Carotid Doppler flowmetry correlates poorly with thermodilution cardiac output following cardiac surgery

BACKGROUND: It remains unclear whether measuring carotid arterial flow by the time velocity integral using vascular Doppler ultrasound can be used to monitor cardiac output and volume responsiveness. METHODS: The carotid Doppler flow (time velocity integral and peak flow velocity variation) was assessed in triplicate by an intensivist with formal vascular ultrasound training. Thirty-three patients admitted following coronary by-pass surgery were studied before and after a passive leg-raising manoeuvre to investigate volume responsiveness (more than 10% increase in cardiac output) along with indices of arterial load measuring cardiac output by thermodilution. Pearson's correlation coefficient and area under the curve (AUC) by receiver operating characteristics were calculated. RESULTS: A significant correlation between carotid Doppler flow and cardiac output was demonstrated in post-operative cardiac surgery patients (r = 0.80 [95%CI 0.61-0.89]), including relative changes following passive leg raising (r = 0.79 [95%CI 0.60-0.89]) that showed a mean difference of 2% with wide limits of agreements (-19% to 16%). Changes in carotid Doppler flow following passive leg raising correlated with the baseline arterial resistance but not with compliance or effective elastance. A peak flow variation > 10% before passive leg raising discriminated responders to the manoeuvre with an AUC of 0.81 [95% CI 0.55-0.95]. CONCLUSIONS: Weak correlations between common carotid Doppler flow and cardiac output mean that the methods cannot be used interchangeably in post-operative cardiac surgery patients.