Research Summary - 3
Comparing a novel culture method using AI and digital image diagnostics to traditional somatic cell counts for detecting intramammary infections and making dry-cow treatment decisions.
| Date/Time: |
9/14/2024
08:45
|
| Author: |
Ben Davidson |
| Clinic: |
Dairysmart NZ Ltd |
| City, State, ZIP: |
Rangiora, NZ 7471 |
W.A. Mason, B.V.Sc, PhD
1
;
E.L. Cuttance, B.V.Sc, PhD
2
;
R.A. Laven, BVetMed, PhD
3
;
R. Nortje, M.V.Sc
4
;
B.I. Davidson, B.Ag.Sc, PG. Dip, M.V.Sc
5
;
1Epivets, Te Awamutu, 3800, New Zealand
2Epivets, Te Awamutu, 3800, New Zealand
3Massey University, Palmerston North,
4Rangiora Vet Centre, Rangiora, 7471, New Zealand
5Dairysmart NZ Ltd, Ranigora, 7471, New Zealand
Introduction:
In New Zealand, cow records, typically Somatic Cell Count (SCC) and clinical mastitis (CM) data, have been used to predict which cows are most likely to have an IMI at drying off, and thus receive antibiotic DCT. Culture and algorithm guided SDCT have been compared to each other and/or blanket DCAT as methods to reduce AMU at dry off without compromising cow udder health and productivity (Rowe et al. 2020; Kabera et al. 2020; tho Seeth et al. 2017). To date in New Zealand, SDCT guided by culture have predominantly only been used in trial settings due to perceived financial constraints and logistical challenges surrounding the implementation of culture based SDCT in seasonal dairy systems.
A new novel culture method using digital imaging and pattern recognition software to diagnose milk cultures on unique dual agar plates may be the answer to large scale testing events. Being able to more accurately predict infection status with a major pathogen this may be a new tool for guiding better SDCT treatment decisions than the current alternative.
The objective of this study was to compare a novel rapid culture-based protocol where only cows identified as having an IMI due to major pathogens (caused by Staph aureus, Streptococcus agalactiae, Escherichia coli, Klebsiella spp., Mycoplasma spp., Strep uberis, or Strep dysgalactiae) were compared with the current New Zealand industry standard of a SCC and mastitis-based selective dry cow algorithm.
The key outcomes were comparing the sensitivity and specificity of the two dry-off protocols at identifying major IMI from all enrolled animals and comparing individual cow SCC at the first herd test after calving from animals enrolled into one of two protocols.
Materials and methods:
A total of 1541 healthy multiparous pregnant lactating cattle from three 100% spring-calving farms were enrolled in this study. Between 10-14 days prior to dry-off, a composite 4-quarter sample was collected prior to milking. Samples were split at the laboratory into two;, one for conventional culture method and the other for a novel culture method utlilising a custom-made quartered agar plate designed to be rapidly read by a cloud based interpretation, powered by machine learning software. All enrolled animals had a status for mastitis IMI caused by a major pathogen by conventional culture, novel culture protocol (cult-SDCT) and the SCC and mastitis history protocol (alg-SDCT). Alg-SDCT were considered positive for a major pathogen if SCC>150,00 cells/ml at last herd test within 80 days of dry-off or which had an electronic record of clinical mastitis in the current lactation. The sensitivity and specificity of cult-SDCT and alg-SDCT, respectively, were compared against conventional culture results. Animals were then randomized to either cult-SDCT or alg-SDCT group blocked by conventional culture result (major, minor or no growth), where provision of selective DCAT were allocated within protocol group. A total of 776 and 765 cows were enrolled into the cult-SDCT group and alg-SDCT group respectively. Cows allocated to cult-SDCT that had either a major pathogen or a contaminated cult-SDCT result received DCAT. Within the alg-SDCT group, cows defined as positive as above received DCAT.
Results:
Across all enrolled animals, the sensitivity (0.80 vs 0.67) and specificity (0.91 vs 0.81) for major IMI prediction was greater for the cult-SDCT method than the alg-SDCT. After accounting for farm, age an dry-off SCC, compared to animals within the cult-SDCT group, animals within the alg-SDCT group had a SCC that was 1.14 times (95% CI 0.99, 1.32) higher at the post-calving herd test.
Compared to a standard algorithm-based protocol using SCC and CM, a novel culture system identified a higher proportion of major pathogens identified by conventional culture, thereby reducing antibiotic use (25 vs 23% of cows treated with DCAT) without increasing post-calving SCC (estimated marginal mean 129000 vs 113000cells, respectively).
Not treating minor pathogens (CNS) with DCAT had no negative consequences in this study. This method of diagnosis and detection of bacteria in the udder could result in a significant reduction in the quantity of antibiotics used globally in the dairy industry.
Significance:
Every year millions of tubes of dry-cow antibiotics are administered to cows unnecessarily. Overuse of antibiotics contributes to antimicrobial resistance (AMR) in humans and animals and prophylactic use of antibiotics is becoming less acceptable. New diagnostic tools such as this novel culture method with machine learning plate reading is a first-of-its-kind technology, offering AI and digital microbiology image transfer in a cloud based platform, giving efficient and accurate results which are more accurate than using current methods of SCC. This technology has the potential to significantly reduce the prescribing and use of mastitis antibiotics in the global dairy industry.
