Risk-based surveillance of antimicrobial residues in pigs--identification of potential risk indicators

Around 20,000 samples are analysed each year for the presence of antibacterial residues in Danish finisher pigs, and between zero and five samples are detected positive above the maximum residue level (MRL). The intention was to develop a risk-based surveillance programme involving fewer samples while ensuring equal sensitivity. Therefore, risk indicators were searched for. Data were obtained from the Danish slaughterhouse database covering the period from July 2010 to December 2012. Residues were found or suspected in 17 incidents. In nine of these, the farmer had called in to prevent the pigs from being slaughtered. Hence, eight suspect cases were found through the surveillance programme, and two of these were above MRL. For these eight case herds, the number of pigs slaughtered and the number in which each of the following lesions were found were included in a statistical analysis: chronic pleuritis, tail bite, chronic pericarditis, chronic pneumonia, chronic peritonitis, osteomyelitis, abscess in hindquarters, abscess in leg/toe and abscess in forequarters. Only chronic pleuritis was associated with the presence of residues. Next, data from all herds delivering pigs for slaughter to the same abattoir were included covering a 3-month period prior to the residue finding. The prevalence of chronic pleuritis was on average 1.7 times higher in the eight case herds compared to all other herds. In two herds, the prevalence was significantly higher (P≤0.05), and in one herd substantially higher, but only borderline significant (P=0.1). In the remaining herds, the prevalence did not differ from the other herds delivering pigs to the abattoir. This indicates that chronic pleuritis might be considered as a risk indicator for use in surveillance. Other risk indicators/factors - reported in the cases where the farmers called in - were inadequate marking of treated animals and incorrect use of medication dispensers. These factors are not suited for use in surveillance and should be dealt with otherwise.