Evaluating the Predictive Ability of a Path-Dependent Thermal Inactivation Model for Salmonella Subjected to Prior Sublethal Heating in Ground Turkey, Beef, and Pork

Abstract

Pathogen thermal inactivation models currently available to and used by industry consider only the present state of the product when predicting inactivation rates. However, bacteria subjected to sublethal thermal injury can develop partial protection against lethal temperatures. The objective of this study was to extend the capabilities of a previously published path-dependent Salmonella inactivation model by accounting for longer sublethal heating periods and different substrates and to test this new model against independent data. Ground samples of irradiated (> 10 kGy) turkey breast, beef round, and pork loin were inoculated with an eight-serovar Salmonella cocktail and subjected to 53 nonisothermal treatments (in triplicate) that combined a linear heating rate (1, 2, 3, 4, or 7 K/min), a variable length sublethal holding period (at 40, 45, or 50°C), a lethal holding temperature (55, 58, 61, or 64°C), and a nominal target kill (3- or 5-log reductions) (n = 159 for each meat species). When validated against nonisothermal data from similar treatments, traditional state-dependent model predictions resulted in root mean squared errors (RMSEs) of 2.9, 2.2, and 4.6 log CFU/g for turkey, beef, and pork, respectively. RMSEs for the new path-dependent model were 0.90, 0.81, and 0.82 log CFU/g for the same species, respectively, with reductions in error of 63 to 82 % relative to the state-dependent model. This new path-dependent model can significantly reduce error from the state-dependent model and could become a useful tool for assuring product safety, particularly relative to slow heating processes.