Acute kidney injury (AKI) is commonly present in critically ill patients with sepsis. Early prediction of short-term reversibility of AKI is beneficial to risk stratification and clinical treatment decision. The study sought to use machine learning methods to discriminate between transient and persistent sepsis-associated AKI. Septic patients who developed AKI within the first 48 h after ICU admission were identified from the Medical Information Mart for Intensive Care III database. AKI was classified as transient or persistent according to the Acute Disease Quality Initiative workgroup consensus. Five prediction models using logistic regression, random forest, support vector machine, artificial neural network and extreme gradient boosting were constructed, and their performance was evaluated by out-of-sample testing. A simplified risk prediction model was also derived based on logistic regression and features selected by machine learning algorithms. A total of 5984 septic patients with AKI were included, 3805 (63.6%) of whom developed persistent AKI. The artificial neural network and logistic regression models achieved the highest area under the receiver operating characteristic curve (AUC) among the five machine learning models (0.76, 95% confidence interval [CI] 0.74–0.78). The simplified 14-variable model showed adequate discrimination, with the AUC being 0.76 (95% CI 0.73–0.78). At the optimal cutoff of 0.63, the sensitivity and specificity of the simplified model were 63% and 76% respectively. In conclusion, a machine learning-based simplified prediction model including routine clinical variables could be used to differentiate between transient and persistent AKI in critically ill septic patients. An easy-to-use risk calculator can promote its widespread application in daily clinical practice.

BackgroundSepsis-associated acute kidney injury (SA-AKI) is common in critically ill patients, which is associated with significantly increased mortality. Existing mortality prediction tools showed insufficient predictive power or failed to reflect patients' dynamic clinical evolution. Therefore, the study aimed to develop and validate machine learning-based models for real-time mortality prediction in critically ill patients with SA-AKI.MethodsThe multi-center retrospective study included patients from two distinct databases. A total of 12,132 SA-AKI patients from the Medical Information Mart for Intensive Care IV (MIMIC-IV) were randomly allocated to the training, validation, and internal test sets. An additional 3,741 patients from the eICU Collaborative Research Database (eICU-CRD) served as an external test set. For every 12 h during the ICU stays, the state-of-the-art eXtreme Gradient Boosting (XGBoost) algorithm was used to predict the risk of in-hospital death in the following 48, 72, and 120 h and in the first 28 days after ICU admission. Area under the receiver operating characteristic curves (AUCs) were calculated to evaluate the models' performance.ResultsThe XGBoost models, based on routine clinical variables updated every 12 h, showed better performance in mortality prediction than the SOFA score and SAPS-II. The AUCs of the XGBoost models for mortality over different time periods ranged from 0.848 to 0.804 in the internal test set and from 0.818 to 0.748 in the external test set. The shapley additive explanation method provided interpretability for the XGBoost models, which improved the understanding of the association between the predictor variables and future mortality.ConclusionsThe interpretable machine learning XGBoost models showed promising performance in real-time mortality prediction in critically ill patients with SA-AKI, which are useful tools for early identification of high-risk patients and timely clinical interventions.

ObjectiveThe epidemiology and outcomes of acute kidney disease (AKD) after acute kidney injury (AKI) in hospitalized children are poorly described. The aim of this study is to investigate the prevalence, predictive factors, and clinical outcomes of AKD in hospitalized children with AKI.MethodsChildren (1 month–18 years) with AKI during hospitalization in the Second Xiangya Hospital from January 2015 to December 2020 were identified. AKD was defined based on the consensus report of the Acute Disease Quality Initiative 16 workgroup. The endpoints include adverse outcomes in 30 and 90 days. Multivariable logistic regression analyses were used to estimate the odds ratio of 30- and 90-day adverse outcomes associated with AKD and identify the risk factors of AKD.ResultsAKD was developed in 42.3% (419/990) of the study patients, with 186 in AKD stage 1, 107 in AKD stage 2, and 126 in AKD stage 3. Pediatric patients with AKD stages 2–3 had significantly higher rates of developing 30- and 90-day adverse outcomes than those with AKD stage 0 and 1. The adjusted odds ratio of AKD stage 2–3 was 12.18 (95% confidence interval (CI), 7.38 - 20.09) for 30-day adverse outcomes and decreased to 2.49 (95% CI, 1.26 - 4.91) for 90-day adverse outcomes. AKI stages 2 and 3, as well as glomerulonephritis, were the only predictive factors for AKD stage 2–3.ConclusionAKD is frequent among hospitalized pediatric AKI patients. AKD stage 2–3 represents a high-risk subpopulation among pediatric AKI survivors and is independently associated with 30- and 90-day adverse outcomes. Awareness of the potential risks associated with AKD stage 2–3 and its risk factors may help improve outcomes through careful monitoring and timely intervention.

Acute kidney injury (AKI) correlates with increased health-care costs and poor outcomes in older adults. However, there is no good scoring system to predict mortality within 30-day, 1-year after AKI in older adults. We performed a retrospective analysis screening data of 53,944 hospitalized elderly patients (age > 65 years) from multi-centers in China. 944 patients with AKI (acute kidney disease) were included and followed up for 1 year. Multivariable regression analysis was used for developing scoring models in the test group (a randomly 70% of all the patients). The established models have been verified in the validation group (a randomly 30% of all the patients). Model 1 that consisted of the risk factors for death within 30 days after AKI had accurate discrimination (The area under the receiver operating characteristic curves, AUROC: 0.90 (95% CI 0.875–0.932)) in the test group, and performed well in the validation groups (AUROC: 0.907 (95% CI 0.865–0.949)). The scoring formula of all-cause death within 1 year (model 2) is a seven-variable model including AKI type, solid tumor, renal replacement therapy, acute myocardial infarction, mechanical ventilation, the number of organ failures, and proteinuria. The area under the receiver operating characteristic (AUROC) curves of model 2 was > 0.80 both in the test and validation groups. Our newly established risk models can well predict the risk of all-cause death in older hospitalized AKI patients within 30 days or 1 year.

No abstract