Mathematical modelling and validation of Salmonella enterica growth in sushi exposed to different time-temperature scenarios found in real sushi establishments.

In this study mathematical models to predict Salmonella enterica growth in sushi at different temperatures were developed considering data obtained in 26 restaurants in Southern Brazil. The sushi type chosen to develop the models was the one that presented the highest total aerobic mesophilic counts among sushis collected in the establishments. Salmonella was inoculated (2-3 log UFC/g) in this sushi type prepared in the laboratory (pH 4.8; aw 0.98) and incubated under isothermal conditions at 7, 15, 20, 25 and 37 °C. Baranyi and Roberts model was used to describe Salmonella growth curves, generating R values of ≥0.98 and RMSE values of <0.24 log CFU/g/h for primary models. Ratkowsky's equation was used in secondary model, generating R of 0.99 and RMSE of 0.02 log CFU/g/h. The model validation was simulated under non-isothermal conditions, using the worst-case scenario that was built through data from the environmental conditions and data obtained from the restaurants. The non-isothermal conditions were performed at 36.3 °C for 6 h, 10 °C for 24 h and 29.5 °C for 6 h sequentially, reaching 6.7 log CFU/g of Salmonella and generating RMSE of 0.06 log CFU/g/h, Bias factor of 0.97 and Accuracy factor of 1.03. The negligible growth time (ς) for Salmonella, considering the average of higher distribution temperatures of chosen sushi type (approximately 18 °C), was 8.9 h. However, growth rates of total aerobic mesophilic demonstrated that at 15 °C and 20 °C, the lag phases were approximately 11 and 5 h respectively. Based on these results, we suggest for sushi distribution the use of temperatures of ≤15 °C for 6 h (maximum time of distribution allowed in Brazil) considering the Salmonella growth.