Dynamic analysis of growth of Salmonella spp. in raw ground beef – Estimation of kinetic parameters, sensitivity analysis, and Markov Chain Monte Carlo simulation

Abstract Nontyphoidal Salmonella are major foodborne pathogens often associated with raw and undercooked meats. In this study, a cocktail containing 6 strains, 5 serovars of Salmonella enterica was inoculated to raw ground beef to observe the bacterial growth and survival under 6 dynamic temperature profiles ranging from 1 to 45 °C during storage. One-step dynamic analysis was used to directly construct predictive models from dynamic growth curves for both background microbiota and Salmonella. The estimated minimum growth temperature (Tmin) for Salmonella spp. in raw ground beef was 9.9 °C, and its population would gradually decrease at the rate of approximately 0.14 log CFU/g per week per °C away from this temperature. Above 17.3 °C, Salmonella would grow faster than the background microbiota. Scaled sensitivity coefficients were calculated to determine the identifiability of the kinetic parameters. The models were validated using dynamic growth curves set aside for validation. The root-mean-square-errors (RMSE) of prediction were identical to those of model development (0.6 log CFU/g) for the deterministic method. Bayesian Markov Chain Monte Carlo (MCMC) simulation was used to generate the posterior distribution of kinetic parameters and to predict the bacterial growth. The results showed that the observed growth data were mostly within the range of mean ± standard deviation of the stochastic predictions. The RMSE of prediction by MCMC analysis was only 0.2 log CFU/g, showing that Bayesian MCMC simulation can be used to more accurately predict the growth of Salmonella in raw ground beef. The results from this study may help conduct more accurate risk assessments of Salmonella spp. in raw beef during storage and temperature abuse.