Modifying the Renal Angina Index for Predicting AKI and Related Adverse Outcomes in Pediatric Heart Surgery

Prediction of cardiac surgery associated acute kidney injury using response to loop diuretic and urine neutrophil gelatinase associated lipocalin

BACKGROUND

Cardiac surgery associated acute kidney injury (CS-AKI) is common. Urine response to loop diuretic and urine neutrophil gelatinase associated lipocalin (uNGAL) are separately associated with CS-AKI. We aimed to determine whether urine response to loop diuretic and uNGAL together were associated with postoperative day 2-4 CS-AKI.

METHODS

Two-center prospective observational study (ages 0-18 years). uNGAL (8-12 h after admission) (ng/mL) and urine response to loop diuretic (6 h for bolus furosemide and 12 h for infusion bumetanide) (mL/kg/hr) were measured. All diuretic doses were converted to furosemide equivalents. The primary outcome was day 2-4 CS-AKI. Patients were sub-phenotyped using a priori cutoffs (uNGAL +  ≥ 100 ng/mL and UOP +  < 1.5 mL/kg/hr) and optimal cutoffs (uNGAL +  ≥ 127 ng/mL and UOP +  ≤ 0.79 mL/kg/hr): 1) uNGAL-/UOP-, 2) uNGAL-/UOP + , 3) uNGAL + /UOP-, and 4) uNGAL + /UOP + . Multivariable regression was used to assess the association of uNGAL, UOP and each sub-phenotype with outcomes.

RESULTS

476 patients were included. CS-AKI occurred in 52 (10.9%). uNGAL was associated with 2.59-fold greater odds (95%CI: 1.52-4.41) of CS-AKI. UOP was not associated with CS-AKI. Compared with uNGAL + alone, uNGAL + /UOP + improved prediction of CS-AKI using a priori and optimal cutoffs respectively (AUC 0.70 vs. 0.75). Both uNGAL + /UOP + (IQR OR:4.63, 95%CI: 1.74-12.32) and uNGAL + /UOP- (IQR OR:5.94, 95%CI: 2.09-16.84) were associated with CS-AKI when compared with uNGAL-/UOP-.

CONCLUSIONS

uNGAL is associated with CS-AKI. The sub-phenotype association was largely driven by uNGAL. Future studies standardizing diuretic dose and timing may be needed to refine the combined performance for clinical decision making.

Biomarker-based acute kidney injury sub-phenotypes refine risk assessment in children undergoing cardiac surgery

BACKGROUND

Pediatric cardiac surgery-associated acute kidney injury (CS-AKI) is common with variable association with outcomes, possibly because transient serum creatinine (SCr) elevations are unrelated to kidney disease. Sub-phenotypes of CS-AKI with biomarker integration may provide prognostic enrichment. This study aims to determine if combining early postoperative urine neutrophil gelatinase-associated lipocalin (uNGAL) and SCr into sub-phenotypes strengthens associations with AKI and outcomes. We hypothesized that patients with early subclinical (uNGAL + , SCr -) or damage (uNGAL + , SCr +) CS-AKI would have more postoperative day 2-4 KDIGO-defined AKI and worse clinical outcomes than patients with early functional AKI (uNGAL - , SCr +).

METHODS

Two-center prospective observational study evaluating combinations of early uNGAL (8-12 h from ICU admission, ≥ 150 ng/mL) and early postoperative (≤ 8 h of admission) KDIGO SCr-defined AKI to predict CS-AKI on postoperative days (POD) 2-4. Four CS-AKI phenotypes were derived (uNGAL - /SCr - ; uNGAL + /SCr - ; uNGAL - /SCr + and uNGAL + /SCr +). The primary outcome was POD2-4 KDIGO SCr-defined CS-AKI. Secondary outcomes included ventilator and intensive care unit-free days (maximum 28).

RESULTS

Four hundred seventy-six patients (median age 4.8 [IQR 1.4-30.4] months, 39% female) were included. POD2-4 AKI occurred in 44 (9.2%). 27% were uNGAL + /SCr - and 0.4% (n = 2) uNGAL + /SCr + . The adjusted odds of POD2-4 AKI was ninefold higher (aOR: 9.09, 95%CI: 3.84-21.53) in uNGAL + /SCr - when compared to uNGAL - /SCr - . uNGAL + /SCr - was associated with fewer ventilator-free (aOR: 0.30, 95%CI: 0.19-0.48) and ICU-free days (aOR: 0.41, 95%CI: 0.26-0.66) when compared to uNGAL - /SCr - .

CONCLUSION

Early postoperative uNGAL, regardless of SCr elevation, refines risk assessment for pediatric POD2-4 CS-AKI and associated morbidity, enabling earlier AKI identification and prognostics.

Diuretic response after neonatal cardiac surgery: a report from the NEPHRON collaborative

BACKGROUND

Multicenter early diuretic response (DR) analysis of single furosemide dosing following neonatal cardiac surgery is lacking to inform whether early DR predicts adverse clinical outcomes.

METHODS

We performed a retrospective cohort study utilizing data from the NEPHRON registry. Random forest machine learning generated receiver operating characteristic-area under the curve (ROC-AUC) and odds ratios for mechanical ventilation (MV) and respiratory support (RS). Prolonged MV and RS were defined using ≥ 90th percentile of observed/expected ratios. Secondary outcomes were prolonged CICU and hospital length of stay (LOS) and kidney failure (stage III acute kidney injury (AKI), peritoneal dialysis, and/or continuous kidney replacement therapy on postoperative day three) assessed using covariate-adjusted ROC-AUC curves.

RESULTS

A total of 782 children were included. Cumulative urine output (UOP) metrics were lower in prolonged MV and RS patients, but DR poorly predicted prolonged MV (highest AUC 0.611, OR 0.98, sensitivity 0.67, specificity 0.53, p = 0.006, 95% OR CI 0.96-0.99 for cumulative 6-h UOP) and RS (highest AUC 0.674, OR 0.94, sensitivity 0.75, specificity 0.54, p < 0.001, 95% CI 0.91-0.97 UOP between 3 and 6 h). Secondary outcome results were similar. DR had fair discrimination for kidney failure (AUC 0.703, OR 0.94, sensitivity 0.63, specificity 0.71, 95% OR CI 0.91-0.98, p < 0.001, cumulative 6-h UOP).

CONCLUSIONS

Early DR poorly discriminated patients with prolonged MV, RS, and LOS in this cohort, though it may identify severe postoperative AKI phenotype. Future work is warranted to determine if early DR or late postoperative DR later, in combination with other AKI metrics, may identify a higher-risk phenotype.

Dialysis for paediatric acute kidney injury in Cape Town, South Africa

BACKGROUND

Dialysis is lifesaving for acute kidney injury (AKI), but access is poor in less resourced settings. A "peritoneal dialysis (PD) first" policy for paediatric AKI is more feasible than haemodialysis in low-resource settings.

METHODS

Retrospective review of modalities and outcomes of children dialysed acutely at Red Cross War Memorial Children's Hospital between 1998 and 2020.

RESULTS

Of the 593 children with AKI who received dialysis, 463 (78.1%) received PD first. Median age was 9.0 (range 0.03-219.3; IQR 13.0-69.6) months; 57.6% were < 1 year old. Weights ranged from 0.9 to 2.0 kg (median 7.0 kg, IQR 3.0-16.0 kg); 38.6% were < 5 kg. PD was used more in younger children compared to extracorporeal dialysis (ECD), with median ages 6.4 (IQR 0.9-30.4) vs. 73.9 (IQR 17.5-113.9) months, respectively (p = 0.001). PD was performed with Seldinger soft catheters (n = 480/578, 83%), predominantly inserted by paediatricians at the bedside (n = 412/490, 84.1%). Complications occurred in 127/560 (22.7%) children receiving PD. Overall, 314/542 (57.8%) children survived. Survival was significantly lower in neonates (< 1 month old, 47.5%) and infants (1-12 months old, 49.2%) compared with older children (> 1 year old, 70.4%, p < 0.0001). Survival was superior in the ECD (75.4%) than in the PD group (55.6%, p = 0.002).

CONCLUSIONS

"PD First for Paediatric AKI" is a valuable therapeutic approach for children with AKI. It is feasible in low-resourced settings where bedside PD catheter insertion can be safely taught and is an acceptable dialysis modality, especially in settings where children with AKI would otherwise not survive.

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.

A proposed framework for advancing acute kidney injury risk stratification and diagnosis in children: a report from the 26th Acute Disease Quality Initiative (ADQI) conference

Acute kidney injury (AKI) in children is associated with increased morbidity, reduced health-related quality of life, greater resource utilization, and higher mortality. Improvements in the timeliness and precision of AKI diagnosis in children are needed. In this report, we highlight existing, novel, and on-the-horizon diagnostic and risk-stratification tools for pediatric AKI, and outline opportunities for integration into clinical practice. We also summarize pediatric-specific high-risk diagnoses and exposures for AKI, as well as the potential role of real-time risk stratification and clinical decision support to improve outcomes. Lastly, the key characteristics of important pediatric AKI phenotypes will be outlined. Throughout, we identify key knowledge gaps, which represent prioritized areas of focus for future research that will facilitate a comprehensive, timely and personalized approach to pediatric AKI diagnosis and management.

Real-Time Acute Kidney Injury Risk Stratification-Biomarker Directed Fluid Management Improves Outcomes in Critically Ill Children and Young Adults

Introduction

Critically ill admitted patients are at high risk of acute kidney injury (AKI). The renal angina index (RAI) and urinary biomarker neutrophil gelatinase-associated lipocalin (uNGAL) can aid in AKI risk assessment. We implemented the Trial in AKI using NGAL and Fluid Overload to optimize CRRT Use (TAKING FOCUS 2; TF2) to personalize fluid management and continuous renal replacement therapy (CRRT) initiation based on AKI risk and patient fluid accumulation. We compared outcomes pre-TF2 and post-TF2 initiation.

Methods

Patients admitted from July 2017 were followed-up prospectively with the following: (i) an automated RAI result at 12 hours of admission, (ii) a conditional uNGAL order for RAI ≥8, and (iii) a CRRT initiation goal at 10% to 15% weight-based fluid accumulation.

Results

A total of 286 patients comprised 304 intensive care unit (ICU) RAI+ admissions; 178 patients received CRRT over the observation period (2014-2021). Median time from ICU admission to CRRT initiation was 2 days shorter (P < 0.002), and ≥15% pre-CRRT fluid accumulation rate was lower in the TF2 era (P < 0.02). TF2 ICU length of stay (LOS) after CRRT discontinuation and total ICU LOS were 6 and 11 days shorter for CRRT survivors (both P < 0.02). Survival rates to ICU discharge after CRRT discontinuation were higher in the TF2 era (P = 0.001). These associations persisted in each TF2 year; we estimate a conservative $12,500 health care cost savings per CRRT patient treated after TF2 implementation.

Conclusion

We suggest that automated clinical decision support (CDS) combining risk stratification and AKI biomarker assessment can produce durable reductions in pediatric CRRT patient morbidity.

Predicting acute kidney injury with an artificial intelligence-driven model in a pediatric cardiac intensive care unit

BACKGROUND

Acute kidney injury (AKI) is among the most common complications following cardiac surgery in adult and pediatric patients, significantly affecting morbidity and mortality. Artificial Intelligence (AI) with Machine Learning (ML) can be used to predict outcomes. AKI diagnosis anticipation may be an ideal target of these methods. The scope of the study is building a Machine Learning (ML) train model with Random Forest (RF) algorithm, based on electronic health record (EHR) data, able to forecast AKI continuously after 48 h in post-cardiac surgery children, and to test its performance. Four hundred nineteen consecutive patients out of 1115 hospital admissions were enrolled in a single-center retrospective study. Patients were younger than 18 years and admitted from August 2018 to February 2020 in a pediatric cardiac intensive care unit (PCICU) undergoing cardiac surgery, invasive procedure (hemodynamic studies), and medical conditions with complete EHR records and discharged after 48 h or more.

RESULTS

Thirty-six variables were selected to build the algorithm according to commonly described cardiac surgery-associated AKI clinical predictors. We evaluated different models for different outcomes: binary AKI (no AKI vs. AKI), severe AKI (no-mild vs severe AKI), and multiclass classification (maximum AKI and the most frequent level of AKI, mode AKI). The algorithm performance was assessed with the area under the curve receiver operating characteristics (AUC ROC) for binary classification, with accuracy and K for multiclass classification. AUC ROC for binary AKI was 0.93 (95% CI 0.92-0.94), and for severe AKI was 0.99 (95% CI 0.98-1). Mode AKI accuracy was 0.95, and K was 0.80 (95% CI 0.94-0.96); maximum AKI accuracy was 0.92, and K was 0.71 (95% CI 0.91-0.93). The importance matrix plot demonstrated creatinine, basal creatinine, platelets count, adrenaline support, and lactate dehydrogenase for binary AKI with the addition of cardiopulmonary bypass duration for severe AKI as the most relevant variables of the model.

CONCLUSIONS

We validated a ML model to detect AKI occurring after 48 h in a retrospective observational study that could help clinicians in individuating patients at risk of AKI, in which a preventive strategy can be determinant to improve the occurrence of renal dysfunction.

Patient level factors increase risk of acute kidney disease in hospitalized children with acute kidney injury

BACKGROUND

Studies in adults have shown that persistent kidney dysfunction ≥7-90 days following acute kidney injury (AKI), termed acute kidney disease (AKD), increases chronic kidney disease (CKD) and mortality risk. Little is known about the factors associated with the transition of AKI to AKD and the impact of AKD on outcomes in children. The aim of this study is to evaluate risk factors for progression of AKI to AKD in hospitalized children and to determine if AKD is a risk factor for CKD.

METHODS

Retrospective cohort study of children age ≤18 years admitted with AKI to all pediatric units at a single tertiary-care children's hospital between 2015 and 2019. Exclusion criteria included insufficient serum creatinine values to evaluate for AKD, chronic dialysis, or previous kidney transplant.

RESULTS

A total of 528 children with AKI were included in the study. There were 297 (56.3%) hospitalized AKI survivors who developed AKD. Among children with AKD, 45.5% developed CKD compared to 18.7% in the group without AKD (OR 4.0, 95% CI 2.1-7.4, p-value <0.001 using multivariable logistic regression analysis including other covariates). Multivariable logistic regression model identified age at AKI diagnosis, PCICU and NICU admission, prematurity, malignancy, bone marrow transplant, previous AKI, mechanical ventilation, AKI stage, duration of kidney injury, and need for kidney replacement therapy during day 1-7 as risk factors for AKD after AKI.

CONCLUSIONS

AKD is common among hospitalized children with AKI and multiple risk factors are associated with AKD. Children that progress from AKI to AKD are at higher risk of developing CKD. A higher resolution version of the Graphical abstract is available as Supplementary information.

Subphenotypes of acute kidney injury in children

PURPOSE OF REVIEW

The purpose of this review is to describe acute kidney injury (AKI) phenotypes in children.

RECENT FINDINGS

AKI is a heterogenous disease that imposes significant morbidity and mortality on critically ill and noncritically ill patients across the age spectrum. As our understanding of AKI and its association with outcomes has improved, it is becoming increasingly apparent that there are distinct AKI subphenotypes that vary by cause or associated conditions. We have also learned that severity, duration, and repeated episodes of AKI impact outcomes, and that integration of novel urinary biomarkers of tubular injury can also reveal unique subphenotypes of AKI that may not be otherwise readily apparent.

SUMMARY

Studies that further delineate these unique AKI subphenotypes are needed to better understand the impact of AKI in children. Further delineation of these phenotypes has both prognostic and therapeutic implications.

Improving acute kidney injury diagnostic precision using biomarkers

Acute kidney injury (AKI) is common in hospitalized patients of all ages and is associated with significant morbidity and mortality. Accurate prediction and early identification of AKI is of utmost importance because no therapy exists to mitigate AKI once it has occurred. Yet, serum creatinine lacks adequate sensitivity and specificity, and quantification of urine output is challenging in incontinent children without indwelling bladder catheters. Integration of clinically available biomarkers have the potential to delineate unique AKI phenotypes that could have important prognostic and therapeutic implications. Plasma Cystatin C, urine neutrophil gelatinase associated lipocalin (NGAL) and the urinary product of tissue inhibitor metalloproteinase (TIMP-2) and insulin growth factor binding protein-7 (IGFBP7) are clinically available. These biomarkers have been studied in heterogenous populations across the age spectrum and in a variety of clinical settings for prediction of AKI. The purpose of this review is to describe and discuss the clinically available AKI biomarkers including how they have been used to delineate AKI phenotypes.

Filtering Down to Risks and Solutions: Risk Factors and Stratification After Pediatric Cardiac Surgery

Acute kidney injury after cardiac surgery (CS-AKI) is common in neonatal and pediatric populations and is a risk factor for poor outcomes, such as mortality and increased hospital resource utilization. This review presents a summary of CS-AKI risk factors, integration of biomarkers, and the need to improve risk stratification for targeting future clinical trials. To date, studies examining CS-AKI risk factors cannot be generalized easily owing to variability in patient age, surgical complexity or population, AKI definition, and center-specific practices. However, certain risk associations, such as younger age at surgery, history of prematurity, cardiopulmonary bypass time, and surgical complexity, have been identified across multiple, but not all, studies. CS-AKI appears to have different severity and duration phenotypes, and serum creatinine is limited in its ability to identify CS-AKI early and predict CS-AKI course. Treatment strategies are largely supportive, and efforts are ongoing to use biomarkers and clinical features to risk-stratify patients, which in turn may facilitate differential CS-AKI phenotyping and management with supportive care bundles, clinical decision support techniques, and modulation of modifiable risk factors.