Assessment of intra- and interobserver variability of pulmonary vein measurements from CT angiography 1

Abstract

Pulmonary vein (PV) stenosis is a common complication after radiofrequency ablation for atrial fibrillation. This study investigated the intra- and interobserver variability and precision of PV ostial measurements from three-dimensional computed tomography angiography. Four observers measured the maximum and minimum diameters, as well as area, of the four PVs of seven patients who underwent a three-dimensional computed tomography scan before radiofrequency ablation. Each observer performed two sets of measurements. The intra- and interobserver variability of the measurements was calculated using analysis of variance. Intraobserver variability was approximately two times lower than interobserver variability in measurements of diameter and area. The standard error of the measurement (SEM), SEM(intra) and SEM(inter), were lower for the mean diameter than the maximum diameter. The minimum detectable changes in diameters, DeltaD(intra) and DeltaD(inter), and area, DeltaA(intra) and DeltaA(inter), demonstrated the same statistical trend as the corresponding SEM for each of the four veins. Both DeltaD(intra) and DeltaD(inter) and DeltaA(intra) and DeltaA(inter) were smaller than the corresponding lower bounds of the 95% confidence interval for 50% diameter reduction and 75% area reduction. The direct results of DeltaA(intra) and DeltaA(inter) for each of the four veins were consistently less than DeltaA's calculated from the corresponding DeltaD(intra) and DeltaD(inter). PV ostial measurements are less variable when made by a single observer than by multiple observers, and mean diameter measurements are more precise than a single, maximum diameter measurement. Both diameter and area measurements are capable of quantifying the mild PV stenosis. Furthermore, area can be measured with greater precision than mean diameter and should be used in PV ostium caliber determination.