Reproducibility of CT signs of right ventricular dysfunction in acute pulmonary embolism.

Abstract

The purpose of our study was to determine the interobserver reproducibility of CT findings of right ventricular (RV) dysfunction in pulmonary embolism (PE). Two experienced observers independently and retrospectively evaluated pulmonary CT angiography (CTA) studies of 50 patients with acute PE for the following signs: bowing of the interventricular septum, inferior vena cava (IVC) contrast medium reflux, RV diameter (RVD)/left ventricular diameter (LVD) ratio on axial sections and four-chamber (4-CH) views, and RV volume (RVV)/left ventricular volume (LVV) ratio. Analysis used kappa statistics, Spearman's rank correlation, and Bland-Altman statistics. The two observers had fair to moderate agreement (kappa = 0.32-0.44) for septal bowing and moderate to good agreement (kappa = 0.57-0.68) for diagnosing IVC reflux. The Spearman's rank correlation coefficients for RVD(axial)/LVD(axial) ratio and RVD(4-CH)/LVD(4-CH) ratio between the two observers were 0.88 (p < 0.001) and 0.85 (p < 0.001), respectively. On Bland-Altman analysis, the mean differences for RVD(axial)/LVD(axial) ratio and RVD(4-CH)/LVD(4-CH) ratio were 0.014 (+/- 0.195) and 0.001 (+/- 0.242), respectively. The correlation coefficient for RVV/LVV ratio was 0.93 (p < 0.001), and the mean difference was 0.033 (+/- 0.229). Considerable differences exist in the interobserver reproducibility of CT findings of RV dysfunction on pulmonary CTA in patients with acute PE. Cardiac chamber measurements are more reproducible than septal bowing and IVC reflux. Volumetric determination of the RVV/LVV ratio is the least user dependent and most reproducible.