RESUMO
BACKGROUND: Accurate estimation of cardiac preload during liver transplantation is essential. The right ventricular end-diastolic volume index (RVEDVI) is recognized as a good preload indicator in patients undergoing liver transplantation. Recently, dynamic variation parameters including pleth variability index (PVI) have been used as predictors of fluid responsiveness. However, the correlation between PVI and preload status has not been well studied. We evaluated the relationship between PVI and RVEDVI during liver transplantation. METHODS: Eighteen patients undergoing liver transplantation were enrolled in this study. Data of hemodynamic parameters including PVI derived by Masimo Rainbow SET Pulse CO-Oximeter, central venous pressure (CVP), pulmonary arterial occlusion pressure (PAOP), and RVEDI were obtained at 10 defined time points throughout liver transplantation. The correlation between RVEDVI and CVP, PAOP, and PVI was analyzed using Spearman rank test. We also investigated the ability of PVI to accurately differentiate RVEDVI <123 or >142 mL/m(2) using receiver operating characteristic (ROC) analysis. RESULTS: There was fair to good correlation between PVI and RVEDVI (correlation coefficient = -0.492, P < .001). The correlation coefficient between CVP, PAOP, and RVEDVI was 0.345 and 0.463, respectively. A 13.5% cutoff value of PVI estimated the RVEDVI <123 mL/m(2) (area under the curve [AUC] = 0.762). A 12.5% cutoff value of PVI estimated the RVEDVI >142 mL/m(2) (AUC = 0.745). CONCLUSIONS: PVI presented as a reliable estimate of preload status and may be a useful predictor of fluid responsiveness in patients undergoing liver transplantation.
