Herd typologies based on multivariate analysis of biosecurity, productivity, antimicrobial and vaccine use data from Danish sow herds

The use of antimicrobials in livestock constitutes an increasing global concern, and many countries pursue approaches to reduce the amount used, particularly in the pig production industry. The EU Commission has decided, due to environmental concerns, to phase out use of zinc oxide in pigs by 2022. This poses an additional challenge to efforts to reduce antimicrobial use (AMU) in European pig production. The pig production sector needs further information about how to comply with official AMU requirements without losing competitiveness. The most efficient approaches are likely to involve a combination of multiple factors. Therefore, the objective of this study was to explore multidimensional associations between biosecurity, productivity, vaccination and AMU. A cross sectional study was conducted using data from 160 Danish sow herds in 2014-2015. Biosecurity data were collected through computer-assisted telephone interviews using a pre-developed questionnaire (Biocheck.UGent (R)) supplemented with additional country-specific questions and translated into Danish. Herd-specific data, consisting of antimicrobial prescriptions, purchase of vaccines against five endemic infections, herd health status and one productivity measure (i.e. number of weaned piglets per sow per year) were extracted from various databases. Factor analysis was conducted on a subset of the data from 152 herds with sufficiently complete data. The identified factors were explained by evaluating data from herds with extreme loadings on the respective factor. The results were further discussed based on plots combining herd factor loadings on two factors at a time. Four factors were selected based on the break-point in the scree-plot. Factor 1 included herd type, herd size, and age of farm buildings. Factor 2 covered general biosecurity, including several internal and external biosecurity measures. Factor 3 represented preventive measures implying specific focus on avoiding introduction of ASF by foreign employees. Lastly, Factor 4 covered vaccination status, specifically regarding vaccination against Mycoplasma hyopneumoniae and Porcine Reproductive and Respiratory Syndrome virus. These factors were used to group the 152 sow herds into herd typologies. Feasible strategies aimed at improving health by reducing AMU without hampering animal welfare were identified and discussed for each typology. AMU and productivity correlated only weakly with other variables. This is probably due to limited variability in both these variables in study herds, which might be attributed to official restrictions on AMU, and a general high level of health and biosecurity in Danish sow herds.