IAG proficiency test animal proteins 2019

The annual proficiency test for the detection of animal proteins in animal feed of the IAG - International Association for Feeding stuff Analysis, Section Feeding stuff Microscopy was organized by Wageningen Food Safety Research, The Netherlands. The aim of the proficiency test was to provide the participants information on the performance of the implementation of the monitoring methods for their local quality systems. A further aim was to gather information about the current practices in the application of the microscopic method. The current 2019 version of the IAG ring test for animal proteins addressed all analytical sections of the methods for microscopy and PCR as published in Regulation (EC) 51/2013 amending Annex VI of Regulation (EC) 152/2009 together with accompanying SOPs. Three of the four samples used in the proficiency test contained poultry material at the legally required technical limit (0.1% w/w; Regulation (EC) 152/2009), or fish meal at a spike level of 2% (w/w), or both. A fourth sample was left blank. A pig feed, containing 3% (w/w) of bakery by-products and a ruminant feed were used as matrix. None of the samples was labelled as fish feed. A total of 44 participants subscribed to the proficiency test animal proteins. Two participants did not submit their results and one submitted PCR results only, leaving 41 sets for microscopic evaluation. 18 sets of ruminant PCR results were submitted as well. Microscopy All participants were requested to determine the presence or absence of land animal and/or fish, to indicate the type of material found and to describe the method used to achieve these results. In total eight participants (19.5% of 41 participants) deviated from the official method by applying an incorrect number of determination cycles and/or drawing incorrect conclusions (e.g. “presence” for five particles, “absence” for ten particles). Therefore, all evaluations were based on the actual number of particles reported by all participants. Incorrect positive results (positive deviations) were expressed in a specificity score and incorrect negative results (negative deviations) were expressed in a sensitivity score. An optimal score is 1.0. The results are analysed in two ways: numbers below threshold (between 1 and 5 particles per determination cycle inclusive) have been considered positive (complying to the zero tolerance) and as alternative considered as negative (for matching the official evaluation). For all samples several participants did not detect terrestrial animal particles in the presence of fish meal (sensitivity 0.95) in contrast to the optimal result in the absence of fish meal (1.0), or erroneously reported terrestrial animal material when absent (specificity 0.93 and 0.90 in the presence or absence, respectively, of fish material). The absence of fish material in the presence of 0.1% poultry PAP resulted in a specificity score of 0.90. 37 institutes participated in both in the 2018 and 2019 studies. Based on their results an intra-laboratory reproducibility, expressed as concordance between 2018 and 2019 was calculated. Especially for the results representing specificity low concordance was found. This indicates wrong observations seemed incidental in most cases. The documentation for and training of microscopists for correct identification of particles of animal origin would deserve further attention in order to guard specificity and avoid incidental errors. Evaluation of several aspects of the application of the current microscopic methods would be beneficial for improving harmonization among the laboratories applying the microscopic method. PCR In the two samples without addition of ruminant PAP, but still containing the bakery by-products, ruminant DNA was detected by qPCR as far as analysed by the majority of the participants. The list of recognised sources such as milk and milk products, and ruminant gelatine can be extended with bakery by-products, which is important for the recycling of food by-products.