Diagnostic imaging in acute interstitial pneumonia in foals : high variability of interpretation of chest radiographs and good conformity between ultrasonographic and post-mortem findings
Acute interstitial pneumonia (AIP) in foals is a rare but challenging syndrome. Diagnostic imaging is crucial for its diagnosis. While there are some reports on radiographic findings, ultrasound is sparsely described. Variability in the evaluation of radiographs in AIP has been well described in human literature. Part one of this study is a prospective, observational, observer agreement study investigating inter- and intraobserver agreement in the assessment of thoracic radiographs. Part two is a prospective, controlled, descriptive study on thoracic ultrasonography in foals with AIP. Eighteen foals with AIP were examined daily by thoracic ultrasound. Thoracic radiographs were taken on three successive occasions. Blinded radiographs were assessed by three observers based on two semi-quantitative scores (pattern recognition, subjective evaluation). Foals that died underwent postmortem examination (n = 9); postmortem findings were compared to ultrasound findings on the day of death. Ultrasonographic findings were consistent with postmortem findings. Comet tail scores in foals with AIP were significantly higher (p < 0.0001) than in control foals. Interrater agreement for the assessment of radiographs was none to moderate (κ = 0.07-0.65) for pattern recognition and weak to moderate (κ = 0.58-0.62) for subjective scoring. Intrarater agreement varied from minimal to strong (κ = 0.30-0.80) for pattern recognition but was strong (κ = 0.83) for subjective scoring. In conclusion, the diagnostic value of thoracic ultrasound in foals with AIP is high due to good conformity with postmortem findings. The evaluation of thoracic radiographs showed high variability in inter- and intra-agreement.