A probability model describing the interface between survival and death of Escherichia coli O157:H7 in a mayonnaise model system Original Research Article

A probability model was developed to describe the influence of temperature (10–30°C) and factors characterizing the composition of mayonnaise (salt [0·5–16·5%], pH [3·5–6·0], acetic acid [0–4%] and sucrose [0–8%]) on the survival of Escherichia coli O157:H7 in tryptic soy broth. pH was adjusted independently by the use of HCl or NaOH. Logistic polynomial regression with a total of 1820 factor combinations was used to describe the influence of main, quadratic, and cross-product effects of the environmental factors. The model successfully predicted survival or death in 1772 (97·4%) of the samples; of those incorrectly predicted, 28 were false-positive (survival predicted where death occurred) and 20 were false-negative (death predicted where survival occurred). The concordance index was 99·4% and the disconcordance index was 0·6%, indicating a good fit of the model to the observed data. The model was validated using survival data from experimental mayonnaise inoculated with E. coli O157:H7. Of 30 combinations of acetic acid, salt and sucrose tested, survival/death was incorrectly predicted by the model in only six cases and, of these, only one prediction was not ‘fail-safe’. This prediction was attributed to the influence of quadratic effects in the model, and to attempts to make predictions outside the boundary of experimental conditions. The results suggest that the probability model developed here can be usefully applied to predictions of E. coli O157:H7 survival in mayonnaise.