A detailed procedure for in-house validation of methods, including experimental design, statistics and evaluation of assumptions, was proposed and applied to the validation of a method for screening avermectins in cow's milk by high performance liquid chromatography coupled with fluorescence detection. Assays were performed with solvent and matrix-matched calibration curves, milk blanks and spiked samples. The assumptions about the statistics employed were evaluated and confirmed. Linearity was achieved between 25 and 150 ng.mL-1. No significant matrix effects were observed in this range. Mean apparent recoveries varied from 87.2 to 101.4%. The relative standard deviations under conditions of repeatability ranged from 3.3 to 11.2%, while the relative standard deviations of partial reproducibility varied from 7.4 to 14.7%. These results indicated exactness and precision of the method between 10 and 30 µg.L-1 for the four analytes studied. The experimental limits of detection and quantification were 5 and 10 µg.L-1, respectively. Decision limits (12.6 to 13.7 µg.L-1) and detection capabilities (15.1 to 17.4 µg.L-1) were estimated, assuming a maximum residue level of 10 µg.L-1.
in-house validation; abamectin; doramectin; eprinomectin; ivermectin, cow's milk; high performance liquid chromatography
