Visual estimation of the severity of aortic stenosis and the calcium burden by 2-dimensional echocardiography: Is it reliable?

Objectives - Guidelines have recommended aortic valve surgery in asymptomatic patients with severe aortic stenosis and a large aortic valve calcium burden. The purpose of this study was to determine whether visual assessment of aortic valve calcium and stenosis severity are reliable based on 2-dimensional echocardiography alone. Methods - We prospectively enrolled 68 patients with aortic stenosis and compared them with 30 control participants without aortic stenosis. All had aortic valve calcium score assessment by computed tomography. In a random order, 2-dimensional images without hemodynamic data were independently reviewed by 2 level 3-trained echocar-diographers, who then classified these patients into categories based on aortic valve calcium and stenosis severity. Results - The 68 patients (mean age ± SD, 74 ± 10 years) were classified as having mild (n = 28), moderate (n = 22), and severe (n = 18) aortic stenosis. When the observers were asked to grade the degree of valve calcification, the agreement between them was poor (κ = 0.33-0.39). The visual ability to determine stenosis severity compared with Doppler echocardiography had high specificity (81% and 88% for observers 1 and 2). However, sensitivity was unacceptably low (56%-67%), and the positive predictive value was poor (44%-50%). Agreement was fair (κ = 0.58-0.69) between the observers for determining severe stenosis. Conclusions - Our results suggest that visual assessment of aortic valve calcium has high interobserver variability; the visual ability to determine severe aortic stenosis has low sensitivity but high specificity. Our results may have important implications for treatment of patients with aortic stenosis and guiding the use of handheld echocardiography. Further research with larger cohorts is needed to validate the variability, sensitivity, and specificity reported in our study.