Reliability of echocardiography in assessing cardiac output

Because of the potential benefits from a noninvasive technique in assessing cardiac output, we compared cardiac output estimates from left ventricular echocardiograms with results obtained simultaneously by a standard technique, dye dilution in 10 healthy normal volunteers. During rest, cardiac outputs by echocardiographic and dye dilution techniques were reproducible and not significantly different. Increases in cardiac output produced by intravenous infusion of isoproterenol (15 ng/kg/min for 4 min) were accurately estimated by echocardiography in subjects whose stroke volume increased less than 40%, but were significantly underestimated when stroke volume increased more than 40%. Decreased cardiac output produced by intravenous propranolol (0.2 mg/kg) was comparable by both methods. Although echocardiography accurately estimated mean cardiac output for the group it over- or underestimated cardiac output in individual subjects. We propose that echocardiography can reliably estimate cardiac output in groups at rest and when stroke volume changes less than 40%.