The objective of this study was to model and quantify the level of Listeria monocytogenes in raw milk cheese (RMc) and pasteurized milk cheese (PMc) from farm to fork using a Bayesian inference approach combined with a quantitative risk assessment. The modeling approach included a prediction of contamination arising from the farm environment as well from cross-contamination within the cheese-processing facility through storage and subsequent human exposure. The model predicted a high concentration of L. monocytogenes in contaminated RMc (mean 2.19 log10 CFU/g) compared to PMc (mean -1.73 log10 CFU/g). The mean probability of illness (P1 for low-risk population, LR) and (P2 for high-risk population, HR, e.g., immunocompromised) adult Irish consumers following exposure to contaminated cheese was 7 × 10(-8) (P1 ) and 9 × 10(-4) (P2 ) for RMc and 7 × 10(-10) (P1 ) and 8 × 10(-6) (P2 ) for PMc, respectively. In addition, the model was used to evaluate performance objectives at various stages, namely, the cheese making and ripening stages, and to set a food safety objective at the time of consumption. A scenario analysis predicted various probabilities of L. monocytogenes contamination along the cheese-processing chain for both RMc and PMc. The sensitivity analysis showed the critical factors for both cheeses were the serving size of the cheese, storage time, and temperature at the distribution stage. The developed model will allow food processors and policymakers to identify the possible routes of contamination along the cheese-processing chain and to reduce the risk posed to human health.
Keywords: Bayesian inference; L. monocytogenes; dose response; pasteurized milk cheese; raw milk cheese; risk assessment.
© 2015 Society for Risk Analysis.