Value and limitations of Doppler ultrasound in the evaluation of aortic stenosis: a statistical analysis of 70 consecutive patients.

Continuous-mode Doppler ultrasound and cardiac catheterization were performed within 1.4 +/- 2.0 days in 70 patients, aged 26 to 84 (mean 67 +/- 11) years, with suspected aortic stenosis. Optimal Doppler spectral display signal was recorded from the apical window in 43% of the patients, the second or third right parasternal area in 34%, and from the suprasternal notch view in 20% of the patients. Aortic valve gradients by Doppler ultrasound were calculated by the simplified Bernoulli equation: pressure gradient = 4X (velocity). There was an overall fair correlation between the peak gradient by Doppler technique and both the maximal instantaneous (r = 0.66) and peak-to-peak aortic pressure gradient (r = 0.68) obtained at catheterization and a good correlation (r = 0.75) between mean gradient by Doppler technique and catheterization. These correlation coefficients improved significantly in the last 51 patients as compared to the initial 19 patients of the study. Of the last 51 patients, correlation was better in those in normal sinus rhythm as compared to those in atrial fibrillation, and in patients with no or insignificant coexistent AR as compared to those with greater than or equal to 2+ AR. The age of the patient and the status of cardiac output did not significantly affect the accuracy of correlations. We conclude that measurements of aortic valve gradient by continuous-mode Doppler ultrasound may not correlate closely with those by catheterization when the experience with the Doppler technique is limited and when patients in atrial fibrillation or with significant coexistent AR are studied.