Predictors of early extubation after pediatric liver transplantation: a single-center retrospective cohort study

Liver transplantation is a cornerstone therapy for pediatric end-stage liver disease. While Enhanced Recovery After Surgery (ERAS) protocols, including early extubation, are well-established in adult liver transplantation, their application in children is more cautious due to unique challenges like malnutrition and poor postoperative cooperation. This study aimed to identify predictors of successful early extubation (within 3 h postoperatively) in pediatric liver transplant recipients.

This was a single-center retrospective cohort study. We analyzed data from 256 patients under 18 years who underwent liver transplantation between September 2019 and November 2022. Patients were divided into Early Extubation (EE, ≤ 3 h) and Non-Early Extubation (Non-EE, > 3 h) groups. Univariate and multivariate logistic regression analyses were performed to identify factors associated with EE. The efficacy of relevant parameters in predicting early extubation was evaluated using receiver operating characteristic (ROC) curve analysis.

Of the 256 patients, 155 (61%) were successfully extubated early. Multivariate analysis identified the following independent predictors: favorable factors were higher preoperative albumin (ALB) level (adjusted OR 1.167, p = 0.008) and greater intraoperative urine output (adjusted OR 4.686, p < 0.001); unfavorable factors were higher preoperative white blood cell (WBC) count (adjusted OR 0.806, p = 0.038), higher end-of-surgery serum lactate level (adjusted OR 0.548, p < 0.001), presence of a concurrent respiratory tract infection (adjusted OR 0.307, p = 0.021), and receiving a donation after circulatory death (DCD) graft (adjusted OR 0.260, p = 0.030). ROC curve analysis identified optimal cutoff values for predicting early extubation: total urine output showed strong discriminative ability (AUC 0.898, 95% CI 0.859-0.938, sensitivity 79.4%, specificity 88.1%); EOS serum lactate showed moderate discriminative ability (AUC 0.725, 95% CI 0.659-0.791, sensitivity 49.5%, specificity 91.0%); pre-op ALB and pre-op WBC showed modest discriminative ability (AUC 0.663 and 0.623, respectively). Optimal cutoff values were: pre-op ALB ≥ 36.15 g/L, pre-op WBC ≤ 11.82 × 10⁹/L, urine output ≥ 3.05 mL/kg/h, and EOS lactate ≤ 3.55 mmol/L.The EE group had a significantly shorter median ICU stay (3 vs. 3.5 days, p = 0.002), with no significant difference in 30-day complication rates or total hospital stay.

Preoperative physiological reserve and infection status (reflected by ALB as a composite marker of hepatic synthetic function and nutritional status, WBC, and respiratory infection), intraoperative organ perfusion and metabolic status (reflected by urine output and lactate level), and graft quality (DCD grafts) are key independent predictors of early extubation, potentially shortening ICU stay and supporting the implementation of ERAS protocols in this population.