Radiographic scoring lacks predictive value in inflammatory airway disease

REASONS FOR PERFORMING STUDY: Diagnosis of inflammatory airway disease (IAD) currently rests upon the results of bronchoalveolar lavage (BAL) cytology, lung function testing and histamine bronchoprovocation (HBP), none of which provides direct information about structural change in the lung. HYPOTHESIS: That thoracic radiography might better portray structural change in the lungs and therefore offer a good clinical assessment of IAD. METHODS: A radiographic scoring system was developed to assess the extent of bronchial and interstitial pattern on thoracic radiographs in the dorsocaudal, dorsocranial and caudoventral regions in 16 control horses and 33 horses with IAD. Mean scores were compared to ascertain whether thoracic radiographs could distinguish between the 2 groups. In order to determine whether independent observers reliably scored radiographs similarly, an inter-rater reliability score was employed for each radiographic observation. Correlations between radiographic scores, BAL cytology, lung function testing using the forced oscillatory technique and HBP were examined. RESULTS: Inter-rater reliability was only moderate. Radiographic scoring demonstrated no differences between the 2 groups. There were no correlations among BAL cytology, lung function testing, HBP and radiographic scores. CONCLUSIONS AND POTENTIAL RELEVANCE: Thoracic radiographs are a low-yield diagnostic modality in horses with a clinical history compatible with IAD. In the absence of clinical evidence of more extensive, infectious disease, thoracic radiographs neither refine nor improve diagnosis of IAD, but increase diagnostic costs.