Purpose: To develop high-performance microbial risk assessment software, incorporating variable environmental conditions in the supply chain and actual consumer consumption patterns.
Methods: Predictive models were implemented stochastically for a number of pathogen/commodity pairs to simulate all potential variability at various points in the supply chain (e.g., variable initial conditions, storage times, temperatures, pH, etc.). Extensive experimental data from the ComBase database was used to estimate growth parameters. The calculated distribution of pathogen concentrations in a number of commodities was combined with probabilistic dietary exposure and dose-response models in order to estimate the disease incidence in a population of consumers, using national consumption surveys from the EU and the US (NHANES). The power of cloud computing was used in order to handle the large data sets and computational effort required.
Results: The method was successfully applied to a number of scenarios, including E. coli O157:H7 in beef and Salmonella Typhimurium in chocolate. An assessment involving 100,000 simulated consumers using the US NHANES survey, 40 food beef commodities contaminated with varying levels of E. coliO157:H7, under varying multi-stage environmental conditions, can be completed in less than 30 minutes. Complete output for a given assessment was stratified to determine e.g., the drivers of pathogen growth, consumer exposure, vulnerable subpopulations, etc., enabling appropriate risk mitigation strategies to be identified.
Significance: This presents a valuable tool that can be used for routine risk assessment and rapid post-outbreak analysis. The tool was developed for the food industry, regulatory bodies, and academic research in microbial food safety, and is extensible to a broad number of pathogens, commodities, and consumer populations.