IN-HOUSE VALIDATION FOR NITROFURAN METABOLITES ANALYSIS ACCORDING TO THE COMMISSION DECISION 2002/657/ EC (ALTERNATIVE METHOD) PERFORMED BY SOFTWARE 'INTERVAL'

摘要

In the framework of Council Directive 96/23/EC and of Commission Decision 2002/657/EC as well as according to ISO 17025, only validated test methods are to be used in residue control. This paper presents a validated method for determining nitrofuran metabolites Furazolidone (NP-AOZ), Nitrofurazone (NP-SEM), Furaltadone (NP-AMOZ), and Nitrofurantoin (NP-AHD) in 5 different matrices (bovine muscle, porcine muscle, poultry muscle, fish muscle and honey) in accordance with an alternative model validation, as referred to the Commission Decision 2002/657/EK (paragraph 3.1.3.). The analytical technique that was used for the analysis of the samples was that of liquid chromatography-tandem mass spectrometry, triple quadrupole technology. The alternative validation procedure (2002/657/EC) was performed by means of the software "Interval". This validation software allows an efficient and uniform validation of test methods for residues, in order to ensure the comparability of the performances of different test methods and to be able to judge the meaningfulness of the test results. The conventional validation method assess an analytical method by performing repeated measurements. However, the performance of the method depends on many influence factors (e.g. operator, sampling conditions, analytical conditions, matrices...), and these are randomly varying. If you hold the influence factors constant, then the validation is valid only for this setting, then many repetitions are necessary in the conventional validation method. Interval offers optimized orthogonal and randomized experimental design. The effects of influence factors are analyzed directly by controlled variation of the experimental conditions which allows a considerable reduction of the number of measurements. For example, for nitrofuran metabolites two factors were selected ("matrices", "operators") and seven factor levels (five matrices and two operators). According to the conventional validation method described by 2002/657/EC, we have at least: 5 matrices * 6 replicates * 3 days * 3 concentration levels = 270 measurements + additional controls. According to the design with 5 matrices proposed by Interval we have 5 matrices * 8 days (runs) * 4 concentration levels = 160 measurements + additional controls. With this design the number of measurements can be reduced by more than 40%. Other advantages of Interval: (1) Measurement uncertainty as a function of the concentration, (2) Matrix-comprehensive in-house reproducibility accross matrices, (3) Analysis of factorial effects provides insight into specific problems, (4) These experimental designs allow a combined analysis of robustness.