Statistical aspects of food safety sampling

In food safety management, sampling is an important tool for verifying control. Sampling by nature is a stochastic process. However, uncertainty regarding results is made even greater by the uneven distribution of microorganisms in a batch of food. This article reviews statistical aspects of sampling and describes the impact of distributions on the sampling results. Five different batch contamination scenarios are illustrated: a homogeneous batch, a heterogeneous batch with high- or low-level contamination, and a batch with localized high- or low-level contamination. These batch contamination scenarios showed that sampling results have to be interpreted carefully, especially when heterogeneous and localized contamination in food products is expected.