Comparison of continous-wave Doppler ultrasound monitor and echocardiography to assess cardiac output in intensive care patients.

BACKGROUND: Continuous-wave Doppler (CWD) ultrasound through the left ventricular outflow tract is one modality used for non-invasive cardiac output measurement. The ultrasonic cardiac output monitor (USCOM) is a relatively new monitor which uses a small, transcutaneous ultrasound probe to measure cardiac output with CWD via the suprasternal window. It is faster and less complex to train new users than conventional echocardiography. In addition to stroke volume (SV), the USCOM can calculate stroke volume variation (SVV) and the Smith-Madigan inotropy index (SMII), which is an estimate of the pre-load independent contractility of the myocardium. OBJECTIVE: To assess the level of agreement between cardiac output measured with conventional echocardiography and with USCOM. METHODS: A prospective, observational, multicentre trial of patients admitted to the intensive care units of two hospitals. After excluding patients with aortic stenosis, any patient undergoing a clinically indicated echocardiogram also underwent a subsequent USCOM study for comparison. RESULTS: We enrolled 121 patients in the study, with aortic stenosis the main reason for patient exclusion. Of the study patients, 63% were mechanically ventilated, 84% were in sinus rhythm, and the mean age of the study cohort was 66 years (SD, 17 years). There was a very strong correlation between SV as measured by the USCOM and by echocardiography. The mean difference in SV was 0.33 mL (SD, 5.62 mL), r2 = 0.956, and Bland-Altman analysis confirmed no significant bias with acceptable limits of agreement between the methods. Patients who were fluid responsive had an SVV cut point on the receiver operating characteristic curve of 21%, and sensitivity and specificity of 95%. A low SMII (< 1.1 watts/m2) calculated with the USCOM did not correlate well with low cardiac output status, with a sensitivity of only 69%. CONCLUSIONS: SV (and thus cardiac output) measured using the USCOM correlated well with echocardiographic cardiac output measurement, which suggests that the USCOM could be a valuable haemodynamic tool for assessment of cardiac output and fluid responsiveness in critically ill patients if patients with aortic stenosis are excluded. Inotropy, as a parameter of low cardiac output, was not useful in this cohort of patients.