Inspection by variables as an acceptance criterion in bioanalysis--a proposal.

Inspection by variables is proposed as an acceptance criterion for use in bioanalysis. The criteria currently used are deficient either by ignoring the issues of precision (fixed range) and/or accuracy (99% confidence interval), not being able to provide immediate answers (quality charts), or even not being scientifically justified (fixed range). Inspection (sampling) by a variables procedure was originally developed to drive the quality of military supplies (MIL-STD-414) and was consequently incorporated in ISO 3951 as a part of industrial quality control. It is based on the concept of acceptable quality level (AQL), which is the maximum per cent defective (number of results outside of specification per 100 results) that can be considered satisfactory as a process average. It also correlates sample size with batch size. An AQL of 6.50% is proposed as a standard with specification set at +/- 20% or +/- 15%, which should result in coefficients of variation of approximately 15% and 12%, respectively. This concept has been applied post factum to 14 authentic data sets, thus proving its utility and validity.