ABSTRACT
The optimal echocardiographic methodology for predicting need for intervention in children with valvar aortic stenosis (VAS) is not known. We reviewed echocardiograms and catheterization reports of 79 children (aged 9.5 +/- 5.9 years) with isolated VAS. The maximum and mean Doppler-predicted gradients from the apical (MIGAP), MEGAP)) and the suprasternal or right parasternal (MIGHP), MEGHP)) windows were measured. The peak-to-peak catheterization gradient and the intervention (if any) were recorded. All sites and methods of Doppler estimation of VAS gradient correlated in a linear fashion with the invasive gradient (R2 = 0.34-0.50) and with one another (R2 = 0.48-0.86). MIGAP and MIGHP overestimated the invasive gradient in 60% and 86% of patients, whereas MEGAP and MEGHP underestimated the invasive gradient in 94% and 83% of patients, respectively. Age and diameter of the ascending aorta had small but significant effects on the level of agreement. A MIGHP < or = 55 mm Hg predicted no intervention with 100% accuracy, whereas the specificities of a MIGHP > 90 mm Hg, a MEGAP > 50 mm Hg, and a (MIGAP + MIGHP)/2 > 70 mm Hg for intervention were 94%, 100%, and 92%, respectively. The magnitude of overestimation was significantly lower from the apical window. In children with VAS, the best prediction of the catheterization gradient could be based on the average of MIGAP and MIGHP.
