Research Summary - 2
Diagnostic sensitivity and specificity of common bovine respiratory disease sampling methods in pre-weaned dairy calves using a Bayesian latent class analysis
| Date/Time: |
8/28/2026
16:15
|
| Author: |
Alexis Thompson |
| Clinic: |
Texas A&M Veterinary Medical Diagnostic Laboratory |
| City, State, ZIP: |
Canyon, TX 79016 |
A.C. Thompson, DVM, PhD, DACVPM (Epidemiology)
1
;
M.A. Scott, DVM, PhD
2
;
A.R. Woolums, DVM, MVSc, PhD, DACVIM
3
;
B. C. Credille, DVM, PhD, DACVIM
4
;
B. M. Crossley, DVM, MPVM, PhD
5
;
T. W. Lehenbauer, DVM, MPVM, PhD, DACVPM
6
;
S. S. Aly, BVSc, MPVM, PhD
6
;
1Texas A&M Veterinary Medical Diagnostic Laboratory, Canyon, TX
2Texas A&M University College of Veterinary Medicine and Biomedical Sciences, Canyon, TX
3Department of Pathobiology and Population Medicine, College of Veterinary Medicine, Mississippi State University, MS
4Boehringer Ingelheim Animal Health, Duluth, GA
5California Animal Health and Food Safety Laboratory System, University of California-Davis, Davis, CA
6Department of Population Health and Reproduction, University of California, Davis School of Veterinary Medicine, Davis, CA
Introduction:
Calculate the diagnostic sensitivity (Se) and specificity (Sp) of the bronchoalveolar lavage (BAL), nasal swab (NS), deep nasopharyngeal swab (DNP), and transtracheal wash (TTW) to diagnose Mannheimia haemolytica, Pasteurella multocida, Mycoplasma bovis, bovine respiratory syncytial virus (BRSV), and bovine coronavirus (BCoV) in pre-weaned dairy calves with signs of respiratory disease on a commercial operation using a Bayesian latent class model.
Materials and methods:
A retrospective cohort study using an existing dataset investigating sampling methods and bovine respiratory disease (BRD) pathogen detection was used. The target population was pre-weaned dairy calves typical of North American commercial dairy operations. The sampled population were pre-weaned dairy calves from a Californian calf ranch. Included calves had to exhibit clinical signs of BRD; calves that received intranasal modified live respiratory vaccines within the previous 30 days and calves that were ever treated with antimicrobials for BRD were excluded. At sampling, the nostrils were cleaned with a single use paper towel then each calf was sampled by NS, DNP, TTW, and BAL, in that order. Bacterial culture was used to identify Mannheimia haemolytica, Pasteurella multocida, and Mycoplasma sp. Mycoplasma sp. isolates were tested by PCR to identify the species Mycoplasma bovis. Multiplex RT-PCR was used to identify bovine herpesvirus-1 (BHV-1), bovine viral diarrhea virus type 1 or 2 (BVDV1 and BVDV2), BRSV, and BCoV. A conditionally dependent latent class model was fit within a Bayesian framework to estimate antemortem BRD sampling method Se and Sp. Priors were found in published literature. Posterior inferences (median [95% credible intervals]) were estimated using a Gibbons sampler.
Results:
Samples from 100 calves were available for inclusion. Fourteen calves (14%; 14/100), 56 calves (56%; 56/100), 71 calves (84%; 71/85), 36 calves (46%; 36/79), 5 calves (5.9%; 5/85), and 6 calves (7.6%; 6/79) tested positive using all sampling methods for Mannheimia haemolytica, Pasteurella multocida, Mycoplasma sp., Mycoplasma bovis, BRSV, and BCoV, respectively. No samples were positive for BHV-1, BVDV1, or BVDV2. Eighty calves (80%; 80/100), 38 calves (38%; 38/100), 0 calves (0%; 0/85), 31 calves (39%; 31/79), 76 calves (89%; 76/85), and 67 calves (85%; 67/79) tested negative using all sampling methods for Mannheimia haemolytica, Pasteurella multocida, Mycoplasma sp., Mycoplasma bovis, BRSV, and BCoV, respectively. Sampling methods performed similarly across all pathogens tested for. Of the sampling methods, NS consistently had the lowest Se (62%[25-91%]) whereas BAL and DNP performed similarly (81%[55-96%], 84%[55-98%], respectively) and TTW had the highest Se (87%[66-97%]). For Sp, TTW was the lowest (50%[40-60%]) and BAL, DNP, and NS performed similarly (94%[89-98%], 92%[77-99%], 87%[66-97%], respectively).
Significance:
The limited number of concordant positive results resulted in wide credible intervals around the calculated sensitivites. Of the sampling methods evaluated, DNP, which is technically simple and convenient, had similar diagnostic Se and Sp as compared to more invasive and technically difficult TTW and BAL for detecting BRD pathogens in pre-weaned dairy calves with signs consistent with BRD.
