Wang RR, Zhu M, Ren HC, Yu WL. Machine learning models for predicting acute kidney injury after pediatric living donor liver transplantation in biliary atresia. World J Gastrointest Surg 2026; 18(7): 120759 [DOI: 10.4240/wjgs.v18.i7.120759]
Corresponding Author of This Article
Wen-Li Yu, PhD, Department of Anesthesiology, Tianjin First Central Hospital, Nankai University, No. 24 Fukang Road, Tianjin 300192, China. yzxyuwenli@163.com
Research Domain of This Article
Surgery
Article-Type of This Article
research-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Wang RR, Zhu M, Ren HC, Yu WL. Machine learning models for predicting acute kidney injury after pediatric living donor liver transplantation in biliary atresia. World J Gastrointest Surg 2026; 18(7): 120759 [DOI: 10.4240/wjgs.v18.i7.120759]
Author contributions: Wang RR and Yu WL conceived the manuscript; Yu WL coordinated and supervised data collection; Zhu M carried out the statistical analysis; Ren HC wrote and prepared the tables and figures; Zhu M and Wang RR revised the manuscript with additional detail; Ren HC and Yu WL critically reviewed the manuscript for important intellectual content; Yu WL is the study’s guarantor. All authors have read and agreed to the published version of the manuscript.
Supported by Tianjin Key Clinical Specialty Construction Project, Tianjin Key Medical Discipline Construction Project, No. TJYXZDXK-3-022C; and Scientific Research Program of the Tianjin Municipal Education Commission, No. 2025ZD40.
Institutional review board statement: The study protocol was compliant with the principles of the Declaration of Helsinki and was approved by the Institutional Review Board and Ethics Committee of Tianjin First Central Hospital, No. KYAP2025-170.
Informed consent statement: The requirement for informed consent was waived by the Ethics Committee due to the retrospective nature of the study.
Conflict-of-interest statement: The authors declare that they have no competing interests.
Data sharing statement: No additional data are available.
Corresponding author: Wen-Li Yu, PhD, Department of Anesthesiology, Tianjin First Central Hospital, Nankai University, No. 24 Fukang Road, Tianjin 300192, China. yzxyuwenli@163.com
Received: March 13, 2026 Revised: April 14, 2026 Accepted: May 8, 2026 Published online: July 27, 2026 Processing time: 142 Days and 16.1 Hours
Abstract
BACKGROUND
Living donor liver transplantation (LDLT) is an important treatment method for end-stage pediatric liver diseases, e.g., biliary atresia (BA). Acute kidney injury (AKI) after transplantation is a common and serious complication in clinical practice that significantly influences patient mortality and survival rate.
AIM
To construct a clinical prediction model for AKI after pediatric LDLT based on machine learning (ML).
METHODS
This study included 340 children with BA who underwent LDLT at our center between December 2022 and December 2024. Complete clinical data were collected, including baseline characteristics, preoperative assessments, intraoperative variables, and postoperative recovery indicators. Least absolute shrinkage and selection operator regression was used for feature selection, and nine ML models were developed for model training and evaluation. After training on the training set, the predictive performance of each model was tested and compared. Finally, the best-performing model was interpreted and visualized using the SHapley Additive exPanations (SHAP) algorithm.
RESULTS
Excluding postoperative creatinine (Cr) levels, this study identified a total of six potential predictors associated with AKI after LDLT. The random forest model showed comprehensive and optimal predictive performance after 10-fold cross-validation, with an area under the curve of 0.875 (95% confidence interval: 0.805-0.944). In addition, the importance of predictors for AKI occurrence was ranked by SHAP analysis, and preoperative Cr level was identified as the most important predictor.
CONCLUSION
This study employed ML algorithms to construct a predictive model for early AKI following pediatric liver transplantation. The developed model is expected to assist doctors in performing timely treatment interventions, thereby reducing the occurrence of post-transplant complications and improve the survival time and quality of life in children undergoing liver transplantation.
Core Tip: This study included 340 children with biliary atresia who underwent liver transplantation. Seven key predictors of acute kidney injury (AKI) were screened out by least absolute shrinkage and selection operator algorithm, including pre-operative/post-operative creatinine (Cr), blood calcium and lactic acid levels during the anhepatic phase, gender, the amount of fresh frozen plasma infused during the operation, and post-operative aspartate aminotransferase level. Nine machine learning methods, including XGBoost, were used to construct the postoperative AKI prediction model based on other features after excluding postoperative Cr, and their prediction performance was compared. This study aims to assist clinicians in early intervention and improve the prognosis of children.
Citation: Wang RR, Zhu M, Ren HC, Yu WL. Machine learning models for predicting acute kidney injury after pediatric living donor liver transplantation in biliary atresia. World J Gastrointest Surg 2026; 18(7): 120759
Since 1963, pediatric liver transplantation surgeries have been performed and promoted in multiple transplantation centers around the world, and the long-term postoperative survival rate has continuously improved[1-3]. Living donor liver transplantation (LDLT), as the standard treatment for pediatric end-stage liver diseases, e.g., cholestatic liver diseases, genetic metabolic disorders, fulminant liver failure, and liver tumors, holds significant clinical importance[4-6]. This study primarily focused on the pediatric patients who underwent liver transplantation due to biliary atresia (BA) at a transplant center, with the liver grafts sourced from the children’s parents.
Acute kidney injury (AKI) after LT (post-AKI) is a common postoperative complication following LT, with an incidence rate of 40% to 70%[7,8]. It significantly increases the mortality rate, hospital stay duration, and long-term survival rate of liver transplant patients[9,10]. The occurrence of AKI is closely related to various factors, e.g., intraoperative hemodynamic changes, postoperative drug use, and preoperative renal function status[11]. Due to its high incidence rate and the significance of clinical treatment, predicting the risk of postoperative AKI accurately has become a current research focus.
With the widespread application of machine learning (ML) technology, ensemble learning models, e.g., the XGBoost and AdaBoost models, have played a significant role in multi-disease studies and have achieved notable success in the medical field[12-15]. Constructing and training ML models and using clinical data to predict AKI after pediatric liver transplantation is expected to enhance early identification capabilities, thereby enabling more precise personalized treatment. This paper discusses the construction of different ML models, evaluates the performance of the constructed models in predicting postoperative AKI, and compares the strengths and limitations of different algorithms to identify the most suitable model for practical application in clinical settings.
MATERIALS AND METHODS
Study design and patient selection
This retrospective study was approved by the Ethics Committee of Tianjin First Central Hospital (KYAP2025-170), and it followed the Helsinki Declaration. A retrospective and observational study of 340 patients admitted to our hospital from December 2022 to December 2024 was performed, and a unified analysis framework was constructed using Python (version 3.12.7) and the scikit-learn ML library. Inclusion criteria: (1) Age between 0 and 36 months; (2) American society of anesthesiologists physical status of I-III; (3) Pre-operative diagnosis of BA undergoing LDLT; and (4) Complete clinical data. Exclusion criteria: (1) Combined with congenital heart disease or other serious malformations; (2) History of previous liver transplantation; (3) Unplanned reoperation during hospitalization; (4) Pre-operative chronic kidney disease; and (5) Missing clinical key data. The patients were pathologically diagnosed with BA and underwent autologous liver transplantation. The donors underwent left lobe hepatectomy, and the recipients underwent modified piggyback liver transplantation. Specifically, left hepatic vein–liver vein anastomosis was employed to reconstruct the outflow tract. AKI was defined in accordance with the KDIGO criteria as any of the following: An increase in serum creatinine (Cr) of ≥ 0.3 mg/dL (≥ 26.5 μmol/L) within 48 hours, an increase to ≥ 1.5 times the baseline within 7 days, or a urine output < 0.5 mL/kg/hour for at least 6 hours. When baseline serum Cr was unavailable, renal function was estimated based on age and height[16].
Sample size calculation
Sample size estimation was performed for the binary variable of postoperative AKI after LDLT for the primary outcome. Based on previous studies and data from our center, the incidence of AKI was approximately 25%-30%. With the significance level set at α = 0.05, the statistical power at 80%, and the effect size between the main predictors and the occurrence of AKI corresponding to an odds ratio of approximately 2.0. Combining the methods related to random forest and Logistic regression for estimation, the required sample size was approximately 280-320 cases. Considering possible exclusions and missing data, the final study included 340 patients.
Data collection
Clinical data were collected from 340 children with BA who underwent LDLT. Preoperative recipient variables included age, gender, height, weight, left ventricular ejection fraction, corrected QT interval, pediatric end-stage liver disease score, alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TB), international normalized ratio, Cr, and hemoglobin. Intraoperative variables included graft weight, graft cold ischemia time, hemodynamic parameters, blood gas parameters and core temperature before the reperfusion, as well as the durations of the anhepatic period, operation, anesthesia, blood loss, urine output, blood transfusions, and fresh frozen plasma (FFP) infusion. Postoperative variables included mechanical ventilation time, intensive care unit (ICU) stay (in days), hospital stay, peak ALT, AST, Cr and TB levels within the first 24 hours after LDLT, incidence of AKI, and one-year survival rate.
Feature selection and model construction
Least absolute shrinkage and selection operator (LASSO) regression was applied to screen the initial set of candidate variables, with the aim of reducing model dimensionality and mitigating the risk of overfitting. The preliminary feature selection identified several variables with non-zero regression coefficients, including preoperative serum Cr, postoperative serum Cr, calcium ion concentration during the anhepatic phase, and lactic acid concentration during the anhepatic phase, among others. However, postoperative serum Cr constitutes a defining criterion for the clinical diagnosis of AKI. Incorporating this variable into the predictive model would therefore introduce circular reasoning and result in data leakage. Consequently, postoperative serum Cr was excluded from the feature set in all subsequent model development and evaluation procedures.
Based on the refined feature set, multiple ML models-including logistic regression, decision tree, random forest, XGBoost, gradient boosting, AdaBoost, k-nearest neighbors (KNN), multilayer perceptron (MLP) neural network, and support vector machine (SVM)-were constructed and systematically compared to identify the optimal predictive model. Finally, the SHapley Additive exPlanations (SHAP) framework was employed to quantify the contribution of each feature and to enhance the interpretability of the optimal model.
Model training and evaluation
Given the relatively small sample size in this study and the fact that a single training-test set division may lead to unstable model performance evaluation, this study did not use the fixed hold-out method, but used an internal validation strategy based on resampling. Specifically, stratified k-fold cross-validation (k = 10) was used to ensure that the proportion of AKI occurrence in each fold was consistent. Given the moderate class imbalance (29% AKI), class weighting was applied during model training for algorithms sensitive to class imbalance, such as SVM. Model performance was evaluated using accuracy, precision, recall, F1 score, and area under the curve (AUC). The mean and 95% confidence interval (CI) of each metric in cross-validation were reported, and receiver operating characteristic (ROC) curves and precision-recall curves were plotted to compare the predictive performance of different models.
Statistical analysis
Data were entered in a dual-entry mode to ensure accuracy and reliability, and all entries were proofread by a dedicated person. Predictor variables with a missing data rate greater than 10% or those with outliers were excluded, such as pre-albumin, pre-prothrombin time, and pre-C-reactive protein. After splitting the dataset, missing values (< 10%) were imputed using miss Forest fitted on the training set and applied to the validation set without refitting to avoid data leakage. If a patient had missing values in any of the key predictive variables included, the patient was excluded. All continuous variable data were tested for normality using the Shapiro-Wilk test. Variables following a normal distribution were expressed as the mean ± SD and analyzed using an independent samples t-test. Variables that do not follow a normal distribution were expressed as the median and interquartile range [M (IQR)] and analyzed using the Mann-Whitney U test. Categorical variables were represented using the results of the χ2 test in terms of case n (%). Specifically, P < 0.05 is defined as having statistical significance. A correlation analysis was performed using the Pearson or Spearman method based on the data type, and all data processing and statistical analyses were performed using R (version 4.0.3; https://www.r-project.org) and Python (version 3.12.7; https://www.python.org).
RESULTS
The results shown in Table 1 demonstrate that the pre-Cr and post-Cr levels in the AKI group were significantly higher than those in the non-AKI patient group (19.00 vs 13.00, P < 0.001; 21.00 vs 14.00, P < 0.001). In addition, the postoperative AST levels (898.60 vs 666.70, P = 0.011) and the TB (102.11 vs 80.30, P = 0.008) in the AKI group were significantly higher. The total red blood cell transfusion volume at the end of the operation (P = 0.009) and the FFP transfusion volume (P = 0.012) in the AKI group were also significantly higher. Further analysis revealed that compared with the children who did not develop AKI, the proportion of male children with postoperative AKI was significantly higher (59.60% vs 45.64%, P = 0.002). Furthermore, in the AKI group, a significant increase in lactate levels during the anhepatic phase was observed (3.30 vs 2.60, P < 0.001), which suggests that metabolic disorders during the anhepatic phase may be related to the occurrence of postoperative AKI (Table 1).
The correlation analysis of all variables is shown in the heatmap presented in Figure 1. Based on the results shown in Figure 1, we further investigated the variables that demonstrated statistically significant differences in postoperative AKI. These variables were plotted into a simplified heatmap to visualize the key factors related to postoperative AKI in an intuitive manner. Specifically, there was a strong correlation between the pre-Cr and post-Cr levels, gender, electrolytes during the anhepatic period (lactic acid, blood calcium, blood chloride, and blood sodium), pre-reperfusion and total blood transfusion volume, ICU stay days, FFP infusion volume, Post-AST, Post-TB, and Post-AKI.
Figure 1 Correlation heatmap of variables.
A: Correlation heatmap of different variables; B: Heatmap of significantly different variables. Cr: Creatinine; LAC: Lactate; FFP: Fresh frozen plasma; ICU: Intensive care unit; PRS: Postreperfusion syndrome; AST: Aspartate aminotransferase; TB: Total bilirubin; AKI: Acute kidney injury; Ca2+: Calcium ion; Na+: Sodium ion.
In this study, the predictive ability of nine ML models for AKI after pediatric LDLT was systematically evaluated by 10-fold cross-validation. The random forest model exhibited the best overall discrimination in stratified 10-fold cross-validation, with an accuracy of 0.842 (95%CI: 0.783-0.902) and the highest area (AUC = 0.875, 95%CI: 0.805-0.944). The XGBoost and Gradient Boosting models performed similarly, with AUCs of 0.848 (95%CI: 0.771-0.925) and 0.841 (95%CI: 0.764-0.917), respectively, but their recall and F1 scores were lower than those of the random forest model. AdaBoost, KNN, MLP, and SVM had moderate discrimination, with AUCs ranging from 0.810 to 0.824. In contrast, the overall performance of the logistic regression and decision tree models was weaker, especially in terms of recall and F1 values, with the decision tree model having the lowest AUC (0.734, 95%CI: 0.664-0.803). Overall, the random forest model demonstrated the most balanced and robust performance in predicting postoperative AKI (Table 2).
Figure 2 presents the features selected by the LASSO regression model along with the magnitude of their absolute coefficients. The comparison of ROC curves in Figure 3 shows that the random forest model (AUC = 0.87) is significantly better than other models, and the precision-recall curve also confirms this (Figure 3). In addition, the confusion matrix showed that the random forest model accurately identified 67 positive samples and 227 negative samples in this study, with only 46 misclassifications, thus demonstrating excellent classification ability (Figure 4). In summary, the results clearly indicate that the random forest model exhibited the best predictive performance.
Figure 4
Confusion matrix of the best prediction model.
To better explain the complex random forest model, the SHAP value analysis method was used to illustrate the degree of influence of each predictor on the risk of AKI during pediatric LDLT surgery (Figure 5). Given that postoperative Cr was excluded from the modeling in this study, we found that the preoperative Cr level contributed the most to the prediction results and was positively correlated with the outcome indicator. Other important features included lactate concentration during the anhepatic phase and peak postoperative AST, which were negatively correlated with the outcome. Meanwhile, this study found that male children were more likely to develop AKI after LDLT. The SHAP analysis results revealed the contribution of each feature to the model prediction, which helped clinicians understand the prediction mechanism of the model. Figure 6 shows the SHAP feature dependence plots of different features.
Figure 5 SHapley Additive exPlanations analysis explanation diagram of the best model.
A: The ranking of the importance of different features for acute kidney injury prediction; B: The diagram of the influence direction and intensity of different features; C: The prediction analysis of a single sample. Mean SHapley Additive exPlanations (SHAP) value is the SHAP mean/feature importance, base value is the baseline value, higher indicates higher risk, Lower indicates lower risk. Cr: Creatinine; AST: Aspartate aminotransferase; LAC: Lactate; Ca2+: Calcium ion; FFP: Fresh frozen plasma; SHAP: SHapley Additive exPlanations.
Figure 6 The influence of each independent factor on the prediction of acute kidney injury after living donor liver transplantation.
A: The influence of pre- creatinine on acute kidney injury (AKI) prediction; B: The influence of post-aspartate aminotransferase on AKI prediction; C: The influence of anhepatic phase lactate level on AKI prediction; D: The influence of anhepatic phase calcium ions on AKI prediction; E: The influence of fresh frozen plasma infusion situation on AKI prediction; F: The influence of gender on AKI prediction. SHAP: SHapley Additive exPlanations; Cr: Creatinine; AST: Aspartate aminotransferase; LAC: Lactate; Ca2+: Calcium ion; FFP: Fresh frozen plasma.
DISCUSSION
Our study demonstrates that patients who develop AKI exhibit significantly higher male gender ratio, preoperative and postoperative Cr levels, postoperative AST peak values, lactate levels during the anhepatic phase, and postoperative ICU stay duration. This adds further evidence regarding perioperative complications in children undergoing LDLT. Additionally, due to the high incidence of BA in pediatric liver transplantation and the complexity of postoperative management, its pathophysiological characteristics, along with the complexity of liver transplantation surgery, may make these patients more prone to AKI. Therefore, we focus on constructing a model to compare post-operative AKI in BA children undergoing LDLT surgery. The findings of this study demonstrate that the random forest model achieved outstanding performance in predicting AKI after pediatric liver transplantation. Its high AUC value and low misdiagnosis rate proved the great potential of ML models in clinical applications. We found that the logistic regression model performed well in terms of accuracy; however, its low recall rate may lead to missed diagnoses. Thus, when used in clinical practice, careful consideration should be given to models with higher recall rates, e.g., the XGBoost and random forest models.
In addition, the results of the SHAP analysis provided interpretability for the model, thereby enabling clinicians to clearly understand which factors have significant impacts on the occurrence of postoperative AKI. In particular, physiological parameters, e.g., serum Cr, calcium ion concentration during the anhepatic period, and lactate concentration, should be monitored closely in clinical settings. In clinical practice, the attending physician can predict the risk of postoperative AKI based on the preoperative and intraoperative indicators of liver transplant patients so as to achieve early intervention. High-performance ML technologies provide important aid in medical research[17-19]. The use of ML technology enables the analysis and screening of massive amounts of intricate clinical data and helps develop models to predict clinical events. In previous studies, logistic regression models applied to clinical data have often been limited in their ability to capture nonlinear relationships and are susceptible to outliers and multicollinearity, potentially affecting model stability and predictive performance. In parallel, a small number of studies have developed ML models to predict post-transplantation outcomes in liver transplant patients[20,21].
To our knowledge, research on AKI after pediatric liver transplantation is relatively limited, with data being particularly scarce for children with BA. Previous studies have largely focused on adult liver transplant populations. This study is expected to provide new evidence to support this field. Recent studies have shown that serum Cr levels, lactate concentration, and transfusion volume are important factors in predicting AKI[22-24]. Building on these findings, this study further investigates the predictive factors for AKI following pediatric liver transplantation. Specifically, various ML models were developed and compared based on factors such as pre-Cr levels, blood calcium concentration during the anhepatic phase, lactate, gender, FFP infusion volume, and postoperative AST levels. Previous studies have demonstrated that men and boys are more prone to AKI, which is in good agreement with the results of this study. This may be related to the physiological differences and hormone levels in men[22-24]. In addition, the findings of this study indicate that the pre-Cr levels was the most important factors in predicting AKI. Abnormal Cr levels have always been an important indicator of poor renal function recovery. Through SHAP visualization analysis, we found that lactate levels during the anhepatic phase were positively correlated with outcome indicators. Elevated lactate reflects inadequate perioperative tissue perfusion and metabolic disturbances, thereby increasing the risk of AKI to a certain extent[25,26]. Furthermore, blood calcium is crucial in terms of maintaining stable kidney function, and imbalances in calcium levels can exacerbate the formation of kidney stones[27]. The high infusion volume of FFP during surgery may cause hemodynamic instability and immune responses, resulting in insufficient kidney perfusion, which increases the risk of AKI[28].
Previous studies have demonstrated that the occurrence and progression of AKI significantly affect the survival rates of patients undergoing LDLT surgery and the grafts[29-32]. Elevated preoperative TB levels and increased intraoperative blood loss are independently associated with the development of AKI, and AKI is significantly correlated with prolonged hospital stays[33]. In addition, during LDLT surgery, most children will experience the stage of postreperfusion syndrome (PRS). Another study indicated that hypothermia before reperfusion (< 36.0 °C) and increased donor cold ischemia duration were independent predictors of PRS[34]. In a long-term follow-up study of adult liver transplantations, the occurrence of AKI after transplantation was correlated with major adverse renal events. The possibility of adverse renal events after AKI after transplantation is 2.3% in the second year and increases to 10.0% in the tenth year after transplantation[35]. Our results further confirmed the above findings.
In this study, postoperative Cr was not included in the predictive variables during the model construction process, because postoperative Cr itself is one of the diagnostic criteria for AKI. If it is used as a feature input, it may lead to information leakage and circular argumentation, thereby overestimating the predictive performance of the model. It is worth noting that even without the introduction of postoperative Cr indicators, the optimal model still achieved a high discrimination ability (AUC = 0.875), indicating that other clinical variables during the perioperative period have been able to fully reflect the risk of postoperative AKI. The results suggest that the model has the ability to identify the risk early before AKI occurs, and provides a certain reference for clinical early intervention and individualized management.
This study still has some limitations that need to be fully considered. Firstly, this study is a single-center, small-sample retrospective study, which may be subject to selection bias and potential confounding factors. Consequently, the model's performance metrics may be overestimated, which limits the generalizability of the findings to some extent. More importantly, there may be significant differences in surgical methods and perioperative management strategies among different transplant centers, and these center-specific factors may affect the incidence of postoperative AKI and the stability of the prediction model. In addition, the screening criteria and referral patterns for children with BA vary among different centers, which may lead to inconsistent baseline disease severity, nutritional status, and preoperative renal function levels, thereby introducing patient selection bias and further limiting the external applicability of the model. Finally, the absence of external validation represents an important limitation of this study. Although internal validation was conducted using cross-validation, the reported model performance may still be optimistic. Therefore, future multi-center, prospective studies with larger sample sizes and external validation cohorts are still needed to further verify the robustness of the model and its clinical application value.
CONCLUSION
This study constructed and compared the application of various ML models in predicting AKI after pediatric LDLT based on seven key variables, including pre-Cr and post-Cr levels, blood calcium and lactate levels during the anhepatic period, gender, FFP infusion volume, and postoperative AST. The results demonstrated that the random forest model outperformed the other models, with high AUC, accuracy, and recall values, highlighting its potential for widespread clinical application. The findings of this study hold significant clinical importance and are expected to enhance the prevention and treatment of AKI after liver transplantation.
de Ville de Goyet J, Baumann U, Karam V, Adam R, Nadalin S, Heaton N, Reding R, Branchereau S, Mirza D, Klempnauer JL, Fischer L, Kalicinski P, Colledan M, Lopez Santamaria M, de Kleine RH, Chardot C, Yilmaz S, Kilic M, Boillot O, di Francesco F, Polak WG, Verkade HJ; European Liver, Intestine Transplant Association. European Liver Transplant Registry: Donor and transplant surgery aspects of 16,641 liver transplantations in children.Hepatology. 2022;75:634-645.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 11][Cited by in RCA: 75][Article Influence: 18.8][Reference Citation Analysis (2)]
Knackstedt ED, Anderson SG, Anand R, Mitchell J, Arnon R, Book L, Ekong U, Elisofon SA, Furuya KN, Himes R, Jain AK, Ovchinsky N, Sundaram SS, Bucuvalas J, Danziger-Isakov L; Society of Pediatric Liver Transplantation. Cytomegalovirus prophylaxis in pediatric liver transplantation: A comparison of strategies across the Society of Pediatric Liver Transplantation (SPLIT) consortium.Am J Transplant. 2025;25:1098-1106.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 2][Cited by in RCA: 3][Article Influence: 3.0][Reference Citation Analysis (2)]
Ostermann M, Bellomo R, Burdmann EA, Doi K, Endre ZH, Goldstein SL, Kane-Gill SL, Liu KD, Prowle JR, Shaw AD, Srisawat N, Cheung M, Jadoul M, Winkelmayer WC, Kellum JA; Conference Participants. Controversies in acute kidney injury: conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Conference.Kidney Int. 2020;98:294-309.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 456][Cited by in RCA: 406][Article Influence: 67.7][Reference Citation Analysis (0)]
Sakai T, Ko JS, Crouch CE, Kumar S, Little MB, Chae MS, Ganoza A, Gómez-Salinas L, Humar A, Kim SH, Koo BN, Rodriguez G, Sirianni J, Smith NK, Song JG, Ullah A, Hendrickse A. Perioperative management of adult living donor liver transplantation: Part 1 - recipients.Clin Transplant. 2022;36:e14667.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 8][Reference Citation Analysis (0)]
Mukhtar A, Dabbous H. Modulation of splanchnic circulation: Role in perioperative management of liver transplant patients.World J Gastroenterol. 2016;22:1582-1592.
[PubMed] [DOI] [Full Text]
P-Reviewer: Paulin VS, MD, Assistant Professor, Consultant, India; Xu X, MD, PhD, Associate Professor, China S-Editor: Qu XL L-Editor: A P-Editor: Wang CH
