Implementation of sample pooling to strengthen the self-monitoring in the food industry: Case study of mycotoxins in cereals.

Implementation of sample pooling strategy in the food chemical surveillance could lead to a strengthening of the food safety by increasing the number of analyzable samples. The analysis of a pool of samples and no longer individual samples was applied to the issue of self-monitoring mycotoxins in cereal-based foods, thanks to a data set provided by the surveillance based on ELISA-kits of two mycotoxins - zearalenone (1121 samples) and ochratoxin A (1601 samples) - in four different types of cereal products. After fitting the distribution of mycotoxin concentrations determined in this product category by a Pareto distribution and considering the measurement error in the decision threshold, numerical simulations of pooling were implemented using the Dorfman-2-step strategy. Simulations showed promising results for three out of the four case-studies of zearalenone and ochratoxin A. While being as sensitive and specific as the current one-by-one system, the pooling approach led to a reduction of the number of analyzes performed by 75 - 87 % in three out of the four case studies. Nevertheless, in unfavorable analytical conditions, the pooling approach can lead to an increase of the total number of analyzes.