Development of a clinical scoring system for bovine respiratory disease in weaned dairy calves

Abstract

Clinical scoring systems for bovine respiratory disease (BRD) in weaned dairy calves have been developed in the past with calves experimentally infected with specific respiratory pathogens. In this prevalent case control study, a BRD clinical scoring system for weaned calves was developed using field data from 689 dairy calves housed in group pens on 5 dairies in California. Of the 689 calves in the study, 89 were selected because they appeared sick based on the display of lethargy, depression, or separation from the group, whereas the remaining 600 were randomly selected. Clinical signs were recorded for all calves, and BRD case status was determined by thoracic auscultation and ultrasound examinations, which were interpreted in parallel. Of the 689 calves, 238 were identified as BRD cases. Five survey-adjusted generalized linear mixed models with a logit link function, calf as the unit of analysis, and dairy as a random intercept were assessed using 3-fold cross-validation. The best model chosen based on performance and parsimony contained the variables cough (2 points), abnormal respiration (1 point), low body condition (5 points), sunken eyes (4 points), and a 24-h ambient temperature range >15°C (1 point) with a 2-point cutoff for a BRD suspect score. An alternative model did not contain a score for the covariate 24-h ambient temperature range and had a 1-point cutoff. The best model was tested on 174 observations not used for model training and resulted in 77.0% screening sensitivity, 100% diagnostic sensitivity, and 61.9% specificity. Adding rectal temperature ≥39.2°C (102.5°F) as a second-tier test increased specificity to 76.7% and lowered the screening sensitivity to 64.8% and diagnostic sensitivity to 76.9%. The alternative model had a screening sensitivity of 84.2%, diagnostic sensitivity of 100%, and specificity of 45.7%. Adding rectal temperature ≥39.2°C (102.5°F) as a second-tier test for score-positive animals improved specificity of the alternative model to 62.6% while lowering its screening sensitivity to 70.5% and diagnostic sensitivity to 76.9%. Use of a 2-tier California BRD postweaning scoring system may provide producers and veterinarians with a new tool to monitor BRD in group-housed dairy calves. Furthermore, the scoring system may aid in judicious medical intervention for BRD cases and reduce unnecessary treatments of animals with antimicrobials.