Cardiac index quantification by Doppler ultrasound in patients without left ventricular outflow tract abnormalities.

OBJECTIVES: We attempted to ascertain whether cardiac index can be directly estimated from Doppler mean velocity.

BACKGROUND: Although diverse Doppler echocardiographic methods have been described for cardiac output quantification, they are not widely used in clinical practice. Cross-sectional area measurement has been identified as the main source of error in flow volume quantification.

METHODS: A three-phase study by Doppler echocardiography was conducted in 306 patients. In phase I, the normal mean velocity ratio of the left and right ventricular outflow tracts was established in 170 normal subjects. In phase II, cardiac index, calculated as the product of aortic annular area index by mean velocity (conventional method), and mean velocity determined in the left ventricular outflow tract and ascending aorta by pulsed and continuous wave Doppler, respectively, were correlated with thermodilution cardiac index in 66 patients. In phase III, the accuracy of the regression equations obtained was prospectively assessed in an additional 70 patients.

RESULTS: The normal left/right ventricular outflow tract mean velocity ratio by pulsed wave Doppler was 1.1 +/- 0.1. Cardiac index (CI) calculated by the conventional method and thermodilution (TD) showed acceptable correlation (r = 0.90, CITD = 1.20 CIPWD + 357; r = 0.86, CITD = 0.90 CICWD + 262) for pulsed (PWD) and continuous wave (CWD) Doppler, respectively, but with systematic underestimation (-28 +/- 13%, p < 0.01) by pulsed wave Doppler. Mean velocity (MV) showed excellent correlation with the thermodilution cardiac index (r = 0.97, CITD = 172 MVPWD - 172; r = 0.93, CITD = 129 MVCWD - 255). When these regression equations were prospectively applied, better agreement with the thermodilution cardiac index was obtained by pulsed wave Doppler directly from mean velocity (SD 240 ml/min per m2) than when aortic annular area was considered in the calculation (SD 428 ml/min per m2). Similar results were obtained by continuous wave Doppler (SD 433 vs. 599 ml/min per m2) but with less accuracy.

CONCLUSIONS: Left ventricular outflow tract mean velocity determined by pulsed wave Doppler permits easy, accurate cardiac index quantification in the absence of left ventricular outflow abnormalities. The simplicity of this method enhances its clinical applicability in noninvasive monitoring of cardiac index.