Modeling the Risk of Salmonellosis from Consumption of Peanuts in the United States

Peanut products were the target of the largest food recall in United States history from 2008 to 2009, with more than 3,200 products implicated, economic losses estimated at $1 billion, and more than 700 reported illnesses and 9 deaths. Predictive modeling tools such as quantitative microbial risk assessment can be used to aid processors in making risk management decisions that may reduce the chances of foodborne illness, but published risk assessment for peanuts is not currently available. A quantitative microbial risk assessment was performed to quantify salmonellosis risk from consumption of peanuts in the United States. Prevalence and concentration data for Salmonella on raw, shelled peanuts were used in combination with probability distributions of simulated log reductions achieved during production steps before consumption. Data for time-temperature combinations used in each step were obtained from published literature, industry surveys, or expert opinion, and survival data were obtained from the literature. A beta-Poisson dose-response model was used to predict probability of illness from ingestion of Salmonella cells. The model predicted 14.2 (arithmetic mean) or 0.0123 (geometric mean) illnesses per year. Sensitivity analysis showed that thermal inactivation log reductions applied had the biggest impact on predicted salmonellosis risk, followed by consumer storage time, Salmonella starting concentration, Salmonella starting prevalence, and number of originally contaminated 25-g servings per originally positive 375-g sample. Scenario analysis showed that increasing log reduction variability increased mean salmonellosis risk. Removing the effect of storage on Salmonella survival increased the arithmetic and geometric means to 153 and 0.598 illnesses per year, respectively. This study indicated that the risk of salmonellosis from consumption of peanuts can be lowered by reducing field contamination, control of storage steps, and monitoring of appropriate critical limits in peanut roasting.