Antimicrobial use and farmers' attitude toward mastitis treatment on dairy farms with automatic or conventional milking systems

Mastitis is one of the major causes for antimicrobial use on dairy cattle farms. On farms with an automatic milking system (AMS), diagnostics differ from those with a conventional milking system (CMS), with potentially a different attitude toward mastitis treatment. This may result in differences in antimicrobial usage (AMU) between these 2 types of farms. The aims of this study were (1) to compare AMU between AMS and CMS farms, (2) to identify variables associated with AMU in both types of herds, and (3) to describe the distribution of mastitis-causing pathogens and their antimicrobial resistance patterns. Data on AMU was collected for 42 AMS and 254 CMS farms in the Netherlands and was expressed as animal-defined daily dose (ADDD). The ADDD variables were total usage (ADDDTOTAL), intramammary usage during lactation (ADDDIMM), usage for dry cow therapy (ADDDDCT), and usage by injection (ADDDINJ). Eighteen AMS farms and 24 CMS farms participated in a survey on factors potentially related to AMU. These farmers collected 5 quarter milk samples from quarters with clinical mastitis or high somatic cell count, which were subjected to bacteriological culture and antimicrobial susceptibility testing. In addition, routinely collected udder health data of these farms were used in the analysis. Nonlinear principal component analysis (NLPCA) was used to explore associations between AMU, udder health, and questionnaire variables. The ADDDTOTAL and ADDDDCT were comparable between AMS and CMS farms, whereas ADDDIMM tended to be lower and ADDDINJ higher on AMS farms than on CMS farms. The NLPCA yielded 3 principal components (PC) that explained 48% of the variation in all these variables. The AMS farms were not distinguished from CMS farms in the principal component space. The 3 PC represented different aspects of udder health, ADDDTOTAL, and treatment strategy. Differences in treatment strategy were unrelated to total antimicrobial usage or overall udder health. The distributi