Reproducibility of scoring emphysema by HRCT. Comparison of visual interpretation methods with pulmonary function tests among farmer's lung patients

Abstract

We evaluated the reproducibility of three visual scoring methods of emphysema and compared these methods with pulmonary function tests (VC, DLCO, FEV1 and FEV%) among farmer's lung patients and farmers. Three radiologists examined high-resolution CT images of farmer's lung patients and their matched controls (n=70) for chronic interstitial lung diseases. Intraobserver reproducibility and interobserver variability were assessed for three methods: severity, Sanders' (extent) and Sakai. Pulmonary function tests as spirometry and diffusing capacity were measured. Intraobserver kappa-values for all three methods were good (0.51-0.74). Interobserver kappa varied from 0.35 to 0.72. The Sanders' and the severity methods correlated strongly with pulmonary function tests, especially DLCO and FEV1. The Sanders' method proved to be reliable in evaluating emphysema, in terms of good consistency of interpretation and good correlation with pulmonary function tests.