Reproducibility of semi-quantitative assessment of aortic valve calcification and valve motion on echocardiography: a small-scale study

Aortic stenosis (AS) is the most common degenerative valve disease in high income countries. While hemodynamic metrics are commonly used to assess severity of stenosis, they are impacted by loading conditions and stroke volume and are often discordant. Anatomic valve assessments such as aortic valve calcification (AVC) and valve motion (VM) during transthoracic echocardiography (TTE) can offer clues to disease severity. The reliability of these semi-quantitatively assessed anatomic imaging parameters is unknown. This is a retrospective study of semi-quantitative assessment of AVC and valve VM on TTE. TTEs representing a range of AS severities were identified. The degree of calcification of the aortic valve and the degree of restricted VM were assessed in standard fashion. AVC scores and valve motion were assessed by readers with varied training levels blinded to the severity of AS. Correlation and inter-reader reliability between readers were assessed. 420 assessments (210 each for AVC and VM) were collected for 35 TTEs. Correlation of AVC for imaging trainees (fellows and students, respectively), ranged from 0.49 (95% CI 0.18-0.70) to 0.62 (95% CI 0.36-0.79) and 0.58 (95% CI 0.30-0.76) to 0.54 (95% CI 0.25-0.74) for VM. Correlation of anatomic assessments between echocardiographer-assigned AVC grades was r = 0.76 (95% CI 0.57-0.87)). The correlation between echocardiographer-assigned assessment of VM was r = 0.73 (95% CI 0.53-0.86), p