Inter-rater agreement and reliability of thoracic ultrasonographic findings in feedlot calves, with or without naturally occurring bronchopneumonia.

Background: Thoracic ultrasonography (TUS) can be used to assess the extent and severity of lung lesions associated with bronchopneumonia (BP) in feedlot cattle.
Hypothesis/objectives: To assess inter-rater agreement and reliability of TUS findings in feedlot cattle, with or without naturally occurring BP.
Animals: Feedlot steers with (n = 210) or without (n = 107) clinical signs of BP that were assessed by TUS in a previous case-control study.
Methods: A random sample of 50 TUS videos (16-s duration) were scored by 6 raters with various levels of TUS expertise. Lung consolidation, comet tail artifacts, pleural irregularity and effusion were scored. Inter-rater agreement was assessed using raw percentage of agreement (Pa), Cohen's and Fleiss' Kappa (κ), and Gwet agreement coefficient (AC1). Intra-class correlation (ICC) was determined for variables with continuous measurements (mixed factorial design).
Results: Median (interquartile range [IQR]) Pa were 0.84 (0.80-0.89), 0.82 (0.80-0.87), 0.62 (0.53-0.67), and 0.82 (0.75-0.86) for presence of lung consolidation, comet tails, pleural irregularity, and pleural effusion, respectively. For the same lesions, Fleiss κ (95% confidence intervals [CI]) were 0.67 (0.49-0.86), 0.56 (0.33-0.80), 0.20 (-0.05 to 0.44), and 0.36 (0.10-0.61), respectively. AC1 were 0.68 (0.51-0.86), 0.73 (0.58-0.89), 0.21 (-0.01 to 0.44), and 0.71 (0.51-0.92), respectively. Moderate reliability was found among raters for all quantitative variables (ICC ranged from 0.52 to 0.70).
Conclusions and Clinical Importance: Inter-rater agreement was good for presence of lung consolidation, comet tails and pleural effusion (based on Pa and AC1) but was slight to poor for pleural irregularity.
(© 2018 The Authors. Journal of Veterinary Internal Medicine published by Wiley Periodicals, Inc. on behalf of the American College of Veterinary Internal Medicine.)