Cardiac output quantification by Doppler echocardiography in intensive care--limitations and validation

Cardiac output (CO) quantification is primordial to the evaluation of patients with heart failure who are on tailored therapy and under invasive hemodynamic monitoring. Doppler echocardiography can be used to access CO noninvasively, but the concordance between its results and those obtained by invasive methods in paired measurements is still controversial. To our knowledge, no previous studies have assessed the clinical relevance of Doppler echocardiography for CO serial evaluation in patients submitted to tailored therapy.

To evaluate the usefulness of echo-Doppler in the assessment of CO and quantification of changes in CO, compared to thermodilution, in patients with advanced heart failure under hemodynamic monitoring to guide tailored therapy.

In 20 patients (14 male, 62 +/- 14 years old, all in sinus rhythm), with dilated cardiomyopathy and NYHA IV, admitted to the intensive care unit (ICU), CO was simultaneously determined by Doppler echocardiography (dpCO) and thermodilution (tdCO) in three serial evaluations (overall 60). The dpCO was calculated by multiplying the aortic orifice area by the velocity-time integral of aortic continuous wave Doppler flow and by the heart rate. A difference between tdCO and dpCO of more than 20% was considered a major error.

In the overall evaluations, dpCO systematically overestimates tdCO (p = 0.026). The correlation between tdCO and dpCO was 0.81, the mean difference between measurements was 0.40 +/- 0.61 l/min (mean -2SD = -1.62 mean +2SD = 0.81) and 19 (32%) major errors occurred. No significant difference was found between CO percentual variation values assessed by both methods, with a stronger correlation (r = 0.92-p = 0.014) compared to that found for absolute values. On using the dpCO/tdCO ratio in the first evaluation to correct subsequent dpCO, the correlation was fairly good (r = 0.96-p = 0.0002 versus corrected dpCO). The mean difference between paired measurements was significantly lower (0.12 +/- 0.28 l/min-mean 2SD = -0.44 mean +2SD = 0.67), and there were no errors.

CO estimated by Doppler echocardiography has a good correlation with thermodilution although with a weak concordance between paired results in patients with dilated cardiomyopathy and advanced heart failure admitted to the ICU for tailored therapy. Our results with dpCO percentual change in repeated evaluations and with corrected dpCO value after a single simultaneous invasive determination suggest that Doppler echocardiography is a valid method for clinical purposes, allowing us to propose a reduction in the time period of invasive hemodynamic monitoring.