Prediction of fluid responsiveness by dynamic preload indices in patients undergoing major hepatic resection

Background and Goal of the Study: Dynamic preload indices, based on the arterial pressure waveform (APW; semi-invasive) or on the plethysmographic waveform (PW; non-invasive) are increasingly used to assess fluid responsiveness. We compared the ability of the commercially available APW-based stroke volume variation (SVV) and the PW-based plethysmographic variability index (PVI) with self-calculated dynamic indices to predict fluid responsiveness in patients undergoing major hepatic resection. Materials and Methods: After local IRB approval, 30 patients were included. Patients received a fluid bolus (FB) of 15 ml/kg in 30 minutes after completion of resection and were considered responders when stroke volume index (SVI) increased > 20 % above pre-FB values. SVV and SVI were measured by the FloTrac-Vigileo system (Edwards Lifesciences, Irvine, USA), PVI by the Masimo Radical 7 pulse co-oximeter® (Masimo Corp, Irvine, USA). The APW and PW were recorded using RugLoop software (Demed, Temse, Belgium) for calculation of APW-based systolic pressure and pulse pressure variation (SPV, PPV) and PW-based variation in peak amplitude (PWVpeak and pulse amplitude (PWVpulse) using an automated algorithm. Areas under the ROC curve (AUROC) were calculated to assess the ability of all indices to predict fluid responsiveness and to assess the optimal cut-off value. Results and Discussion: 17/30 patients were fluid responsive. Mean SVI increased from 36 (12) to 45 (13) ml/m². ROC analysis revealed an AUROC for the APW based SVV, SPV and PPV of0.81, 0.75 and 0.77 and for the PW based PVI, PWVpeak and PWVpulse of 0.78, 0.87 and 0.77, respectively. Optimal cut-off values with associated sensitivity and specificity are shown in table 1. Conclusion: All investigated semi- and non-invasive dynamic preload indices are able to predict fluid responsiveness with similar sensitivity and specificity in patients undergoing major hepatic resection. However, the differences in cut-off values of the various dynamic preload indices should be acknowledged