strong claims on your website. Fine.

If you tell me that you have a sheep in your garden, I will believe you.
If you tell me that you have a unicorn, I will go and see it.