Poultry food assess risk model for salmonella and chicken gizzards: I. Initial contamination

Authors: Oscar, Thomas

Submitted to: Journal of Food Protection
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/27/2022
Publication Date: 1/2/2023
Citation: Oscar, T.P. 2023. Poultry food assess risk model for salmonella and chicken gizzards: I. Initial contamination. Journal of Food Protection. 86(2). https://doi.org/10.1016/j.jfp.2022.100036.
DOI: https://doi.org/10.1016/j.jfp.2022.100036

Interpretive Summary: The Poultry Food Assess Risk Model (PFARM) for Salmonella can be used to identify higher risk poultry production chains and thus, could serve as the basis of a new food safety system. In this study, the hazard identification step of PFARM for Salmonella was demonstrated using chicken gizzards as a case study. Results indicated that at meal preparation just before cooking 35% of chicken gizzard servings examined were contaminated with Salmonella at levels from 1 to 1,000 cells with a most likely level of 2 to 3 cells per 56 g or 2 ounces or per two whole chicken gizzards. The next step in the project is to use PFARM to determine consumer exposure and response to Salmonella contamination of chicken gizzards at meal preparation by considering meal preparation practices, food consumption behavior, and consumer resistance to salmonellosis in the production chain.

Technical Abstract: The Poultry Food Assess Risk Model (PFARM) was developed to assess risk of salmonellosis from poultry food contaminated with Salmonella and produced by different farm-to-table scenarios. In the present study, the first step of the PFARM process, which is hazard identification, was demonstrated using Salmonella and chicken gizzards as a case study. Data for Salmonella contamination (prevalence, number, serotype) of a single brand and source of chicken gizzards were collected at meal preparation using a combination of methods: 1) whole sample enrichment (WSE); 2) quantitative polymerase chain reaction (qPCR); 3) culture isolation; 4) serotyping; and 5) Monte Carlo simulation. Data (n = 100) were collected with one sample size (56 g) and then Monte Carlo simulation was used to obtain data for other sample sizes (112, 168, 224, 280 g). The PFARM was developed in Excel and was simulated with @Risk. Data were simulated in sets of 60 consecutive samples to determine how Salmonella contamination of chicken gizzards at meal preparation changed over time in the production chain. Of 100 samples examined, 35 were found to harbor Salmonella at levels from 0 to 0.4 (mode) to 2.8 log per 56 g. Salmonella serotype prevalence per 56 g was 65% for none, 16% for Kentucky, 9% for Infantis, 6% for Enteritidis, 3% for Typhimurium, and 1% for Thompson. Distributions for Salmonella serotype prevalence and number among portions of chicken gizzards at meal preparation changed (P = 0.05) over time in the production chain. Although Salmonella prevalence and number on chicken gizzards at meal preparation decreased (P = 0.05) over time in the production chain, the zoonotic potential (ZP) of Salmonella on the chicken gizzards at meal preparation increased (P = 0.05) over time because the predominant serotype changed from Salmonella Kentucky, which has low ZP, to Salmonella Infantis, which has high ZP. Thus, it was not clear from the hazard identification in PFARM, how poultry food safety was affected over time in the production chain.