Jump to Main Content
Improved estimation of thermal resistance of Escherichia coli O157:H7, Salmonella spp., and Listeria monocytogenes in meat and poultry – The effect of temperature and fat and A global analysis
- Huang, Lihan, Hwang, Cheng-An, Fang, Ting
- Food control 2019 v.96 pp. 29-38
- Escherichia coli O157, Listeria monocytogenes, Salmonella, acidulants, anti-infective agents, beef, food pathogens, health hazards, heat tolerance, human diseases, ingredients, microorganisms, model validation, models, poultry, poultry meat, prediction, public health, regression analysis, serotypes, temperature
- Escherichia coli O157: H7, Salmonella spp., and Listeria monocytogenes are three major foodborne pathogens in meats that frequently cause serious human infections and are significant public health hazards. Many studies have reported the thermal resistance of these pathogens in various meats. Although showing some general trends, the published data vary considerably.This study aimed to develop more accurate regression models for estimating the thermal resistance by incorporating the effect of temperature and fat. Both reduced model (temperature only, or D-z model) and expanded model (temperature and fat) were developed by linear regression. The results showed that the expanded models improved the accuracy of estimation of log D. For E. coli O157:H7 in beef, greater than 93% of the variations in log D can be attributed to the expanded model, while it is greater than 96% for E. coli O157:H7 in non-beef meats, L. monocytogenes, and Salmonella spp. in poultry meats. For Salmonella spp. in non-poultry meats, 90.4% of the variations of the log D values can be attributed to the expanded model, which is significantly greater than 74% in the reduced model. For Salmonella spp. in poultry meats, both reduced and expanded models can be used, achieving greater than 94.8% accuracy in predicting the log D values.Based on Akaike Information Criterion (AIC) and F-test, the improvement in accuracy by each expanded model is statistically significant (α = 0.05), irrespective of the difference in the sources, serotypes, and isolates. The expanded models can be used to design and evaluate thermal processes for inactivating these pathogens in meat and poultry products with normal fat contents and without ingredients (salt, acidulants, or other antimicrobials) that may significantly affect the survival of these microorganisms. However, the users should know the limits of regression models and validate the models before using them in real-world applications.