Poultry Food Assess Risk Model for Salmonella and Chicken Gizzards: I. Initial Contamination

The Poultry Food Assess Risk Model (PFARM) project was initiated in 1995 to develop data collection and mod-eling methods for simulating the risk of salmonellosis from poultry food produced by individual production chains. In the present study, the Initial Contamination (IC) step of PFARM for Salmonella and chicken gizzards (CG) was conducted as a case study. Salmonella prevalence (Pr), number (N), and serotype/zoonotic potential (ZP) data (n = 100) for one sample size (56 g) of CG were collected at meal preparation (MP), and then Monte Carlo simulation (MCS) 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 using a moving window of 60 samples to determine how Salmonella Pr, N, and ZP changed over time in the production chain. The ability of Salmonella to survive, grow, and spread in the production chain and food, and then cause disease in humans was ZP, which was based on U. S. Centers for Disease Control and Prevention data for salmonellosis. Of 100 CG samples tested, 35 were contaminated with Salmonella with N from 0 to 0.809 (median) to 2.788 log per 56 g. Salmonella serotype Pr per 56 g was 16% for Kentucky (ZPmode = 1.1), 9% for Infantis (ZPmode= 4.4), 6% for Enteritidis (ZPmode = 5.0), 3% for Typhimurium (ZPmode = 4.9), and 1% for Thompson (ZPmode = 3.7). Results from MCS indicated that Salmonella Pr, N, and ZP among portions of CG at MP changed (P <= 0.05) over time in the production chain. Notably, the main serotype changed from Kentucky (low ZP) to Infantis (high ZP). However, the pattern of change for Salmonella Pr, N, and ZP differed over time in the production chain and by the statistic used to characterize it. Thus, a performance standard (PS) based on Salmonella Pr, N, or ZP at testing or MP will likely not be a good indicator of poultry food safety or risk of salmonellosis.

Automated summary

The PFARM project was developed in 1995 to simulate the risk of salmonellosis from poultry food. In this study, the Initial Contamination step was conducted as a case study. Data for Salmonella prevalence, number, and serotype/zoonotic potential were collected and simulated over time in the production chain. Results showed that the pattern of change for Salmonella differed over time and by the statistic used, indicating that a performance standard based on Salmonella may not be a good indicator of food safety or risk of salmonellosis.