Organic Contaminants and Additives, Scientific Directorate of Chemical and Physical Health Risks, Sciensano, Leuvensesteenweg 17, 3080 Tervuren, Belgium.
Toxins (Basel). 2021 Aug 24;13(9):591. doi: 10.3390/toxins13090591.
In accordance with the International Standard Organization ISO 17043, two proficiency tests (PTs) for the simultaneous determination of aflatoxins (AFB, AFB, AFG, AFG); deoxynivalenol; fumonisins FB, FB and B; ochratoxin A, the T-2 toxin; and the HT-2 toxin were conducted in 2019 and 2020 using cornflakes and rusk flours that were prepared in house. The homogeneity and the stability of these materials were verified according to the criteria laid down in ISO 13528 using randomly selected samples. Most of the targeted toxins were found to be homogenously distributed in both materials with no significant changes during the timescale of the PTs. Next, the materials were distributed to approximately 25 participating laboratories from Europe, Canada, and the United States. The obtained datasets were computed using robust statistics. The outliers were checked and removed, and the toxin concentrations were assigned as the consensus value of the results of the participants at Horwitz ratios <1.2. The z scores were generated for all mycotoxins, and the results were pooled to calculate the relative sum of squared z scores (SZ2) indexes and were clustered according to the triple A rating. Overall, at least 80% of the participating laboratories achieved good and acceptable performances. The most frequent categories assigned to good performances (SZ2 ≤ 2) were AAA (51%) and BAA (13%). Clusters of BBA + CBA (6%) included laboratories reporting acceptable z scores <90% of the total z scores for less than 90% or 50% of the mycotoxins targeted in the 2 matrices. The triple A rating seems to be appropriate in evaluating the performances of laboratories involved in multi-mycotoxin analyses. Accredited and non-accredited analytical methods achieved good and acceptable performances.
根据国际标准化组织（ISO）17043 的规定，于 2019 年和 2020 年使用自制的玉米片和脆皮面包粉进行了两次同时测定黄曲霉毒素（AFB₁、AFB₂、AFG₁、AFG₂）、脱氧雪腐镰刀菌烯醇、伏马菌素 FB₁、FB₂ 和 B、赭曲霉毒素 A、T-2 毒素和 HT-2 毒素的能力验证试验（PT）。根据 ISO 13528 中规定的标准，使用随机选择的样品验证了这些材料的均一性和稳定性。结果发现，大多数目标毒素在这两种材料中均匀分布，在 PT 期间没有明显变化。然后，将这些材料分发给来自欧洲、加拿大和美国的约 25 个参与实验室。使用稳健统计方法计算获得的数据组。检查并剔除离群值，将毒素浓度指定为参与者在霍维茨比（Horwitz ratio）<1.2 时的结果的共识值。为所有真菌毒素生成 Z 分数，将结果汇总以计算相对平方和 Z 分数（SZ2）指数，并根据三重 A 评级进行聚类。总体而言，至少 80%的参与实验室表现良好和可接受。分配给良好表现（SZ2 ≤ 2）的最常见类别是 AAA（51%）和 BAA（13%）。BBA+CBA（6%）的聚类包括报告可接受 Z 分数（<90%的总 Z 分数）<90%或<50%的目标真菌毒素的实验室，这些真菌毒素分别来自 2 个基质中的不到 90%或 50%。三重 A 评级似乎适合评估参与多真菌毒素分析的实验室的表现。经认可和未经认可的分析方法均取得了良好和可接受的结果。
