Probabilistic MIC modelling for AMR risk assessment

Understanding the minimum inhibitory concentration (MIC) of antibiotics is crucial for developing targeted strategies to combat antimicrobial resistance (AMR) in the food chain. Traditional MIC testing methods provide a single and deterministic value without taking into account the inherent inter-individual variability in the response of foodborne pathogens to antimicrobials. A more accurate representation of the complex interactions influencing AMR in foodborne pathogens could be probabilistic MIC modelling. The incorporation of probabilistic MIC modelling into a classic quantitative microbiological risk assessment (QMRA) of AMR for food pathogens could provide a more realistic risk estimate. The EU-FORA fellowship's objective was to develop a probabilistic QMRA for an antimicrobial-resistant foodborne pathogen in a food product, incorporating the variability of the MIC of individual bacterial cells. The project involved the collection of data on Listeria monocytogenes in ready-to-eat (RTE) cooked ham, the selection of an appropriate growth model, the conducting of MIC assessment laboratory experiments for ampicillin based on the plate-count agar method and Monte Carlo analysis. The QMRA model was constructed using the R programming language. The final outputs obtained were a total of 1000 simulated doses of L. monocytogenes in servings of cooked ham at the time of consumption, as well as the maximum and 95th percentile of single-cells MIC values of ampicillin for each dose. Moreover, a sensitivity analysis was conducted.