Development of predictive models for the growth of Listeria monocytogenes on ready-to-eat meat and poultry products

Listeria monocytogenes (L. monocytogenes ) is a foodborne pathogen which has been implicated in several outbreaks of foodborne disease (listeriosis). The high risk population includes pregnant women, neonates and immunocompromised adults; among these populations the mortality rate can reach about 25%. The U.S. Centers for Disease and Control and Prevention (CDC) estimates that 2,500 cases of listeriosis occur each year, resulting in 500 deaths. L. monocytogenes is ubiquitous in the environment and has been isolated from the air, water, soil and the gastrointestinal tract of humans. The USDA-FSIS requires RTE meat and poultry processors to adopt measures to minimize the risk of L. monocytogenes contamination of the post-lethality exposed products. The processors can adopt one of the three alternatives: (1) a post-lethality treatment and the addition of antimicrobial agents or process; (2) either a post-lethality treatment or the addition of antimicrobial agents to inhibit the growth of L. monocytogenes or (3) rely on sanitation to minimize L. monocytogenes contamination of the products. The objective of this study was to develop and validate predictive models for the growth of L. monocytogenes on RTE cured and non-cured roast beef and turkey hams during refrigerated storage. Cured and non-cured RTE roast beef and turkey hams were inoculated with a 5-strain L. monocytogenes cocktail. Isothermal growth at 1, 4, 7, 10, 13 and 16°C was used to develop primary and secondary models for L. monocytogenes growth in cured and non-cured roast beef and turkey hams formulated with a 0, 1, 2, and 3% blend of buffered sodium citrate and sodium diacetate (BSC+SDA). The products were vacuum-packaged and stored at 1, 4, 7, 10, 13 or 16°C. L. monocytogenes populations were determined at weekly intervals (10, 13, and 16°C), or biweekly intervals (1, 4, and 7°C) over a 12 or 16-week storage period. Dynamic models were developed by numerically integrating the primary model (Baranyi equation) and a cubical spline function to interpolate the growth rate at various temperatures. A total of 14 dynamic models (six for the cured and non-cured RTE roast beef and eight for the cured and non-cured RTE turkey hams) were developed. Each dynamic model was validated using a non-isothermal, sinusoidal temperature profile for both RTE roast beef and turkey hams for a period of 90 days. The dynamic model under-predicted the growth data. To make the models fail-safe, the initial physiological state of the cells (ho) value was reduced by 25%. The model was then used to predict the growth data at isothermal conditions and fail-safe predictions of L. monocytogenes growth were obtained. Accuracy factor (AF), bias factor (BF), and RMSE were calculated and evaluated between observed data and predicted values for the sinusoidal temperature profiles. The models for cured and non-cured RTE roast beef and turkey hams were capable of predicting growth with low RMSE values, and BF and AF values close to 1, indicating a good fit. These models will allow regulatory agencies as well as RTE processors to evaluate and validate their processing operations to control L. monocytogenes growth on RTE meat products and poultry products. Keywords. Listeria monocytogenes, RTE, predictive microbiology, BSC+SDA, cured meats, non-cured meats.