Comparison between thoracic ultrasonography and thoracic radiography for the detection of thoracic lesions in dairy calves using a two-stage Bayesian method

Comparison of thoracic ultrasonography and thoracic radiography to detect active infectious bronchopneumonia in hospitalized dairy calves

“…The Se and Sp of TR for the detection of thoracic lesions has recently been described without regard to the active aspect of disease. 18 Despite wide CI of accuracy estimates, our results show that TR is not only sensitive in detecting thoracic lesions, but also in recognizing hospitalized calves with active BP, as supported by a numerical estimate of 89%. Similarly, TUS revealed a numerical estimation of the Se of 84%.…”

“…4 This result is also similar to the Se of TUS T A B L E 5 Test sensitivity, test specificity, positive likelihood ratio (LR+) and negative likelihood ratio (LRÀ) along with their 95% confidence intervals for thoracic ultrasonography (TUS) and thoracic radiography (TR) using a panel diagnosis method as a gold standard and in detecting thoracic lesions with an Se of 81%. 18 Interestingly, the Sp of 58% and 74% for TR and TUS, respectively, was relatively low compared with studies reporting an Sp of TR of 89% for the detection of thoracic lesions in calves 18 and an Sp of TUS > 90% in detecting chronic, subclinical or active BP. 4,6,40 Different hypotheses could explain these low Sp.…”

“…Finally, this study relied on experts' clinical judgment to make a diagnosis of active BP based on a comprehensive set of clinical information and blood work results. This approach differs from studies using the presence of thoracic lesions on diagnostic imaging examinations as a gold standard 18 or latent class analysis. 4 In this study, calves were classified as having active or nonactive BP based on a PDM.…”

“…A 500 mA and 150 kVp Siemens Vertix machine (Siemens, Mississauga, ON, Canada) and an AGFA DX-S CR system, using a focused grid (grid ratio of 6:1; Reina Imaging, Crystal Lake, IL, USA), were utilized. 18 Radiographs were transferred to a Picture Archiving and Communication System (PACS) and viewed on a single workstation using AGFA Impax 6 (Agfa, Toronto, ON, Canada). A single American College of Veterinary Radiology (ACVR) certified diagnostic imaging specialist, unaware of the component tests and TUS results, reviewed the radiographs 10 months after enrolment of the last calf in the study.…”

“…Descriptive data summarizing thoracic lesions detected on thoracic ultrasonography and thoracic radiography in 50 calves with (active BP +) or without (active BPÀ) active bronchopneumonia (BP) based on a panel diagnosis method Thoracic ultrasonography: (a) lung consolidation defined as hypoechoic lung tissue resembling the sonographic appearance of hepatic parenchyma; (b) pneumothorax diagnosed when the sliding movement of the pleurae was no longer visible; (c) pleural effusion diagnosed when the pleurae were separated from one another by anechoic or echogenic fluid; and (d) nodule described as a well-defined discrete hypoechoic circular region inside the lung parenchyma 17. b Thoracic radiography: (a) bronchial pattern defined as the presence of airways with prominent and/or thickened walls; (b) interstitial pattern defined as the presence of an area of the lung field with increased opacity combined with blurred vascular borders; (c) nodule defined as the presence of a focal area of increased soft tissue opacity (<3 cm) with relatively well-defined borders; (d) alveolar pattern defined as the presence of an area of increased soft tissue opacity silhouetting with pulmonary vessels located within the opacified area; (e) pleural effusion diagnosed based on the retraction of lung lobe margins, pleural fissure lines, and fluid opacity between the thoracic wall and the lung margins; and (f) pneumothorax diagnosed based on the presence of air between the lung lobes and the thoracic wall causing retraction of the lung margins 18. an assumption of covariance between both tests and the paired design of the study, a 2-sided McNemar's test was performed (Additional File 3).Additionally, as recommended by Pepe,36 a comparison between both tests was performed by estimating the relative probabilities of TUS and TR.…”