Introduction: The aim of this retrospective study was to evaluate any relationship between cardiac power index (CPI) and preload indexes during liver transplantation (LT).
Methods: Thirty-three patients with normal preoperative cardiac evaluation undergoing LT were included. Anesthesia management was standardized. Monitoring included continuous cardiac output determination by pulmonary artery catheter. CPI was calculated throughout LT by using the following standard formula: Mean Arterial Pressure [mm Hg] × Cardiac Index [L/min/m2] × k, where k = 0.0022. A logistic regression to determine which preload indexes predicted an adequate CPI (≥ 0.4 watt/m2) was performed. Postregression analysis was carried out to calculate a cutoff of right ventricle end diastolic volume index (RVEDVI) able to guarantee an adequate CPI after establishing a sensitivity >0.9. The area under receiver operating characteristic curve (AUC) was also run separately for patients with a Model for End-Stage Liver Disease (MELD) score < or ≥ 25 to establish an accurate level of prediction in these subgroups (post-hoc analysis).
Results: Logistic regression showed that RVEDVI was the only predictor of CPI (AUC = 0.81). A cutoff value for RVEDVI of 105 mL/m2 was found (sensitivity = 90.5%; specificity = 50%). RVEDVI predicted CPI with moderate accuracy (AUC = 0.80) in patients with MELD < 25 (n = 25), whereas the prediction was highly accurate (AUC = 0.96) in patients with MELD ≥ 25 (n = 8).
Conclusion: An RVEDVI = 105 mL/m2 can be considered a valid cutoff to perform a fluid challenge to optimize preload during LT. Sicker recipients (with MELD ≥ 25) could exhibit less tolerance to preload reduction, proven by a decrease of CPI below the minimum value considered safe (0.40 watt/m2).
Copyright © 2020 Elsevier Inc. All rights reserved.