Modeling the impact of between-lot and within-lot variability in Listeria monocytogenes contamination on risk reduction from sampling ready-to-eat foods.

Microbiological sampling and testing are widely utilized in food safety risk management. We developed risk assessments to quantify the impact of various sampling plans on the risk of invasive listeriosis to consumers. We used the FDA-iRISK® tool and adapted available process, consumption, and dose response modules of published L. monocytogenes risk assessments to predict cases per billion servings from consumption of ready-to-eat foods. We also developed an ad hoc quantitative risk assessment application using R (the "FDA-LmQRA App") to evaluate more complex scenarios and provide additional metrics. Data and model inputs included the prevalence and levels of L. monocytogenes, sampling plan parameters, and operating characteristic curves data. We derived prevalence and between-lot distributions from data from market basket surveys of L. monocytogenes in ready-to-eat foods in the U.S. and examined two assumptions for the within-lot contamination: a lognormal distribution, or a heterogeneous distribution with a defined proportion of exceptional (higher level) contamination in addition to a single lognormal distribution. We found that testing each lot using 2-class plans (e.g., n=5 or 10, m=0/25 g or 0/5 g, and c=0) or 3-class mixed plans (e.g., n=5 or 10, m=0/25 g or 0/5 g, M=20 CFU/g or 100 CFU/g, and c=1) and replacing positive lots by non-contaminated lots predicted quantifiable, but relatively low, risk reduction. The risk estimates were highly influenced by the variability of the between-lot concentration distribution as well as the presence of exceptional contamination for the within-lot contamination. In the presence of exceptional contamination, a 3-class mixed plan (c=1) was predicted to have comparable performance based on risk estimates to a 2-class plan (corresponding n and m but c=0). Results from this study may inform the choice of sampling plans to optimize sampling and testing strategies for reducing listeriosis associated with ready-to-eat foods.