Report on the 2017 proficiency test of the European Union reference laboratory for mycotoxins : determination of deoxynivalenol in wheat

The Joint Research Centre (JRC), a Directorate-General of the European Commission operates the European Union Reference Laboratory (EURL) for Mycotoxins. One of its core tasks is to organise proficiency tests (PTs) among designated National Reference Laboratories (NRLs). This report presents the results of the PT on the determination of deoxynivalenol in wheat. The test materials for this PT were four naturally contaminated wheat materials. The test items were dispatched to the participants at the end of April 2017. Each participant received one test item per material containing approximately 55 g each. Fifty-nine participants from 32 countries (among them 41 NRLs and 18 official food control laboratories-OCLs) registered for the exercise and 59 sets of results for test items A, B, C and D) were reported. The assigned values, established by an exact-matching double isotope dilution mass spectrometric technique (EMD-IDMS), were 551 μg/kg (± 37 μg/kg) deoxynivalenol for material A, 1556 μg/kg (± 83 μg/kg) for material B, 4405 μg/kg (± 265 μg/kg) for material C and 1160 μg/kg (± 60 μg/kg) for material D. Participants' results were rated with z-scores and zeta-scores in accordance with ISO 13528:2015. The z-score compares the participant's deviation from the reference value with the target standard deviation accepted for the PT, whereas the zeta-score indicates whether the participant's estimate of uncertainty is consistent with the observed deviation from the assigned value. Only z-scores were used for the evaluation of whether an individual laboratory underperformed. In total, 93 % of the attributed z-scores were below an absolute value of two for test items A and C, 95 % for test item B and 92 % for test item D. This indicates that most of the participants performed satisfactorily. One NRL had a z-score above an absolute value of 3 and will have to investigate the reasons for the deviation (root-cause analysis) and report the planned corrective actions to the EURL. Participants were requested to assess the compliance of the sample against legislative limits. Eighty-five percent to 100 % of the participants assessed the compliance/non-compliance of the test materials A, C and D correctly. Only 36 % of the laboratories assessed correctly the non-compliance of Material B, and 56 % classified the test material as compliant providing a proper justification (taking into account their measurement result and reported uncertainty).