Comparison of a Remote Early Disease Identification (REDI) System to Visual Observations to Identify Cattle with Bovine Respiratory Diseases

ABSTRACT Background Bovine respiratory disease (BRD) is the most common disease in the beef industry, and diagnosis based on visual observation of clinical signs is challenging. Cattle may alter behavior when diseased and technology exists to remotely monitor these behavioral changes. The objective of this study was a preliminary assessment of the ability of a remote early disease identification system (REDI) to identify cases of BRD compared to a trained observer. Methods Eighty cross breed bull calves (239 ± 19 kg) were used to evaluate potential differences in the two BRD diagnosis modalities. Bulls were randomly assigned to one diagnostic modality, and were then treated for respiratory disease only when that modality called them sick and their rectal temperature was above 40° C. Agreement and accuracy between the two diagnostic modalities were compared. Results Within each treatment group, the numbers of pulls were nearly identical (REDI n = 17/40, visual observer n = 16/40), and the average days to first treatment were also similar. Both systems agreed on 83% of all daily individual observations, and this agreement is deemed as fair when evaluated using a kappa statistic on both a calf (κ = 0.26) and individual observation level (κ = 0.23). When using agreement of both systems and the temperature threshold to estimate true calf health status, there were no statistical differences between the two modalities and both REDI and visual observer methods illustrated a high probability of making the correct disease call, 94% ± 1.4 and 92% ± 1.8 respectively. This research tested the initial iteration of the system and further refinements to REDI algorithms evaluated post-hoc improved agreement to (κ = 0.30), and these algorithms would have identified sick cattle an average of 0.75 days before Comparison of a Remote Early Disease Identification (REDI) System to Visual Observations to Identify Cattle with Bovine Respiratory Diseases