Salmonella in slaughter pigs in Northern Ireland: prevalence and use of statistical modelling to investigate sample and abattoir effects.

A cross-sectional survey of pigs at slaughter in Northern Ireland was undertaken to determine the overall prevalence of Salmonella infection. In total 513 pigs were sampled across four abattoirs, with Salmonella spp. isolated from the caecal contents of 31.4% (95% confidence interval [CI] 27.4%-35.4%) and from 40.0% (95% CI 35.8%-44.3%) of swabs taken from the surface of carcasses post-evisceration. Two serovars, S. Typhimurium and S. Derby, were predominant and accounted for 52% and 35% respectively, of isolates from caecal contents. Antimicrobial resistance was most common amongst isolates of S. Typhimurium with 63.9% multiresistant compared to 10.8% of S. Derby isolates and 8.0% of other Salmonella spp. The proportion of pigs showing serological evidence of infection was significantly lower, with 11.5% (95% CI 8.9%-14.6%) and 10.1% (95% CI 7.7%-13.1%) of meat-juice samples giving positive and suspect reactions, respectively. The ratio of caecal positive to serologically positive animals is higher than in a number of other studies and may suggest recent infection, such as infection occurring during transport or lairage, in a proportion of animals. Statistical (logistic regression) modelling was used to investigate the association between the risk of Salmonella on carcasses and the isolation of Salmonella from caecal contents, and/or the serological status of the animal, while adjusting for other possible explanatory and confounding variables such as abattoir, season, day and time of sampling. The occurrence of Salmonella in caecal contents (odds ratio [OR] 2.39; 95% CI 1.52-3.77) or a suspect/positive serological reaction (OR 2.15; 95% CI 1.28-3.61) were both independently associated with the occurrence of Salmonella on carcasses in homebred, but interestingly not in imported animals. In most multivariable models there were also significant differences in carcass contamination between seasons with the highest odds of carcass contamination occurring in the April to June quarter and the lowest in the October to December quarter. Differences between sampling days were also evident with the highest odds of carcass contamination at the end of the week (Fridays) and the lowest at the start of the week (Mondays). These associations, after adjusting for the caecal or serological result, would suggest the occurrence of abattoir effects, such varying residual levels of abattoir contamination, which are independent of the individual pig status.