Data from routine meat inspection is a poor indicator of the prevalence of tail lesions in undocked pigs

We investigated the prevalence of tail lesions in batches of undocked slaughter pigs in herds just before delivery to an abattoir. At the abattoir, dehaired and scalded carcasses were submitted to routine meat inspection which included recording of tail lesions. The purpose of the study was to investigate the relationship between clinically and abattoir- detected tail lesions in undocked pigs. During visits in 15 label-production herds, 2346 slaughter pigs from 24 batches were examined. Tail lesions were registered as mild healed, mild unhealed or severe. The median prevalence of the three categories in batches was 13, 9 and 6%, respectively. At the abattoir, tails were evaluated by public inspectors. Between 0 and 10% of pigs within batches (median: 1%) were registered with tail lesions at the abattoir. A linear regression model was used to compare the proportions of severe tail lesions registered in each batch within the herds with the proportions registered at the abattoir. We applied a leave-one-batch-out internal cross-validation on the model in order to explore a systematic relationship. The mean absolute difference between the predicted and the observed proportion was 9%-points. The coefficient of determination (r 2 ) was 0.006. Our results indicate that there is no systematic relationship between clinically and abattoir-registered tail lesions in undocked pigs. Thus, abattoir registrations as carried out in the present study did not mirror the clinical situation properly. If meat inspection recordings should be used to reflect tail lesions in the herds, efforts must be undertaken to ensure a positive correlation between the two. Thus, abattoir registrations used as an indicator of tail bite prevalence in herds are currently not reliable.