Comparison of lactate concentration and model for end-stage liver disease to predict early mortality after liver transplantation / 临床麻醉学杂志

ABSTRACT

Objective To compare the accuracy of lactate concentration and model for end-stage liver disease (MELD) to predict early mortality after liver transplantation, and to provide a convenient and timely predictive tool for clinical work. Methods A total of 121 patients who underwent Allograft liver transplantation in 2017 in our hospital, 92 males and 29 females, aged 25-78 years old, ASA physical status Ⅲ or Ⅳ, were selected to the retrospective study. Patients were divided into survival group and non-survival group, according whether survived within 30 days after the surgery. The preoperative and postoperative lactate concentrations were gotten and the preoperative and postoperative MELD scores were calculated. The preoperative lactate concentrations, postoperative lactate concentrations, preoperative MELD scores and postoperative MELD scores were compared between survival group and non-survival group, respectively. The predictive accuracy about early mortality (30 days) of lactate concentration and MELD score were compared at preoperation and postoperation through receiver operating characteristic, respectively. Results There were 109 patients (90.0％) survived and 12 patients (10.0％) non-survived within 30 days after the surgery. Compared with survival group, the preoperative lactate concentrations, postoperative lactate concentrations, preoperative MELD scores and postoperative MELD scores in the non-survival group were significantly increased (P ＜ 0.05). According to the ROC analysis, the area under the curve (AUC) of ROC of preoperative lactate concentration was 0.78 (95％ CI 0.62-0.93) with a cutoff value of 2.43 mmol/L, the AUC of preoperative MELD score was 0.70 (95％ CI 0.53-0.87) with a cutoff value of 24.50 scores. There was not statistically significant between the AUC of preoperative MELD score and lactate concentration. The AUC of postoperative lactate concentration was 0.85 (95％ CI 0.70-0.99) with a cutoff value of 9.57 mmol/L, the AUC of postoperative MELD score was 0.74 (95％ CI 0.61-0.88) with a cutoff value of 25.42 scores. The AUC of postoperative lactate concentration was statistically higher than that of MELD score (P ＜ 005). Conclusion The predictive capability of lactate concentration, especially that of postoperative lactate concentrations, was better than the predictive capability of MELD score.