Research Summary - 2
Evaluating the diagnostic utility of on-farm culture for mastitis pathogen group identification on commercial dairy farms
| Date/Time: |
9/12/2025
13:45
|
| Author: |
Felipe Peņa-Mosca |
| Clinic: |
Cornell University |
| City, State, ZIP: |
Ithaca, NY 14850 |
Felipe Peņa-Mosca, DVM, MS, PhD
1
;
Paulina Avondet-Benzano, Student
1
;
Jocelyn Rivera, DVM
1
;
Lowell Midla, VMD, MS
2
;
Tiago Tomazi, DVM, PhD
2
;
Cainan Florentino, DVM
3
;
Luciano Caixeta, DVM, PhD
3
;
Daryl Nydam, DVM, PhD
1
;
1Department of Public and Ecosystem Health, College of Veterinary Medicine, Cornell University, Ithaca, NY, 14853.
2Merck Animal Health, Kenilworth, NJ, 07033.
3Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, 55108.
Introduction:
Recent research has demonstrated that on-farm culture (OFC) can identify mastitis pathogens in milk samples and guide selective treatment of clinical mastitis. However, most studies evaluated OFC performance under controlled conditions, typically conducted by research personnel at a single farm, which may not reflect its diagnostic performance when performed on commercial dairy farms. The objective of this study was to evaluate the diagnostic performance of OFC compared to laboratory methods for identifying mastitis pathogens in milk samples collected from cows with clinical mastitis in four commercial dairy farms, and to explore whether diagnostic performance varied across farms.
Materials and methods:
This longitudinal study was conducted on four commercial dairy farms in New York State between August 2024 and March 2025. Aseptic milk samples were collected from cows with clinical mastitis identified by milking technicians and initially processed using OFC by farm personnel, using the Minnesota Tri-plate or Accumast plate systems. Samples were transported on ice by the research team and submitted to the Quality Milk Production Services laboratory (Ithaca, NY), where they were incubated aerobically at 37°C and read at 18–24 and 48 hours. Samples with three or more distinct colony types were classified as contaminated and excluded from further analysis (n = 85). Isolates from non-contaminated samples were identified using MALDI-TOF MS. Statistical analyses were performed using R (version 4.3.2) to estimate kappa, sensitivity and specificity for pathogen detection within each farm, using laboratory identification as the reference.
Results:
A total of 2,170 milk samples were collected among four farms. Gram-positive pathogens were identified in 31.8% of the milk samples (n = 691; range: 21.8 – 44.9%) using laboratory methods and in 49.1% (n = 1,066; range: 37.6 – 64.8%) by OFC. Across farms, sensitivity and specificity for Gram-positive pathogen detection ranged from 0.71 to 0.94 and 0.49 to 0.79, respectively. Agreement between methods was fair to moderate (kappa: 0.26 – 0.58).
The most frequently isolated Gram-positive pathogens were Streptococcus spp. and Streptococcus-like organisms (SSLO), identified in 25.3% (n = 549) of samples by laboratory methods and 45.6% (n = 990) by OFC. For SSLO detection, sensitivity ranged from 0.76 to 0.98, and specificity ranged from 0.52 to 0.77 across farms. Agreement varied from fair to moderate (kappa: 0.27 – 0.54).
Gram-negative pathogens were identified in 11.2% (n = 242) of the samples by laboratory culture and in 12.5% (n = 272) by OFC, with farm-level ranges of 6.7 – 16.7% and 4.4 – 23.5%, respectively. Detected Gram-negative pathogens included Escherichia coli (7.3% [n = 158]; range: 3.7 – 11.2% across farms) and Klebsiella spp. (0.6% [n = 14]; range: 0 – 1.8%). Diagnostic performance for Gram-negative pathogens varied across farms, with sensitivity ranging from 0.52 to 0.82, specificity ranging from 0.82 to 0.99, and moderate to substantial agreement between methods (kappa: 0.50 – 0.71).
Significance:
On-farm culture showed good sensitivity and specificity for identifying groups of mastitis pathogens in milk samples collected from four commercial dairy farms. Agreement between OFC and laboratory culture varied across pathogen groups and farms, with higher agreement for Gram-negative compared to Gram-positive pathogens, including SSLO. Sensitivity for Gram-positive pathogens and SSLO was high, while specificity was relatively lower, suggesting that farmers prioritized avoiding missed treatments, which might be beneficial for these pathogens. Variability in OFC diagnostic performance across farms emphasizes the need to consider farm-level factors and to standardize protocols and training when implementing OFC. Although OFC is a practical tool for guiding on-farm treatment decisions, systematic laboratory submissions remain important for quality control.
