Bioequivalence and Bioavailability Forum

Dear Dan,

❝ IMHO you should also consider if the internal investigation took place before or after statistical evaluation and thus was triggered by an unexpected or unfavorable study result.

Agreed.

I interpreted Roberto's reference to failed runs as meaning that the calibration and / or QCs failed, or that significant interferences were seen, or this kind of things. Considering also that the data available on extracts stability are usually limited to 72 h max, I also assumed that the internal investigation took place before statistical evaluation. To me it would not be acceptable to run the stats, see that the study fails, go back to the chromatograms, say "oh we had a problem with this run" and re-analyse the samples. Whether this would be done by re-injecting the extracts or by completely re-processing them would make no difference.