PREDICTION OF THE 3RD AND 4TH HEART SOUNDS BY DOPPLER-ECHOCARDIOGRAPHY

Doppler echocardiographic variables were sought for predicting the thi rd and fourth heart sounds, as documented by phonocardiography. Phonoc ardiographic recordings of gallop sounds and Doppler echocardiographic investigations of mitral inflow and pulmonary venous flow were evalua ted 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 ve locity (FV), the interval front peak ECG R wave to peak early diastoli c mitral FV, and the early diastolic mitral FV deceleration time. At t he observed prevalence of the third heart sound (47%), the predictive positive value was 81% and the predictive negative value was 86%. A co rrect identification of 72% of the subjects with, and 83% without, a f ourth sound was possible by discriminant analysis using the ratio of p eak 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 prevale nce 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 predict able by a combination of Doppler echocardiographic variables.