Assessment of fluid responsiveness in children using respiratory variations in descending aortic flow.

ABSTRACT

BACKGROUND: The primary aim of the current study was to investigate the ability of respiratory variations in descending aortic flow, measured with two-dimensional echo at the suprasternal notch (ΔVpeak dAo), to predict fluid responsiveness in anesthetized mechanically ventilated children. In addition, variations in peak descending aortic flow measured with apical transthoracic echo (ΔVpeak LVOT) were examined for the same properties. METHODS: Twenty-seven patients under general anesthesia were investigated in this prospective observational study. Cardiac output, ΔVpeak dAo, and ΔVpeak LVOT were measured at stable conditions after anesthesia induction. The measurements were repeated after a 10 mL kg-1 fluid bolus. Patients were classified as responders if stroke volume index increased by >15% after fluid bolus. The ability of each parameter to predict fluid responsiveness was assessed using receiver operating characteristic curves. RESULTS: Twenty-seven patients were analyzed, mean age and weight 43 months and 16 kg, respectively. Twelve responders and 15 non-responders were identified. ΔVpeak dAo was significantly higher in the responder group (14%, 95% confidence interval [CI] 12%-17%) compared to the non-responder group (11%, 95% CI 9%-13%) (p = .04) at baseline. Area under the ROC curve for ΔVpeak dAo and ΔVpeak LVOT was 0.73 (95% CI 0.52-0.89, p = .02) and 0.56 (0.34-0.78, p = .3), respectively. A baseline level of ΔVpeak dAo of >14% predicted fluid responsiveness with a sensitivity of 58% (95% CI 28%-85%) and specificity of 73% (95% CI 45%-92%). CONCLUSION: In mechanically ventilated children, ΔVpeak dAo identified fluid responders with moderate diagnostic power in the current study. ΔVpeak LVOT failed to predict fluid responders in the current study.