Calculation of the decision limit (CCα) and the detection capability (CCβ) for banned substances: The imperfect marriage between the quantitative and the qualitative criteria

Abstract Initially in the Decision 2002/657/EC the criteria for the calculation of the decision limit (CC α ) and the detection capability (CC β ) have been estimated as purely quantitative ( α -error is 1% and β -error is 5%). In 2004, the European Commission has issued a document to provide guidance for the interpretation of the 2002/657/EC. In this document it is mentioned that also qualitative criteria should be fulfilled. Therefore, the calculated CC α and CC β must be verified by using fortified samples. The method should be able to detect/identify the target component in 50% of the cases at CC α and in 95% of the cases at CC β . Analytical methods for the analysis of nitroimidazoles, nitrofurans and corticosteroids with LC–MS/MS have been validated by fortifying blank samples below and above the MRPL. CC α and CC β were calculated using the ISO 11843 approach. In addition, the frequency of methodical compliance for the qualitative criteria was determined at each concentration level. It was observed that at the calculated CC α and CC β levels the qualitative criteria were not fulfilled. It was concluded that the detection capability of the analytical method should be calculated by using decreasing fortification levels at and below the MRPL. A protocol validating methods for banned substances by limiting the number of samples is presented and the qualitative criteria for the assessment of CC α and CC β were verified based on the same set of data without the need of performing additional validation experiments.