Reply from Helmut, 2012-10-18 16:14:
Merhaba Haydonat!
» We determine the medium QC […] by using the geometric mean of LLOQ and ULOQ to be in the range of concentration levels in study samples. But in EMEA bioanalytical method validation guideline, medium QC is described to be around 50% of the calibration curve range.
Good point!
» Does using the geometric mean instead of the 50% of calibration curve range for calculating medium QC pose a problem during an EMEA audit?
Don’t know. Personally I haven’t seen a deficiency letter covering this topic yet – but the new GL is in force for a short time now… I’m interested in opinions/experiences of other forum members.
Actually the geometric mean was suggested in the Arlington III conference report – but only for Ligand Binding Assays. Strange IMHO because a 4-parameter logistic function is symmetric. The geometric mean makes more sense for conventional assays, where the calibration range covers several orders of magnitude and calibrators are logarithmically spaced.
I will prepare are more detailed answer later on.