Evaluation of intraoperative fluid responsiveness with two different monitoring method in liver transplantation

Objective: The aim of this study was to examine whether pleth variability index (PVI), obtained with Masimo® SET Rainbow system, could predict fluid responsiveness in liver transplantation (LT), and to compare PVI with the preload and fluid responsiveness parameters that measured by PICCO2 monitoring.Materials and Methods: Twenty-five patients, older than 18 years, and undergoing orthotopic LT due to end-stage liver failure, were enrolled to the study. After standart anesthesia induction, patients were monitored with Masimo that had PVI software for measurement of perfusion index and PVI were measured. Through PICCO2 and central venous pressure monitorings, cardiac index (CI), stroke volume index (SVI), stroke volume variation (SVV), pulse pressure variation (PPV), global end-diastolic index (GEDI) and intratorasic blood volume (ITBV) were recorded by transpulmonary termodilution method. Parameters obtained by Masimo and PICCO2 monitoring at the specified times were recorded simultaneously. In addition, 10 mL/kg % 6 hydroxyetyl starch boluses were given after discharge of ascite during dissection and after clamping of the vena cava during anhepatic phase. Patients were divided into two groups as had changed higher than 15% in CI (responders) or lower than 15% (nonresponders) after fluid infusion.Results: In dissection phase the ratio of responders and nonresponders were 14/11, while the ratio were 18/6 in anhepatic phase. There were no differences between responders and nonresponders according to the baseline PVI values. Baseline SVV values of responders were higher than responders in both phases. The area under ROC curve (AUC) for PVI was 0.56 (sensitivity 35% ve specifity 90%, p=0.58), and the threshold value was ?7 in dissection phase. The AUC for PVI was 0.55 (sensitivity 55% ve specifity 66%, p=0.58), and the threshold value was >16 in dissection phase. The higher AUC value was belonged to SVV among the parameters. Baseline SVV correlated significantly with ?CI (dissection r=0.60, p=0.001; anhepatic r=0.44, p=0.02) whereas there were no correlation between values of PVI, SVB, PPV, GEDI, ITBV and ?CI before fluid loading. Furthermore, there was no correlation between ?PVI values and ?SVV, ?PPV.Conclusion: We conclude that PVI is not a reliable predictor for fluid responsiveness in operation that has variable peripheral perfusion and frequent uses of vasoconstrictors like LT, however, SVV that obtained PICCO2 system is a reliable indicator.