How far can end-product testing of quality be anticipated, and superseded, by the introduction of parametric release criteria enforced during production? Dr Klaus Haberer (a microbiologist from Hoechst, Germany) defined parametric release as "a system of release that gives assurance of quality, based on information collected during manufacture and on compliance with GMP [good manufacturing practice]”. He posed two key questions of conflict consequential on operating parametric release: did end product testing then have any meaning or value, and would agencies impose many extra requirements?
He instanced heavily loaded permission for release of sterile products from the Food and Drug Administration for two companies, Pfizer and Baxter. He suggested that parametric release was probably most useful and feasible where there were long production runs.
Dr Haberer distinguished "useful and non-useful "situations: parametric release was useful if based on existing batch and validation data, derived by internationally agreed analytical principles. But it was not useful if a great deal of additional data had to be compiled and different prerequisites were imposed by different regulatory agencies. He urged removal of terminal biological testing, which he regarded as "meaningless "and "beset by statistical absurdity "(eg, the need to test three million containers for 95 per cent confidence that only one in a million would fail). He quoted Annex 17 of the European GMP guide, which recognised a final biotest as an "only "opportunity to detect major failure, but also "a last chance in an inherently insecure operation like aseptic processing”. Dr Haberer remained pessimistic about official enthusiasm for parametric release and he quoted a series of prerequisites that had been proposed. Agencies would expect critical "hazard analysis "and a "sterility assurance system”, with explicit "operational qualification”, experience of automated systems, and a qualified microbiologist on site. Nevertheless, he summarised how parametric release, ideally, should operate: agency approval in principle should be sought at the time of GMP inspection, and consideration should be taken of previous non-sterility experience, coupled with a "firm commitment to compliance with GMP”.