A second-order Monte Carlo simulation model to describe coagulase-positive Staphylococci growth in mayonnaise served in Algerian pizzerias

To bridge the data gap on food poisoning caused by coagulase-positive staphylococci (CoPS), especially related to mayonnaise sauce served at Algerian pizzerias, this study aimed to assess the concentration of CoPS in mayonnaise and the probability of exceeding a critical concentration of ≥ 5 log CFU/g. The city of Ain Témouchent in West Algeria was taken as a case study.

A probabilistic assessment model was built, taking into account the initial contamination in freshly made mayonnaise and the potential growth before serving. Uncertainty and variability were integrated separately in the model. Uncertainty came from lack of data and model fitting error, variability from natural heterogeneity of biological materials (e.g., microbial strains) and temperature during cold storage.

The second-order Monte Carlo procedure was implemented in R using the mc2d package. The following pieces of data were generated to populate the model: CoPS were enumerated and characterized from 57 samples of mayonnaise served at pizzeria in Ain Témouchent city; challenge tests at 23 °C were performed in mayonnaise using three CoPS isolates. The following existing data were also gathered: meteorological data from Ain Témouchent were analysed to build a realistic scenario of storage, while a set of 43 and 35 values of the minimal and maximal growth limits of CoPS, respectively, were collected from the literature and analysed to inform a secondary predictive model describing the growth rate at various storage temperature conditions. A sensitivity analysis was performed to facilitate the interpretation of the results.

The results revealed a CoPS prevalence in freshly made mayonnaise of 25% [15%; 37%] with concentrations varying from 0.4 [0.3; 0.9] to 2.9 [2.4; 3.0] log CFU/g. The growth rates at 23 °C, based on challenge tests in mayonnaise, had a median value estimated to be 1.41 [1.17; 1.65] h ^− 1.

Concentration levels according to various scenarios of temperature and serving conditions were calculated. For instance, the median contamination of CoPS in mayonnaise after storage in a refrigerated display counter for 14 h during the hottest months of the year was estimated to be zero. However, the 95th percentile was estimated to be 3.6 [2.9; 4.2] log CFU/g. In this scenario, the probability of exceeding a critical concentration of ≥ 5 log CFU/g was estimated to be 1% [0.3%; 2%], which is low but not negligible.

These findings could be used to improve food safety policies and develop a risk management strategy to reduce the food poisoning associated with the consumption of ready-to-use foods in Algerian fast food restaurants.