Background: Liver transplantation (LT) is a high-risk surgery associated with significant hemodynamic changes requiring advanced hemodynamic monitoring. Pulmonary Artery Catheter (PAC) is still considered as a gold-standard for Cardiac Index (CI) measurement during LT despite association with an increased risk of complications. Noninvasive impedance cardiography (ICG) could be an interesting alternative tool for CI monitoring. The aim of this study was to compare the precision and trending ability of ICG versus PAC methods during LT.
Methods: Patients undergoing LT were prospectively included. CI was measured with PAC and ICG at 4 time points (T1: before surgical incision, T2: during anhepatic phase, T3: after portal reperfusion, T4: during wound closure). Bias and percentage error (PE) between CI measured with PAC and ICG were analyzed with the Bland-Altman method for repeated measurements. Trending ability was studied with 4-quadrant and polar plots and correlation coefficient.
Results: We included 43 patients with 156 measures. Mean bias was -0.95 L.min-1.m-2, SD±1.07, limits of agreement -3.73 to 1.83 L.min-1.m-2 and PE 58%. There was a significant increase in bias during LT (P<0.001). Assessment of trending ability displayed a concordance rate of 72% on the 4-quadrant plot and a mean angular bias of -8.4° (SD±28°) and radial limits of agreement ±55° on the polar plot.
Conclusions: CI measurements using ICG exhibited a low precision and a poor trending ability when compared to thermodilution method during LT. Consequently, ICG is not an adequate hemodynamic tool to monitor CI during LT.