Prognostic Impact of Parameters of Metabolic Acidosis in Critically Ill Children with Acute Kidney Injury: A Retrospective Observational Analysis Using the PIC Database

Acute kidney injury: the experience of a tertiary center of Pediatric Nephrology

INTRODUCTION

Acute kidney injury (AKI) is an abrupt deterioration of kidney function. The incidence of pediatric AKI is increasing worldwide, both in critically and non-critically ill settings. We aimed to characterize the presentation, etiology, evolution, and outcome of AKI in pediatric patients admitted to a tertiary care center.

METHODS

We performed a retrospective observational single-center study of patients aged 29 days to 17 years and 365 days admitted to our Pediatric Nephrology Unit from January 2012 to December 2021, with the diagnosis of AKI. AKI severity was categorized according to Kidney Disease Improving Global Outcomes (KDIGO) criteria. The outcomes considered were death or sequelae (proteinuria, hypertension, or changes in renal function at 3 to 6 months follow-up assessments).

RESULTS

Forty-six patients with a median age of 13.0 (3.5-15.5) years were included. About half of the patients (n = 24, 52.2%) had an identifiable risk factor for the development of AKI. Thirteen patients (28.3%) were anuric, and all of those were categorized as AKI KDIGO stage 3 (p < 0.001). Almost one quarter (n = 10, 21.7%) of patients required renal replacement therapy. Approximately 60% of patients (n = 26) had at least one sequelae, with proteinuria being the most common (n = 15, 38.5%; median (P25-75) urinary protein-to-creatinine ratio 0.30 (0.27-0.44) mg/mg), followed by reduced glomerular filtration rate (GFR) (n = 11, 27.5%; median (P25-75) GFR 75 (62-83) mL/min/1.73 m2).

CONCLUSIONS

Pediatric AKI is associated with substantial morbidity, with potential for proteinuria development and renal function impairment and a relevant impact on long-term prognosis.

Predicting pediatric cardiac surgery-associated acute kidney injury using machine learning

BACKGROUND

Prediction of cardiac surgery-associated acute kidney injury (CS-AKI) in pediatric patients is crucial to improve outcomes and guide clinical decision-making. This study aimed to develop a supervised machine learning (ML) model for predicting moderate to severe CS-AKI at postoperative day 2 (POD2).

METHODS

This retrospective cohort study analyzed data from 402 pediatric patients who underwent cardiac surgery at a university-affiliated children's hospital, who were separated into an 80%-20% train-test split. The ML model utilized demographic, preoperative, intraoperative, and POD0 clinical and laboratory data to predict moderate to severe AKI categorized by Kidney Disease: Improving Global Outcomes (KDIGO) stage 2 or 3 at POD2. Input feature importance was assessed by SHapley Additive exPlanations (SHAP) values. Model performance was evaluated using accuracy, area under the receiver operating curve (AUROC), precision, recall, area under the precision-recall curve (AUPRC), F1-score, and Brier score.

RESULTS

Overall, 13.7% of children in the test set experienced moderate to severe AKI. The ML model achieved promising performance, with accuracy of 0.91 (95% CI: 0.82-1.00), AUROC of 0.88 (95% CI: 0.72-1.00), precision of 0.92 (95% CI: 0.70-1.00), recall of 0.63 (95% CI: 0.32-0.96), AUPRC of 0.81 (95% CI: 0.61-1.00), F1-score of 0.73 (95% CI: 0.46-0.99), and Brier score loss of 0.09 (95% CI: 0.00-0.17). The top ten most important features assessed by SHAP analyses in this model were preoperative serum creatinine, surgery duration, POD0 serum pH, POD0 lactate, cardiopulmonary bypass duration, POD0 vasoactive inotropic score, sex, POD0 hematocrit, preoperative weight, and POD0 serum creatinine.

CONCLUSIONS

A supervised ML model utilizing demographic, preoperative, intraoperative, and immediate postoperative clinical and laboratory data showed promising performance in predicting moderate to severe CS-AKI at POD2 in pediatric patients.

Splenic SUMO1 controls systemic inflammation in experimental sepsis

Introduction

The recent discovery of TAK981(Subasumstat), the first-in-class selective inhibitor of SUMOylation, enables new immune treatments. TAK981 is already in clinical trials to potentiate immunotherapy in metastatic tumors and hematologic malignancies. Cancer patients have more than ten times higher risk of infections, but the effects of TAK981 in sepsis are unknown and previous studies on SUMO in infections are conflicting.

Methods

We used TAK981 in two sepsis models; polymicrobial peritonitis (CLP) and LPS endotoxemia. Splenectomy was done in both models to study the role of spleen. Western blotting of SUMO-conjugated proteins in spleen lysates was done. Global SUMO1 and SUMO3 knockout mice were used to study the specific SUMO regulation of inflammation in LPS endotoxemia. Splenocytes adoptive transfer was done from SUMO knockouts to wild type mice to study the role of spleen SUMOylation in experimental sepsis.

Results and discussion

Here, we report that inhibition of SUMOylation with TAK981 improved survival in mild polymicrobial peritonitis by enhancing innate immune responses and peritoneal bacterial clearance. Thus, we focused on the effects of TAK981 on the immune responses to bacterial endotoxin, showing that TAK981 enhanced early TNFα production but did not affect the resolution of inflammation. Splenectomy decreased serum TNFα levels by nearly 60% and TAK981-induced TNFα responses. In the spleen, endotoxemia induced a distinct temporal and substrate specificity for SUMO1 and SUMO2/3, and both were inhibited by TAK981. Global genetic depletion of SUMO1, but not SUMO3, enhanced TNFα production and metabolic acidosis. The transfer of SUMO1-null, but not wild-type, splenocytes into splenectomized wild-type mice exacerbated TNFα production and metabolic acidosis in endotoxemia.

Conclusion

These results suggest that specific regulation of splenic SUMO1 can modulate immune and metabolic responses to bacterial infection.

Association between serum anion gap and all-cause mortality in patients with acute myocardial infarction: A retrospective study based on MIMIC-IV database

Background

Although previous studies have reported that many biomarkers can determine the prognosis of patients with acute myocardial infarction (AMI), serum anion gap (AG) has not been well studied. We aimed to investigate the association between serum AG and mortality in patients with AMI.

Methods

Adult patients first admitted to the ICU and diagnosed with AMI from 2008 to 2019 in the MIMIC-IV database were included. Patients were divided into the survival and non-survival groups based on 30-day and 90-day outcomes. According to the AG value (15.12 mmol/L) with a hazard ratio of 1 in the restricted cubic spline (RCS) analysis, patients were further divided into high and low AG groups. The Kaplan-Meier survival curve was plotted, and all-cause mortality was compared between the high and low groups using the log-rank test. Multivariate Cox regression analysis and RCS analysis were constructed to assess the relationship between AG and recent all-cause mortality in patients with AMI.

Results

4446 patients were enrolled. The 30-day and 90-day mortality rates in the high AG group (25.53%, 31.75%) were higher than that in the low AG group (9.73%, 14.01%, P < 0.001) independently. The Kaplan-Meier curve showed that the 30-day and 90-day cumulative survival rates were lower in the high AG group than that in the low AG group (P < 0.001). RCS analysis showed that there was a non-linear relationship between AG and the risk of 90-day all-cause mortality in patients with AMI (χ2 = 18.680 P < 0.001). When AG was 15.12 mmol/L, its HR was about 1. Multivariable Cox regression analysis confirmed that increased AG was associated with higher 30-day and 90-day mortality.

Conclusion

Elevated serum AG (≥15.12 mmol/L) is an independent predictor for short-term mortality in patients with AMI, and it may provide a basis for clinicians to identify patients with poor prognosis as early as possible.

Construction and validation of prognostic models in critically Ill patients with sepsis-associated acute kidney injury: interpretable machine learning approach

BACKGROUND

Acute kidney injury (AKI) is a common complication in critically ill patients with sepsis and is often associated with a poor prognosis. We aimed to construct and validate an interpretable prognostic prediction model for patients with sepsis-associated AKI (S-AKI) using machine learning (ML) methods.

METHODS

Data on the training cohort were collected from the Medical Information Mart for Intensive Care IV database version 2.2 to build the model, and data of patients were extracted from Hangzhou First People's Hospital Affiliated to Zhejiang University School of Medicine for external validation of model. Predictors of mortality were identified using Recursive Feature Elimination (RFE). Then, random forest, extreme gradient boosting (XGBoost), multilayer perceptron classifier, support vector classifier, and logistic regression were used to establish a prognosis prediction model for 7, 14, and 28 days after intensive care unit (ICU) admission, respectively. Prediction performance was assessed using the receiver operating characteristic (ROC) curve and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) were used to interpret the ML models.

RESULTS

In total, 2599 patients with S-AKI were included in the analysis. Forty variables were selected for the model development. According to the areas under the ROC curve (AUC) and DCA results for the training cohort, XGBoost model exhibited excellent performance with F1 Score of 0.847, 0.715, 0.765 and AUC (95% CI) of 0.91 (0.90, 0.92), 0.78 (0.76, 0.80), and 0.83 (0.81, 0.85) in 7 days, 14 days and 28 days group, respectively. It also demonstrated excellent discrimination in the external validation cohort. Its AUC (95% CI) was 0.81 (0.79, 0.83), 0.75 (0.73, 0.77), 0.79 (0.77, 0.81) in 7 days, 14 days and 28 days group, respectively. SHAP-based summary plot and force plot were used to interpret the XGBoost model globally and locally.

CONCLUSIONS

ML is a reliable tool for predicting the prognosis of patients with S-AKI. SHAP methods were used to explain intrinsic information of the XGBoost model, which may prove clinically useful and help clinicians tailor precise management.

Lactate acid level and prognosis of neonatal necrotizing enterocolitis: a retrospective cohort study based on pediatric-specific critical care database

OBJECTIVE

To investigate the relationship between lactate acid level and hospitalization mortality in neonatal necrotizing enterocolitis (NEC).

METHOD

Paediatric-specific critical care database collected clinical data from the intensive care unit of Children's Hospital Affiliated to Zhejiang University Medical College from 2010 to 2018. Clinical and laboratory examination information of NEC patients was collected and divided into the death group and discharge group to find out the risk factors affecting the prognosis through univariate and multivariate analysis.

RESULTS

Among 104 NEC neonates, the admission age was 7.5 days and the weight was 2.03 kg. Comparing the death group with the discharge group, there were significant differences in therapeutic regimen, pH, serum albumin, total protein, creatinine and lactate acid. Multivariate and threshold effect analysis showed that lactate acid had a linear correlation with hospital mortality, and newborns who died in the hospital had much higher lactate levels than those who were discharged. The mortality of NEC newborns increased by 40-45% for every 1 mmol/L increase in lactate acid level.

CONCLUSIONS

There was a correlation between lactate acid level and hospital mortality in newborns with NEC, and lactate acid level was an important index to evaluate the prognosis of NEC.

Acute Kidney Injury and Hemolytic Uremic Syndrome in Severe Pneumococcal Pneumonia—A Retrospective Analysis in Pediatric Intensive Care Unit

Objectives: This study aimed to evaluate the prevalence of acute kidney injury (AKI) and hemolytic uremic syndrome (HUS) in severe pediatric pneumonia due to Streptococcus pneumoniae and to identify factors associated with AKI and HUS in these patients.

Methods: We retrospectively analyzed pediatric patients who were admitted to our pediatric intensive care unit due to severe pneumococcal pneumonia between 2013 and 2019.

Results: Forty-two patients with a median age of 4.3 years were included. Among these patients, 14 (33.3%) developed AKI, including seven (16.7%) stage 1, two (4.8%) stage 2, and five (11.9%) stage 3 AKI. Features of HUS were present in all of the patients with stage 3 AKI, and four required renal replacement therapy (RRT), with a median duration of 10.5 days (range 3 to 16 days). All patients with HUS required mechanical ventilation and inotropic supports. Patients with lower leukocyte and platelet counts, serum sodium and bicarbonate levels, positive urine dipstick (heme or protein ≥ 2 + ), and presence of bacteremia were associated with stage 2 and 3 AKI.

Conclusions: Pediatricians should be aware of the relatively high prevalence of kidney involvement in severe pneumococcal pneumonia, with one-third having AKI and 11.9% developing HUS. Majority (80%) of HUS patients required RRT. Positive urine dipstick, serum sodium, and bicarbonate at presentation, which can be measured in point-of-care tests, may potentially be useful as quick tests to stratify the risks of moderate-to-severe AKI.

Normal Anion Gap Metabolic Acidosis in Pediatric Acute Diarrhea: A Menace or an Innocent Bystander?

How to cite this article: Bhowmick R. Normal Anion Gap Metabolic Acidosis in Pediatric Acute Diarrhea: A Menace or an Innocent Bystander? Indian J Crit Care Med 2022;26(12):1235-1236.

Abnormal magnesium levels and their impact on death and acute kidney injury in critically ill children

BACKGROUND

The prevalence of magnesium imbalance in critically ill children is very high. However, its significance in the development of acute kidney injury (AKI) and mortality remains unknown.

METHODS

In this retrospective observational study from 2010 to 2018, the pediatric-specific intensive care database was analyzed. We included critically ill children aged > 3 months and those without chronic kidney disease. Patients were diagnosed with AKI, according to the Kidney Disease Improving Global Outcomes (KDIGO) study. We calculated the initial corrected magnesium levels (cMg) within 24 h and used a spline regression model to evaluate the cut-off values for cMg. We analyzed 28-day mortality and its association with AKI. The interaction between AKI and magnesium imbalance was evaluated.

RESULTS

The study included 3,669 children, of whom 105 died within 28 days, while 1,823 were diagnosed with AKI. The cut-off values for cMg were 0.72 and 0.94 mmol/L. Both hypermagnesemia and hypomagnesemia were associated with 28-day mortality (odds ratio [OR] = 2.99, 95% confidence interval [CI] = 1.89-4.71, p < 0.001; OR = 2.80, 95% CI = 1.60-4.89, p < 0.001). Hypermagnesemia was associated with AKI (OR = 1.52, 95% CI = 1.27-1.82, p < 0.001), while neither hypermagnesemia nor hypomagnesemia interacted with the AKI stage on the 28-day mortality.

CONCLUSIONS

Abnormal magnesium levels were associated with 28-day mortality in critically ill children. AKI and hypermagnesemia had a strong association. "A higher resolution version of the Graphical abstract is available as Supplementary information".

Predicting Risk of Mortality in Pediatric ICU Based on Ensemble Step-Wise Feature Selection

Background. Prediction of mortality risk in intensive care units (ICU) is an important task. Data-driven methods such as scoring systems, machine learning methods, and deep learning methods have been investigated for a long time. However, few data-driven methods are specially developed for pediatric ICU. In this paper, we aim to amend this gap-build a simple yet effective linear machine learning model from a number of hand-crafted features for mortality prediction in pediatric ICU.Methods. We use a recently released publicly available pediatric ICU dataset named pediatric intensive care (PIC) from Children's Hospital of Zhejiang University School of Medicine in China. Unlike previous sophisticated machine learning methods, we want our method to keep simple that can be easily understood by clinical staffs. Thus, an ensemble step-wise feature ranking and selection method is proposed to select a small subset of effective features from the entire feature set. A logistic regression classifier is built upon selected features for mortality prediction.Results. The final predictive linear model with 11 features achieves a 0.7531 ROC-AUC score on the hold-out test set, which is comparable with a logistic regression classifier using all 397 features (0.7610 ROC-AUC score) and is higher than the existing well known pediatric mortality risk scorer PRISM III (0.6895 ROC-AUC score).Conclusions. Our method improves feature ranking and selection by utilizing an ensemble method while keeping a simple linear form of the predictive model and therefore achieves better generalizability and performance on mortality prediction in pediatric ICU.