PROBABILISTIC MODELLING IN FOOD SAFETY: A SCIENCE-BASED APPROACH FOR POLICY DECISIONS

This thesis deals with use of qualitative and quantitative probabilistic models for the animal-derived food safety management. Four unrelated models are presented: three quantitative and one qualitative. Two of the quantitative models concern the risk posed by pathogens in raw milk, in the first study, a probabilistic approach for the inclusion of the variability and the uncertainty in the consumers’ habits and the bacterial pathogenic potential is proposed while the second study, demonstrate how the overlook of the relationship between the storage time and temperature has led to overestimated results in raw milk-related models published so far and an equation to address the issue is provided. In the third study, quantitative modelling techniques are used to simulate the dynamics underlying the spread of Campylobacter in broiler flocks and quantify the potential effects that different on-farm mitigation strategies or management measures have on the microbial load in the intestine of infected birds at the end of the rearing period. In the qualitative study, a general approach for the estimation of the likelihoods of introduction of live parasites in aquaculture implants and the commercialization of infested product is outlined by using the example of Anisakids in farmed Atlantic salmon.