Application of EU guidelines for the validation of screening methods for veterinary drugs

Commission Decision (CD) 2002/657/EC describes detailed rules for method validation within the framework of residue monitoring programmes. The approach described in this CD is based on criteria. For (qualitative) screening methods, the most important criteria is that the CCβ has to be below any regulatory limit. Especially when microbiological or immunochemical methods are involved, the approach described in the CD is not easily applied. For example, by those methods, a large number of analytes (all antibiotics) within several different matrices (meat, milk, fish, eggs, etc.) are detected. It is not completely clear whether all those analytes and all matrices have to be taken into account during method validation. To clarify this, a working group - from EU Reference Laboratories - came up with a practical approach to validate multi-analyte multi-matrix screening methods. It describes how many analyte/matrix combinations have to be tested and how these combinations are selected. Furthermore it describes how to determine CCβ for screening methods in relation to a large list of compounds and maximum residue limits (MRLs). First for each analyte/matrix combination the 'cut-off' level - i.e. the level at which the method separates blanks from contaminated samples - is established. The validation is preferably at the concentration of 50% of the regulatory limit. A minimum set of 20 different samples has to be tested. From the experiences with applying these guidelines it was concluded that the validation approach is very 'practical'; however, there are some remarks. One has to be careful with selecting 'representative' analytes and matrices and it is strongly recommended to collect additional validation data during the routine application of the method.