Abstract
BACKGROUND: The prediction of fluid responsiveness in critical patients helps clinicians in decision making to avoid either under- or overloading of fluid. This study was designed to determine whether lung recruitment maneuver (LRM) would have an effect on the predictability of fluid responsiveness by the changes of hemodynamic parameters in pediatric patients who were receiving lung-protective ventilation and one-lung ventilation (OLV). METHODS: A total of 34 children, aged 1-6 years old, scheduled for heart surgeries via right thoracotomy were enrolled. Patients were anesthetized and OLV with lung-protection ventilation settings was established, and then, positioned on left lateral decubitus. LRM and volume expansion (VE) were performed in sequence. Heart rate (HR), systolic arterial pressure (SAP), mean arterial pressure (MAP) diastolic arterial pressure (DAP), stroke volume (SV), stroke volume variation (SVV), and pulse pressure variation (PPV) were recorded via an A-line based monitor system at the following time points: before and after LRM (T1 and T2) and before and after VE (T3 and T4). An increase in stroke volume (SV) or mean arterial pressure (MAP) of ≥10% following fluid loading identified fluid responders. The predictability of fluid responsiveness by the changes of SV (ΔSV(LRM)) and MAP (ΔMAP(LRM)) after LRM and VE were statistically evaluated by receiver operating characteristic curves [area under the curves (AUC)]. RESULTS: SVs in all patients were significantly decreased after LRM (p < 0.01) and then, increased and returned to baseline after VE (p < 0.01). In total, 16 out of 34 patients who were fluid responders had significantly lower SV after LRM compared to that in fluid non-responders. The area under the receiver operating characteristic curves for ΔSV(LRM) was 0.828 (95% confidence interval [CI], 0.660 to 0.935; p < 0.001) and it indicated that ΔSV(LRM) was able to predict the fluid responsiveness of pediatric patients. MAPs in all patients were also decreased significantly after LRM, and 12 of them fell into the category of fluid responders after VE. Statistically, ΔMAP(LRM) did not predict fluid responsiveness when LRM was considered as an influential factor (p = 0.07). CONCLUSIONS: ΔSV(LRM), but not ΔMAP(LRM), showed great reliability in the prediction of the fluid responsiveness following VE in children during one-lung ventilation with lung-protective settings. TRIAL REGISTRATION: ChiCTR2300070690.