Validation of methods used in the Florida Department of Agricultural and Consumer Services' Chemical Residue Laboratory.

Very few methods for detecting residues of pesticides in food or agricultural samples have undergone rigorous colloborative study and possess official AOAC status. The Chemical Residue Laboratory has formalized a method validation scheme to use when incorporating or developing new, unofficial methods. These methods are validated by assessing certain performance parameters: scope, specificity, linear range, accuracy, precision, limit of detection (LOD), and limit of quantitation (LOQ). For accuracy and precision assessment, 12 replicate fortifications must yield recoveries within the range of 70-120% with a coefficient of variation (CV) that compares favorably to the Horwitz CV. LOD and LOQ are equivalent to 3 and 10 times, respectively, the background signal contributed by a sample matrix blank. This criterion that we use for LOD/LOQ is not universal. In fact, because of differing definitions, we have encountered difficulties in enforcing a tolerance by using a registrant's method. This paper also presents an example of our method validation scheme, using a recent method development project for detecting sulfamethazine in raw milk. The sulfamethazine project also revealed unanticipated personnel problems, underscoring the importance of the human factor in quality assurance.