Evaluation of an automated on-farm device to identify groups of bacteria associated with clinical mastitis on a United States dairy farm

Recent studies suggest that selectively treating non-severe clinical mastitis during lactation, guided by on-farm culture, can be prudent. However, on-farm culture presents logistical challenges. This study investigated Mastatest’s® (MT; Masta­plex, New Zealand) ability to identify broad groups of patho­gens in clinical mastitis milk samples compared to laboratory culture and matrix-assisted laser desorption ionization-time of flight mass spectrometry (LC+M) on one U.S. dairy farm. Milk samples from quarters with clinical mastitis were col­lected on a dairy farm located in Colorado (n = 162). The test’s ability to identify groups of pathogens was assessed against LC+M (reference method) using contingency tables. Test char­acteristics (sensitivity [Se], specificity [Sp], positive predictive value and negative predictive value) and their 95% confidence intervals were estimated. Agreement was assessed using the Kappa statistic. Prevalences (0.20-0.60) for Gram-positive and Gram-negative bacteria were used to estimate predictive values. Before analysis, contaminated or missing samples in either MT or LC+M were excluded (n = 36). Our findings indi­cated moderate to substantial agreement, with Kappa values ranging from 0.53 to 0.66. For Gram-positive pathogens, Se and Sp estimates, and their 95% confidence intervals were: Se: 0.83 (0.70, 0.92), Sp: 0.81 (0.70, 0.89), and for Gram-negative bac­teria (Se: 0.73 [0.50, 0.89], Sp: 0.93 [0.87, 0.97]). Negative predic­tive value remained above 0.70 across different prevalence sce­narios, particularly where values exceeded 0.90 for prevalences below 0.30. Our findings suggest MT is a suitable technology for on-farm detection of mastitis pathogens commonly identified on U.S. dairy farms in clinical mastitis milk samples.