Predictive modeling and scenario-based risk estimation of Escherichia coli growth in meatball products: A comparison of aerobic and vacuum packaging conditions

This study evaluated the growth of Escherichia coli , a hygiene indicator with a high pathogenic potential, in meatballs under aerobic and vacuum packaging (AP and VP) conditions using the Baranyi model. The model was validated using various wild-type E. coli strains isolated from different food sources, including both commensal and pathogenic strains. E. coli showed faster growth under VP conditions, with significantly longer lag phases and higher maximum growth rates compared to AP. The coefficients of determination (R 2 ) of the secondary models for AP and VP were 0.986 and 0.987. The bias factor and accuracy factor for validation were 0.942 and 1.062 for AP and 1.102 and 1.102 for VP, respectively. All validation values were within the acceptable range for various E. coli strains, including acid-resistant, multidrug-resistant, virulent gene, and biofilm-forming strains, supporting the robustness of the model. The dynamic model exhibited a root mean square error below 0.855 ln CFU/g and acceptable simulation zone values above 86%. Scenario-based risk estimation indicated higher risk from consumption of VP meatballs than AP, emphasizing the potential hazards of VP foods. These findings highlight the need for improved food safety management to address the risk of E. coli in VP foods. ● E. coli growth in meatball was modeled under aerobic (AP) and vacuum packaging (VP). ● Lag phase duration and maximum growth rate were higher under VP than AP. ● Predictive models could be applied to pathogenic and commensal E. coli strains. ● Predictive models performed well under various dynamic temperature profiles. ● Scenario-based risk estimation shows higher E. coli risk in VP meatballs.