Determining the Likelihood that Salmonella Develops Heat Resistance during Thermal Processing of Commercial, Whole-Muscle, Ready-to-Eat Meat Products

This research adapted and validated a model to predict the rate of Salmonella thermal inactivation as a function of both product temperature and prior sublethal thermal history. It also evaluated whether any resulting increase in Salmonella thermal resistance would have an impact on the compliance of typical commercial cooking operations with USDA-FSIS lethality performance standards for RTE products.

Download final report

Objectives

The objectives of this research were to modifying a new thermal inactivation model that accounts for the effect of sub-lethal injury on subsequent heat resistance of Salmonella; validate the model via laboratory- and pilot-scale challenge studies, and; evaluate whether the resulting increase in Salmonella thermal resistance has practical impact on the compliance of typical commercial cooking operations with USDA-FSIS lethality performance standards for ready-to-eat (RTE) products.

Conclusions

Salmonella can develop significantly increased thermal resistance due to sub-lethal injury that can occur during slow cooking of whole-muscle meat and poultry products.  Traditional inactivation models (D and z) based on isothermal inactivation studies can significantly over-predict the actual lethality of Salmonella in slow-cooked meat and poultry products, with the degree of over-prediction increasing with the extent of sub-lethal heating.  The uncertainty underlying thermal process validations increases significantly when scaling predictions from laboratory- to pilot-scale (and presumably commercial-scale) applications.  Whole-muscle turkey and beef products cooked in a moist-air convection oven to a core temperature of 71.1°C (160°C) all exceeded the lethality performance standards.  There was a significant risk of not achieving the lethality performance standards for whole-muscle turkey and beef products cooked just to the target lethality (i.e., 7.0 or 6.5 log10 reductions, respectively), computed via traditional methods (D and z from laboratory studies).  Therefore, particular caution (and/or improved modeling methods) should be exercised for marginally-processed products.

Deliverable

These results can be translated with a concurrent funded USDA project into a web-based process lethality tool that will account for the effects of product species, structure, composition, and heating profile (including sub-lethal injury) in computing process lethality and reliable estimates of uncertainty.