Prediction of the Third and Fourth Heart Sounds by Doppler Echocardiography

Doppler echocardiographic variables were sought for predicting the third and fourth heart sounds, as documented by phonocardiography. Phonocardiographic recordings of gallop sounds and Doppler echocardiographic investigations of mitral inflow and pulmonary venous flow were evaluated in 85 subjects by discriminant and multiple regression analysis. Of 85 subjects 47% had a third sound and 72% a fourth sound, evaluated by phonocardiography. A correct identification of 85% subjects with, and 82% without, the third sound was possible by discriminant analysis using the ratio of peak early diastolic to peak atrial mitral flow velocity (FV), the interval from peak ECG R wave to peak early diastolic mitral FV, and the early diastolic mitral FV deceleration time. At the observed prevalence of the third heart sound (47%), the predictive positive value was 81% and the predictive negative value was 86%. A correct identification of 72% of the subjects with, and 83% without, a fourth sound was possible by discriminant analysis using the ratio of peak early diastolic to peak atrial mitral FV, the interval between the end of atrial mitral FV and the peak ECG R wave, and the duration of pulmonary venous reverse FV at atrial systole. At the observed prevalence of the fourth heart sound (72%) the predictive positive value was 92% and the predictive negative value was 54%. By multiple regression analysis, up to 50% of the amplitude of both gallop sounds was predictable by a combination of Doppler echocardiographic variables.