Bayesian modelling to identify the risk factors for Yersinia enterocolitica contamination of pork carcasses and pluck sets in slaughterhouses.

ABSTRACT

The probability of contamination by pathogenic Yersinia enterocolitica of carcasses and pluck sets at slaughterhouse was determined by means of a Bayesian analysis. Prior information of the prevalence of Y. enterocolitica in faeces and the seroprevalence of Yersinia in serum of pigs collected at farms were obtained from previous studies and introduced in the models as beta prior informative distributions. Samples of intestinal content, tonsils, and swabs of carcasses and pluck set surfaces were collected at slaughterhouses. The posterior probabilities, odds ratio (OR) and their probability interval (PI) were calculated by means of a generalized linear model constructed in WinBugs. Occurrence of Y. enterocolitica in intestinal content (OR 35.6, 95%PI 2.8-8285), tonsils (OR 38.4, 95%PI 5.0-854), and pluck set (OR 16.6, 95%PI 1.9-1111) was a risk for the contamination of pork carcasses, and an increased risk of contaminated pluck set was observed when Y. enterocolitica was isolated in intestinal content (OR 40.6, 95%PI 2.1-10510) and tonsils (OR 17.6, 95%PI 3.4-230.6). This increased risk indicated a potential cross-contamination at the slaughterhouse.