Submitted to: International Journal of Food Microbiology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 1, 2003
Publication Date: February 2, 2004
Citation: Oscar, T.P. 2004. A quantitative risk assessment model for salmonella and whole chickens. International Journal of Food Microbiology. 93:231-247. Interpretive Summary: The rate of salmonellosis in the United States is between 15 and 20 cases per 100,000 people. Approximately 10% of salmonellosis cases are caused by poultry meat with 6.6% from turkey and 4.4% from chicken for a rate of 0.66 to 0.88 cases of salmonellosis per 100,000 consumers of chicken. Preventing outbreaks and sporadic cases of salmonellosis from chicken requires a holistic approach to chicken safety. One such holistic approach is risk assessment. Published data and models for growth and death of Salmonella were used to create a risk assessment model for Salmonella and whole chickens that considered the effects of consumer transport, cooking and handling on the risk of salmonellosis from whole chickens. Simulation results of the model indicated that consumer transport was not an important risk factor, whereas undercooking and cross-contamination of cooked chicken with Salmonella from uncooked chicken during serving were important risk factors. The model predicted a rate of 0.44 cases of salmonellosis per 100,000 consumers of chicken, which is consistent with the aforementioned estimate based on surveillance data.
Technical Abstract: A previously developed quantitative risk assessment model (QRAM) for Salmonella and whole chickens was modified and used to demonstrate how existing data and predictive models could be used to define its input settings. The QRAM was constructed in an Excel spreadsheet and was simulated using @Risk. The product testing-to-table pathway was modeled as a series of pathogen events that included initial contamination at product testing, growth during consumer transport, thermal inactivation during cooking, cross-contamination during serving, and dose-response after consumption. Published data as well as predictive models for growth and thermal inactivation of Salmonella were used to establish input settings. Non-contaminated chickens were simulated so that the QRAM could predict changes in the incidence of Salmonella contamination. The incidence of Salmonella contamination changed from 30% at product testing to 0.16% after cooking to 4% at consumption. Salmonella growth on chickens during consumer transport was the only pathogen event that did not impact the risk of salmonellosis. For the scenario simulated, the QRAM predicted 0.44 cases of salmonellosis per 100,000 consumers, which was consistent with recent epidemiological data that indicates a rate of 0.66 to 0.88 cases of salmonellosis per 100,000 consumers of chicken.