Validation of a rapid on-farm culture system for group classification of clinical mastitis-causing pathogens

The objective of this diagnostic accuracy study was to compare the use of a rapid tube test system (MastDecide [MD]; Quidee GmbH, Homberg, Germany) to aerobic milk culture for group classification of mastitis pathogens in dairy cattle. A total of 204 milk samples from cows with clinical mastitis (CM) were collected across 60 dairy herds in Ontario, Canada. Samples were collected by dairy producers and their staff and transported refrigerated to the University of Guelph Animal Health Laboratory (Guelph, ON, Canada) on the day of collection. Each sample was tested via MD (gram-positive, gram-negative, or "no growth") and via aerobic milk culture followed by MALDI-TOF (reference method [RM]). The MD and RM results were interpreted at 14 and 24 h after incubation, respectively. An additional assessment was performed regarding the accuracy of intramammary antibiotic treatment decisions (gram-positive: to treat; gram-negative and "no growth" samples: not to treat). Test performance characteristics (overall accuracy, sensitivity [Se], specificity [Sp], and positive [PPV] and negative predictive value [NPV]) as well as Cohen's kappa coefficient (κ) were calculated. The RM results were classified as gram-positive (n = 107, 53%), gram-negative (n = 20, 10%), "no growth" (n = 51, 25%), other pathogens (n = 21, 10%), and mixed growth (gram-positive and gram-negative; n = 5, 3%). The Se of the MD test was 58% (95% CI: 47%-67%), 40% (19%-64%), and 61% (50%-72%) for gram-positive, gram-negative, and "no growth," respectively. The Sp of the MD test was 73% (63%-81%), 84% (78%-89%), and 75% (66%-82%) for gram-positive, gram-negative, and "no growth," respectively. The PPV of the MD test was 68% (57%-78%), 22% (10%-39%), and 62% (50%-73%) for gram-positive, gram-negative, and "no growth," respectively; and the NPV was 63% (54%-72%), 93% (87%-96%), and 74% (65%-82%) for gram-positive, gram-negative, and "no growth," respectively. The overall test accuracy for Gram stain classification analysis was 57%, with a κ of 0.30. Regarding intramammary antibiotic treatment, the MD test yielded a correct decision relative to the RM for 65% of the cases. Using this on-farm test as a primary guide in CM treatment decisions should be approached with caution, as the risk of false negatives may affect the ability to effectively treat CM cases.