Purpose: This research aims to determine the critical control points (CCPs) for L. monocytogenes contamination along the production chain in the Scottish cold-smoked salmon industry and to estimate the probability of illness in consumers.
Methods: An Excel spreadsheet model was built with variables accounting for: processing and storage conditions along the food chain, prevalence, demographics and consumption data. Monte Carlo simulations and sensitivity analysis were performed using @Risk. The probability of illness per serving was estimated through an exponential dose-response model.
Results: The prevalence of Listeria in cold-smoked salmon was most sensitive to cross-contamination during transport to smoking plant (Spearman Rank coefficient 0.87), contamination during slicing (0.87), filleting (0.86), and salting (0.84). The dose-response model was most sensitive to Listeria prevalence in the final product (0.78) and Listeria contamination levels (0.50). The model used a species specific constant R (1.18×10-10) that helps define the dose-response curve. The estimated mean risk of illness per serving was 6.41×10-06. Assuming 100% virulence of L. monocytogenesstrains, the average number of predicted listeriosis cases in Scotland was 49, whereas the reported annual average is 17.
Significance: This farm-to-fork risk assessment of L. monocytogenes in cold-smoked salmon suggests that CCPs are transport to smoking plant, slicing, filleting and salting. Appropriate controls in these stages would improve food safety through reduction of Listeria prevalence and levels in cold-smoked salmon.