Evaluation of Mycotoxin Screening Tests in a Verification Study Involving First Time Users

Rapid screening methods are currently recognized as a strategic tool for mycotoxin issues management. Specific guidelines for validation and verification of mycotoxin screening methods are set in the Commission Regulation (EU) No 2014/519. This regulation establishes that the “aim of the validation is to demonstrate the fitness-for-purpose of the screening method” and focuses the entire validation procedure on determining specific cut-off values ensuring a maximum rate of false negative results of 5%. In addition, the assessment of the rate of false suspect results is addressed. With regard to rapid test-kits, ‘fitness-for-purpose’ includes not only the criteria more commonly considered when discussing laboratory-based methods (specificity, accuracy, and precision), but also more “practical” parameters such as speed and ease of implementation in a new operational environment. The latter means demonstrating under local conditions that performance parameters, as established during the validation, can be achieved by first time users. This goal can be achieved through “method verification”. The aim of the present study was to verify the fitness-for-purpose of mycotoxin screening methods when applied by first time users. This was achieved in one laboratory facility via results of a training course with multiple technicians attending. The verification study was organized similarly to a collaborative exercise and involved two groups comprising of 10 technicians each that used the methods for the first time. Different screening methods were applied for deoxynivalenol (DON) in wheat, which was mainly Enzyme Linked Immunosorbent Assay (ELISA), lateral flow device (LFD), fluorescence polarization immunoassay (FPIA), and liquid chromatography-high resolution mass spectrometry (LC-HRMS). An additional verification was done for aflatoxin B(1) (AFB(1)) in maize and wheat using LFD and LC-HRMS, respectively. The results of analyses were used to calculate intermediate precision (RSD(ip), covering the inter-analyst variability in preparing the analytical samples and the precision under repeatability conditions) cut-off values and false suspect rates. RSD(ip) ranged from 6.5% to 30% for DON, and from 16% to 33% for AFB(1). The highest obtained variances were associated with the AFB(1) analyses due to working with much lower mass fractions. The rate of false suspect results were lower than 0.1% for all tested methods. All methods showed a fit-for-purpose method performance profile, which allowed a clear distinction of samples containing the analytes at the screening target concentration (STC) from negative control samples. Moreover, the first time users obtained method performances similar to those obtained for validation studies previously performed on the screening methods included in the training course.