Validation of On-Farm Bacteriological Systems for Endometritis Diagnosis in Postpartum Dairy Cows

Simple Summary Endometritis is a disease affecting reproductive performance in dairy cows. Considering the modern issues concerning the use of antibiotics in animal production, it is important to refine our criteria for diagnosing this disease. As such, confirming the presence of bacteria in the uterus before implementing an intrauterine antibiotic treatment is critical. To be able to achieve this on dairy farms, the accuracy of currently available on-farm bacteriological culture systems (Tri-plate and Petrifilm) needs to be validated. This study used data from 189 dairy cows to assess this objective. Uterine samples were collected on cows between 30 and 43 days in milk and were submitted for bacteriological culture using three different approaches: standard laboratory and two on-farm systems (Tri-plate and Petrifilm). Our results showed that the optimal criteria for using the Tri-plate and Petrifilm on-farm systems were >90 and >100 colonies, respectively, when compared with the results from the standard laboratory. These results support the possibility of using the Tri-plate on-farm bacteriological culture system to diagnose endometritis. The objective of this study was to validate the accuracy of the results of on-farm bacteriological culture media (Tri-plate and Petrifilm) from endometrial samples compared with the ones from the diagnostic laboratory. A cross-sectional observational study was set up within two dairy herd clients of the Universite de Montreal. A total of 189 cows in the postpartum period were systematically enrolled to collect two uterine samples from cytobrushes during the same examination. The first cytobrush was used to inoculate the Tri-plate medium directly and then was sent to the reference laboratory for aerobic bacterial culture. The second cytobrush was used to make a microscopic smear for cytological analysis (proportion of polymorphonuclear cells) and subsequently diluted in 1 mL of saline to inoculate the Petrifilm medium. From these data, statistical analyses were computed to optimize the summation of sensitivity and specificity of the two systems compared with the results of the reference laboratory. For the Tri-plate and Petrifilm media, the cutoffs of >90 and >100 colonies gave the maximum sum of sensitivity and specificity, respectively. In conclusion, Tri-plate media was best at reproducing the results obtained by laboratory analysis using a threshold of >90 colonies.