Modeling the prevalence of Bacillus cereus spores during the production of a cooked chilled vegetable product

Abstract

In minimally processed vegetable foods, pathogenic spore-forming bacteria pose a significant hazard. As part of a quantitative risk assessment, we used Bayesian belief methods to model the uncertainty and variability of the number of Bacillus cereus spores that can be found in packets of a vegetable puree. The model combines specific information from the manufacturer, experimental data on inactivation of spores, and expert opinion concerning spore concentrations in the raw vegetables and ingredients. Sensitivity analysis revealed that spore contamination of added ingredients contributes most uncertainty to the assessment. The assessment produced a quantitative estimate of the prevalence of B. cereus spores in packets of vegetable puree at the end point of the manufacturing process.