Visual Abstract
Keywords: AKI, pediatric nephrology
Abstract
Key Points
Using indwelling urinary catheters, urine output (UO) shows good performance in neonates and younger children.
Using higher UO thresholds in neonates post-cardiac surgery improves discriminatory capacity for outcomes compared to neonatal Kidney Disease Improving Global Outcomes.
In younger children (1–24 months), higher UO thresholds were not better than the adult Kidney Disease Improving Global Outcomes criteria.
Background
Pediatric AKI is associated with significant morbidity and mortality, yet a precise definition, especially concerning urine output (UO) thresholds, remains unproven. We evaluate UO thresholds for AKI in neonates and children aged 1–24 months with indwelling urinary catheters undergoing cardiac surgery.
Methods
A 6-year prospective cohort study (2018–2023) after cardiac surgery was conducted at a reference center in Brazil. All patients had indwelling urinary catheters up to 48 hours after surgery and at least two serum creatinine measurements, including one before surgery. The main objective of this study was to determine the optimal UO thresholds for AKI definition and staging in neonates and younger children compared with the currently used criteria—neonatal and adult Kidney Disease Improving Global Outcomes (KDIGO) definitions. The outcome was a composite of severe AKI (stage 3 AKI diagnosed by the serum creatinine criterion only), KRT, or hospital mortality.
Results
The study included 1024 patients: 253 in the neonatal group and 772 in the younger children group. In both groups, the lowest UO at 24 hours as a continuous variable had good discriminatory capacity for the composite outcome (area under the curve-receiver operating characteristic 0.75 [95% confidence interval, 0.70 to 0.81] and 0.74 [95% confidence interval, 0.68 to 0.79]). In neonates, the best thresholds were 3.0, 2.0, and 1.0 ml/kg per hour, and in younger children, the thresholds were 1.8, 1.0, and 0.5 ml/kg per hour. These values were used for modified AKI staging for each age group. In neonates, this modified criterion was associated with the best discriminatory capacity (area under the curve-receiver operating characteristic 0.74 [0.67 to 0.80] versus 0.68 [0.61 to 0.75], P < 0.05) and net reclassification improvement in comparison with the neonatal KDIGO criteria. In younger children, the modified criteria had good discriminatory capacity but were comparable with the adult KDIGO criteria, and the net reclassification improvement was near zero.
Conclusions
Using indwelling catheters for UO measurements, our study reinforced that the current KDIGO criteria may require adjustments to better serve the neonate population. In addition, using the UO criteria, we validated the adult KDIGO criteria in children aged 1–24 months.
Introduction
AKI is prevalent among critically ill neonates and is linked to increased morbidity and mortality.1–3 Despite challenges in comparing studies because of variations in practices for assessing AKI in neonatal intensive care units (ICUs), such as urine output (UO) reporting and the frequency of serum creatinine (sCr) measurements, the estimated incidence of AKI from the largest epidemiological study to date is nearly 30%, increasing to 48% in neonates with a gestational age of <28 weeks.1,4 In neonates undergoing cardiac surgery, its incidence reaches more than 50%, with increased in-hospital mortality.5
Although AKI is common in neonates and has severe consequences, a precise definition of AKI is lacking.6 With respect to the sCr criteria, the Kidney Disease Improving Global Outcomes (KDIGO) guidelines extrapolated increment thresholds from adults, modifying the cutoff for stage 3—sCr higher than 2.5 in neonates versus 4 mg/dl in adults.1 Despite various definitions being used for UO, the current threshold for oliguria integrated into the neonatal KDIGO AKI definition has been empirically set by an expert panel as 1 ml/kg per hour for 24 hours, which is >0.5 ml/kg per hour in adults.6
Owing to challenges in documenting UO accurately in this population, often stemming from technical measurement difficulties, AKI diagnosis has frequently relied solely on sCr values,1 with a few studies including UO criteria.5 In 2013, our group initially proposed a higher UO threshold (1.5 ml/kg per hour), which better predicts mortality in preterm infants with AKI.7 Recently, an observational study in two centers (France and Switzerland) confirmed our findings.8 However, in both studies, UO was measured using diaper weights to estimate UO. Nevertheless, this method is susceptible to imprecise measurements because of output mixing with stool, often resulting in an overestimation of urine volume. In addition, a recent editorial highlighted the lack of prospective studies with precise UO measurement as a barrier to altering urine thresholds in the neonatal AKI definition.9 Hasson et al.10 reported a specific neonate cohort subjected to the Norwood procedure (a cardiac surgery with high rates of AKI) in which all patients had UO measured with an indwelling catheter; however, UO thresholds other than those defined by the neonatal KDIGO criteria were not explored.
Regarding younger children (1–24 months), studies evaluating AKI use only sCr,11–13, and to the best of our knowledge, there are no studies validating UO thresholds in younger children aged 1–24 months or older. We think this is an urgent topic because although tubular maturity is complete at this age, the hydric requirements can be different from those of adults because of a larger water percentage in body composition.
In this study, we evaluated two large cohorts of neonatal and young children (age 1–24 months) who underwent cardiac surgery between 2018 and 2023. Using this cohort, we had the opportunity to use indwelling urinary catheters for 48 hours after surgery. Our main objective was to assess UO and determine the optimal thresholds for defining AKI in neonatal and younger children using a composite outcome—severe stage 3 AKI by the sCr criterion only, the need for KRT or death.
Methods
Study Design, Setting, and Patient Selection
This prospective cohort study was conducted at a single center, Hospital do Coração de Messejana (Ceará, Brazil), which serves as a reference center for pediatric cardiac surgery in a population of more than six million patients. Our center is a participant in the International Quality Improvement Collaborative,14 which has been collecting data on congenital heart disease surgeries from low- and middle-income countries since 2007 to compare outcomes and foster collaborative quality improvement efforts. Children aged 2 years or younger who underwent cardiac surgery between January 2018 and December 2023 were eligible. Children who received renal transplants, who had preoperative sCr levels >2.5 mg/dl, or who were undergoing chronic dialysis were excluded. Additional exclusion criteria included death within 24 hours after surgical procedure, absence of sCr measurement before surgery and within 48 hours after the procedure, absence of UO registration for at least 24 hours after surgery or achievement of the composite outcome (see below) before 24 hours after surgery. Children were recruited preoperatively and followed postoperatively. The study was approved by the institution's research ethics board (no. 12950413.7.0000.5039), and informed consent was obtained from patients or parents/guardians before participation.
Data Collection and Study Procedures
Demographic data and medical history were recorded, and the preoperative eGFR was determined using the updated Schwartz equation.15 Height was recorded to calculate the preoperative eGFR. Additional clinical variables collected included age, sex, use of cardiopulmonary bypass (CPB), CPB time, preterm status (<37 weeks), preoperative sCr, and the Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery risk of mortality (STAT) Categories category.16 Variables collected up to 48 hours after surgery included the maximum vasoactive inotropic score,17 diuretic use and dosage, intraoperative and 48 hours after ICU admission fluid balance, maximum sCr and lactate value, and minimum UO measurement on a 24-hour basis. In addition, urinary tract infections and major mechanical/hemorrhagic complications related to indwelling urinary catheters were recorded. Details about the methodology used to collect and calculate the variables are provided in the Supplemental Material.
AKI Staging
Initially, AKI was defined according to two different criteria: the adult and neonatal KDIGO criteria.6,18 The adult KDIGO criteria were modified to adjust our 24-hour UO measurements instead of the 6–12 hour intervals. Patients were classified according to AKI severity based on the maximum increase in sCr compared with the preoperative value. Patients were assigned to the appropriate AKI stage if they met either the UO or sCr criterion or both. Because the adult UO KDIGO criterion between stages 1 and 2 differs only in the time window, patients with UO between 0.3 and 0.5 ml/kg per hour were classified as having at least stage 2 AKI. In addition, patients were reclassified according to UO cutoff points, as described below, and reclassified into the newly modified AKI classifications. For all classifications, only UO and sCr up to 48 hours after surgery were considered. Table 1 displays all four staging systems—two currently used in neonatal and pediatric populations and two evaluated in the present paper.
Table 1.
Adult and neonatal Kidney Disease Improving Global Outcomes (AKI) definition and modified AKI definitions for neonatal and younger children
| Stage | Neonatal Population | Younger Children Population | ||
|---|---|---|---|---|
| Neonatal KDIGO AKI Definition | Modified Neonatal AKI Definition | Adult KDIGO AKI Definition (Modified for UO for 24 h) | Modified Younger Children AKI Definition | |
| 1 | Urine <1 ml/kg per hour for 24 h; creatinine=0.3-mg/dl rise within 48 h or ≥1.5×–1.9× baseline | Urine <3 ml/kg per hour for 24 h; creatinine=0.3-mg/dl rise within 48 h or ≥1.5×–1.9× baseline | Creatinine=0.3-mg/dl rise within 48 h or ≥1.5×–1.9× baseline | Urine <1.8 ml/kg per hour for 24 h; creatinine=0.3-mg/dl rise within 48 h or ≥1.5×–1.9× baseline |
| 2 | Urine <0.5 ml/kg per hour for 24 h; creatinine ≥2.0×–2.9× baseline | Urine <2 ml/kg per hour for 24 h; creatinine ≥2.0×–2.9× baseline | Urine <0.5 ml/kg per hour for 24 h; creatinine ≥2.0×–2.9× baseline | Urine <1 ml/kg per hour for 24 h; creatinine ≥2.0×–2.9× baseline |
| 3 | Urine <0.3 ml/kg per hour for 24 h; creatinine ≥3× baseline | Urine <1 ml/kg per hour for 24 h; creatinine ≥3× baseline | Urine <0.3 ml/kg per hour for 24 h; creatinine ≥3× baseline | Urine <0.5 ml/kg per hour for 24 h; creatinine ≥3× baseline |
AKI classifications currently used for neonatal and pediatric populations and the classifications with changes in urine output thresholds evaluated in this study. KDIGO, Kidney Disease Improving Global Outcomes; UO, urine output.
Outcomes
Because only the first 48 hours after surgery were evaluated, and this timeframe could be too early to predict hospital mortality, we chose to assess a composite outcome, which included developing severe stage 3 AKI by the sCr criterion only or the need for KRT both during the ICU stay and hospital mortality. Each outcome was also analyzed separately for the main findings.
Statistical Analysis
A detailed description is provided in the Supplemental Material. Briefly, descriptive statistics are expressed as the mean±SD or median (interquartile range [IQR]), and comparisons between groups were performed as appropriate. The association between AKI stage and the composite outcome was assessed after adjusting for confounders. Receiver operating characteristic (ROC) curves were generated to evaluate the accuracy of the lowest 24-hour UO up to 48 hours after surgery in ml/kg per hour and AKI classifications for predicting mortality. For UOs, cutoff points were chosen according to the highest Youden index. These points were tested with ROC curves against continuous data to determine the cutoff points that best preserved the area under the curve-ROC (AUC-ROC). Furthermore, we determined the net reclassification improvement (NRI) between adult and neonatal KDIGO and modified AKI criteria, defining KDIGO as the classification variable and modified criteria as the reclassification variable.19 P values < 0.05 were considered to indicate statistical significance. The statistical analysis was performed using SPSS 21.0 for Windows and R. Data are presented as the mean±SD.
Results
Population Characteristics
In total, 1104 patients younger than 2 years underwent cardiac surgery between January 2018 and December 2023. Among these patients, nine were excluded because they died within 24 hours after the surgical procedure, 20 because they had no sCr measurements taken before surgery, 14 because they initiated KRT within 24 hours after surgery, and 36 because they had no UO data available. A total of 1024 patients remained in the final analysis—253 in the neonatal group and 772 in the younger children group (Figure 1).
Figure 1.
Flowchart of patients and exclusion criteria.
In the neonatal group, 44 (17.4%) patients were preterm. Most neonatal patients had STAT category 3 or 4 (79.4%) while the great majority of younger pediatric patients had STAT category 2 or 4 (75.2%). The mean eGFRs before surgery were 46.7±21.6 and 94.4±41.7 ml/min per 1.73 m2 in neonates and younger children, respectively. In addition, 65.2% and 78.6% of the patients in the neonatal and younger children groups, respectively, underwent CPB.
More than 85% of patients had at least three sCr measurements after surgery (at ICU admission and in the morning of the first and second postoperative days). The median sCRs were 5.5 (IQR 4–9) and 5 (IQR 3–11) in neonates and young children, respectively. Overall, more patients in the neonatal group achieved our composite outcome than did those in the younger children group (48.6% versus 16.9%, P < 0.001). All three outcomes (severe AKI by sCr, need for dialysis, and hospital death) were more frequent in neonates, and the earliest outcome was KRT, followed by severe AKI and death; the incidence and median time from ICU admission to each outcome are shown in Supplemental Table 1. Indwelling urinary catheter-related complications were rare: only three urinary tract infection episodes and no mechanical or bleeding complications were reported.
Factors Associated with Composite Outcomes
Generally, patients with poor composite outcomes had reduced eGFRs preoperatively, increased CPB use, and increased vasoactive inotropic score, lactate, and fluid balance. In addition, the STAT score was associated with poor composite outcomes. Diuretic use was not associated with a poor outcome. A complete description of the baseline data and patient characteristics up to 48 hours after surgery according to age group and composite outcome is shown in Table 2. The multivariable analysis, including the evaluated AKI staging system, is presented below and in Supplemental Table 2.
Table 2.
Cohort description of neonates and younger children according to composite outcome
| Variable | Neonates | Younger Children | ||||
|---|---|---|---|---|---|---|
| All (N=253) | No composite Outcome (n=130) | Composite Outcome (n=123) | All (N=772) | No Composite Outcome (n=642) | Composite Outcome (n=130) | |
| Male sex, No. (%) | 166 (65.6) | 79 (60.8) | 87 (70.7) | 454 (58.8) | 367 (57.2) | 87 (66.9) |
| Preterm, No. (%) | 44 (17.4) | 24 (18.5) | 20 (16.3) | 120 (15.5) | 96 (15.0) | 24 (18.5) |
| eGFR preoperative (ml/min per 1.73 m2), mean±SD | 46.7±21.6 | 50.4±22.7 | 42.7±19.8 | 92.1±39.6 | 94.4±41.7 | 80.4±35.4 |
| STAT category, No. (%) | ||||||
| 1 | 15 (5.9) | 14 (10.8) | 1(0.1) | 90 (11.7) | 85 (13.2) | 5 (3.8) |
| 2 | 22 (8.7) | 18 (13.8) | 4 (3.3) | 255 (33.0) | 232 (36.1) | 23 (17.7) |
| 3 | 36 (14.2) | 20 (15.4) | 16 (13.0) | 89 (11.5) | 70 (10.9) | 19 (14.6) |
| 4 | 165 (65.2) | 75 (57.7) | 90 (73.2) | 326 (42.2) | 248 (38.6) | 78 (60.0) |
| 5 | 15 (5.9) | 3 (2.3) | 12 (9.8) | 12 (15.5) | 7 (0.1) | 5 (3.8) |
| CPB use, No. (%) | 165 (65.2) | 74 (56.9) | 91 (74.0) | 607 (78.6) | 496 (77.3) | 111 (85.4) |
| CPB duration (min), median (IQR)a | 80 (40–123) | 67.5 (35–115) | 103 (45–135) | 70 (50–95) | 70 (50–90) | 89 (60–128) |
| Intraoperative fluid balance (ml/kg), median (IQR) | 11.9 (4.2–28.1) | 10.2 (3.4–24.1) | 13.4 (5.1–36.3) | 33.8 (0.6–99.0) | 3.1 (0.5–8.7) | 5.3 (1.1–17.2) |
| Diuretic use first 24 h after surgery, No. (%) | 175 (69.2) | 92 (70.1) | 83 (67.5) | 421(54.5) | 350 (54.5) | 71(54.6) |
| Diuretic use, No. (%) | 182 (71.9) | 97 (74.6) | 85 (69.1) | 470 (60.9) | 391 (60.9) | 79 (60.8) |
| Median frusemide dosage (mg/kg per day), median (IQR)b | 3 (2–4) | 3 (2.5–4) | 3 (2–4) | 1.5 (0–2.0) | 1.4 (0–2.0) | 1.8 (0–2.1) |
| Max vasoactive-inotropic score, median (IQR)b | 12.5 (5.9–21.0) | 9.2 (5.0–15.7) | 17.8 (10.0–28.0) | 7.0 (5.0–13.6) | 6.4 (4.5–12.0) | 14.1 (7.8–25.0) |
| Max lactate (mmol/L), median (IQR)b | 2.2 (1.6–3.2) | 1.9 (1.4–2.5) | 2.7 (2.1–4.2) | 1.6 (1.3–2.2) | 1.5 (1.2–2.0) | 2.2 (1.7–3.4) |
| Cumulative fluid balance (ml/kg), median (IQR)b | −6.8 (−23.5 to 7.5) | −11.4 (−28.5 to 3.6) | −1.6 (−20.0 to 16.0) | 14.8 (−5.4 to 7.9) | −0.5 (−6.1 to 5.6) | 7.9 (0.2–19.7) |
| Mechanical ventilation duration (d), median (IQR) | 9 (4–19) | 7 (3–14.5) | 13 (6–23) | 3 (1.5–10) | 2 (1.5–7) | 13 (6–23) |
| Length of ICU stay (d), median (IQR) | 15 (8–21) | 13 (7–18) | 22 (10–26) | 7 (3–18) | 6 (3–15) | 18 (9–24) |
CPB, cardiopulmonary bypass; ICU, intensive care unit; STAT, Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery risk of mortality.
Only in those with cardiopulmonary bypass use.
Values obtained up to 48 hours after surgery.
UO as a Predictor of Composite Outcomes
In both neonates and younger children, the lowest UO had good performance as a continuous variable for the composite outcome (AUC-ROC 0.75 [95% confidence interval (CI), 0.70 to 0.81] and 0.74 [95% CI, 0.68 to 0.79], P < 0.001 for both). Neonatal patients were divided according to UO values using the Youden index into the following UO stages: stage 0: >3.0 ml/kg per hour; stage 1: 2.0–3.0 ml/kg per hour; stage 2: 1.0–2.0 ml/kg per hour; and stage 3: <1.0 ml/kg per hour. For younger children, the values were as follows: stage 0, >1.8 ml/kg per hour; stage 1, 1.0–1.8 ml/kg per hour; stage 2, 0.5–1.0 ml/kg per hour; and stage 3, <0.5 ml/kg per hour—Table 1. The AUC-ROC curves comparing UO as a continuous variable or as a categorical variable were comparable (Supplemental Figure 1, A and B).
The Incidence of AKI in the First 48 Hours Using UO Stages Combined with the Definitions of Neonatal and Adult AKI from KDIGO
We used the abovementioned UO stages to create a modified AKI staging system using the same cutoff of sCr increments suggested by KDIGO—see Table 1. Neonatal patients were classified according to the modified neonatal criteria and neonatal KDIGO criteria while younger children were classified according to the modified younger children criteria and adult KDIGO criteria. In neonates, the incidence of AKI increased from 40.7% according to the neonatal KDIGO definition to 66.8% according to the modified neonatal AKI definition. In younger children, the incidence of AKI increased from 42.5% according to the adult KDIGO definition to 56.6% according to the modified younger children criteria. In both age groups, the increase occurred across all AKI stages (Table 3).
Table 3.
AKI incidence among neonates and younger children using traditional (neonatal or adult Kidney Disease Improving Global Outcomes criteria) and modified definitions for each age range and net reclassification according to composite outcome
| Modified Neonatal AKI Criteria | Neonatal KDIGO AKI Criteria | ||||
| Positive Composite Outcome | |||||
| No-AKI | AKI Stage 1 | AKI Stage 2 | AKI Stage 3 | ||
| No-AKI | 23 | — | — | — | |
| AKI stage 1 | 12a | 11 | — | — | |
| AKI stage 2 | 17a | 10a | 15 | — | |
| AKI stage 3 | —a | 10a | 8a | 17 | |
| Negative Composite Outcome | |||||
| No-AKI | AKI Stage 1 | AKI Stage 2 | AKI Stage 3 | ||
| No-AKI | 61 | —a | —a | —a | |
| AKI stage 1 | 17 | 15 | —a | —a | |
| AKI stage 2 | 20 | 1 | 15 | —a | |
| AKI stage 3 | — | 1 | — | — | |
| Proportion reclassified up correctly: positive composite outcome | 46.3% | NRI (95% CI) favoring modified classification: 16.3% (4.5 to 28.2) | |||
| Proportion reclassified up incorrectly: negative composite outcome | 30.0% | ||||
| Modified Younger Children AKI Criteria | Adult KDIGO AKI Criteria | ||||
| Positive Composite Outcome | |||||
| No-AKI | AKI Stage 1 | AKI Stage 2 | AKI Stage 3 | ||
| No-AKI | 17 | — | — | — | |
| AKI stage 1 | 12a | 17 | — | — | |
| AKI stage 2 | 1a | 5a | 42 | — | |
| AKI stage 3 | —a | —a | 4a | 32 | |
| Negative Composite Outcome | |||||
| No-AKI | AKI Stage 1 | AKI Stage 2 | AKI Stage 3 | ||
| No-AKI | 318 | —a | —a | —a | |
| AKI stage 1 | 81 | 116 | —a | —a | |
| AKI stage 2 | 15 | 7 | 79 | —a | |
| AKI stage 3 | — | — | 2 | 24 | |
| Proportion reclassified up correctly: positive composite outcome | 16.9% | NRI (95% CI) favoring modified classification: 0.6% (−6.5 to 7.6) | |||
| Proportion reclassified up incorrectly: negative composite outcome | 16.3% | ||||
CI, confidence interval; KDIGO, Kidney Disease Improving Global Outcomes; NRI, net reclassification improvement.
Indicates patients who were correctly reclassified.
Associations of Different AKI Staging Criteria with Composite Outcomes in Neonatal and Younger Children
In younger children, there was a stepwise increase in the odds ratio for the composite outcome according to AKI staging when both the adult KDIGO and modified younger children criteria were used. However, in the neonatal population, according to the neonatal KDIGO criteria, both no-AKI and stage 3 AKI were associated with a low or greater frequency of poor outcomes; however, there was no clear distinction between AKI stages 1 and 2, in contrast with the stepwise increase in composite outcomes when modified neonatal AKI criteria were applied (Table 4).
Table 4.
Adjusted odds ratios for composite outcomes according to traditional (neonatal or adult Kidney Disease Improving Global Outcomes criteria) and modified definitions for each age range
| AKI Stage | Neonatal Population | |||
|---|---|---|---|---|
| Neonatal KDIGO Criteria | Modified Neonatal Criteria | |||
| Adjusted OR | 95% CI | Adjusted OR | 95% CI | |
| No-AKI | Reference | Reference | ||
| AKI stage 1 | 4.1 | 1.8 to 9.6 | 1.9 | 0.8 to 4.0 |
| AKI stage 2 | 4.2 | 2.0 to 12.6 | 3.6 | 1.7 to 7.1 |
| AKI stage 3 | a | a | 38.2 | 8.8 to 69.7 |
| Younger Children Population | ||||
|---|---|---|---|---|
| Adult KDIGO Criteria | Modified Younger Children Criteria | |||
| Adjusted OR | 95% CI | Adjusted OR | 95% CI | |
| No-AKI | Reference | No-AKI | Reference | No-AKI |
| AKI stage 1 | 2.8 | 1.4 to 5.6 | 2.9 | 1.4 to 5.8 |
| AKI stage 2 | 9.6 | 4.6 to 16.3 | 9.4 | 4.7 to 16.9 |
| AKI stage 3 | 20.3 | 10.8 to 37.2 | 21.5 | 12.3 to 49.9 |
CI, confidence interval; CPB, cardiopulmonary bypass; KDIGO, Kidney Disease Improving Global Outcomes; STAT, Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery risk of mortality.
All patients had a poor composite outcome. Adjusted for sex, preoperative eGFR, Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery risk of mortality category, cardiopulmonary bypass use and duration, maximum vasoactive-inotropic score, maximum lactate, and intraoperative and cumulative fluid balance.
Discrimination and NRI
Compared with the neonatal KDIGO criteria, modified neonatal AKI staging had a greater discriminatory capacity for predicting composite outcomes in the neonatal population—AUC-ROC 0.74 (95% CI, 0.67 to 0.80) versus 0.68 (95% CI, 0.61 to 0.75), P < 0.05. However, in younger children, the AUC-ROC curves were very similar between the modified younger children and adult KDIGO staging criteria—AUC-ROC 0.77 (95% CI, 0.73 to 0.82) versus 0.76 (95% CI, 0.11 to 0.81), P = 0.78 Figure 2, A and B. Supplemental Table 3 displays the AUC-ROC for both criteria for each individual outcome according to age group. As expected, the discriminatory capacity was better for the earliest outcomes; however, even for hospital mortality, a later event, modified neonatal AKI criteria had good discriminatory capacity. Supplemental Table 4 displays the sensitivity, specificity, and positive and negative predictive values for each AKI stage.
Figure 2.
Discriminatory capacities for KDIGO and modified criteria. AUC-ROC curve for (A) neonatal KDIGO and modified neonatal criteria—P for comparison <0.05—and (B) adult KDIGO and modified infant criteria—P for comparison=0.78. The AUC-ROC curve was used to predict the composite outcome. AUC-ROC, area under the curve-receiver operating characteristic; CI, confidence interval; KDIGO, Kidney Disease Improving Global Outcomes.
We also compared the different AKI staging criteria using the NRI. Compared with the neonatal KDIGO criteria, the modified AKI staging criteria for neonatal patients had a total NRI of 16.3%. By contrast, in the younger children group, the modified AKI staging criteria correctly classified 16.9% of the patients compared with the adult KDIGO criteria but incorrectly classified 16.3% of the controls, with a total NRI of only 0.6% (Table 3).
Sensitivity Analysis
To further explore whether diuretics can be confounders in our analysis, we performed a sensitivity analysis according to diuretic use in neonates and younger children. In younger children receiving diuretics (n=470), there was no significant difference in the AUC-ROC when comparing modified younger children (more liberal in UO thresholds) with adults according to the KDIGO AKI criteria—0.78 (95% CI, 0.72 to 0.84) versus 0.78 (0.71 to 0.84), P = 0.95. Neonatal patients who did not receive diuretics up to 48 hours after surgery (n=71) were included. Overall, the AUC-ROC of the modified neonatal AKI criteria remained better than that of the neonatal KDIGO criteria (AUC-ROC 0.71 [95% CI, 058 to 0.82] versus 0.66 [95% CI, 0.54 to 0.79], P < 0.05).
Discussion
In this study, we evaluated two populations for up to 48 hours after cardiac surgery: one neonatal population and one ranging from 1 month to 2 years old. Using UO measurements obtained by indwelling catheters, we demonstrated different thresholds for 24-hour UO in neonates than those suggested by the neonatal KDIGO criteria. We found that a UO of <3 ml/kg per hour over 24 hours had an impact on various outcomes. Our study also assessed a cohort of younger children, enabling us to validate the UO values currently used for adults in children age up to 2 years.
One main finding of our study is that a higher threshold for UO (3 ml/kg per hour) was already associated with the main outcomes. This threshold is even greater than that previously suggested by two previous studies that explored UO in the neonatal population.7,8 One important characteristic of the study to be discussed is whether the high-dose use of diuretics post-cardiac surgery can affect these UO thresholds. Several reasons argue against this hypothesis. First, studies in adults have demonstrated that diuretic use does not alter the relationship between reduced UO and poor outcomes.20 Additionally, although substantial center heterogeneity is detected in multicenter studies,5 the UO rate in this study is comparable with that in other studies evaluating UO and AKI diagnosis in neonates; for example, the mean UO rate in this study is comparable with that in our previously published cohort of neonates in a general ICU (2.5 versus 2.1 ml/kg per hour, respectively).7 In other cohorts of general neonatal ICUs, the prevalence of patients within different ranges was also very similar.18 Finally, in a sensitivity analysis with only younger children who received diuretics, no increase in discriminatory capacity was observed when UO thresholds were increased compared with those of the adult KDIGO UO criteria. By contrast, in neonates who did not receive diuretics, the modified neonatal criteria performed better than did the neonatal KDIGO criteria.
We believe that a key strength of the present research lies in the utilization of indwelling urinary catheters to enable precise measurement of UO in all patients. However, the hospital protocol mandates the removal of catheters on the third day postsurgery if deemed unnecessary. Given the significant knowledge gap in precisely defining UO in neonates and early infancy, we opted to collect data only up to 48 hours after surgery. Because AKI during this timeframe could be too early to predict hospital mortality, we chose to assess a composite outcome comprising hospital mortality, the need for dialysis, or severe AKI, as indicated by sCr levels during the ICU stay, all of which represent robust end points. Notably, when analyzing each outcome separately, the results remained consistent.
Recently, the Neonatal and Pediatric Heart and Renal Outcomes Network (NEPHRON) Investigators published two multicenter studies (n=22) evaluating AKI after cardiac surgery in neonates. Despite significant differences between centers in the NEPHRON group, as evidenced by the wide variation in AKI incidence (from 27% to 86%), it is important to highlight some differences between our center and the consortium's published data. We believe that one main difference is the greater number of patients who underwent KRT within the first 24 hours (referred to as prophylactic peritoneal dialysis) in the NEPHRON data. These patients were excluded from our study. There are two key reasons that may explain this discrepancy: (1) Our newborn patients had three times fewer STAT 5 category cases than did patients in the NEPHRON database. These high-risk patients are more prone to initiate KRT early. (2) Patients in our group were more likely to receive early diuretic therapy (52 versus 69%), reflecting an approach aimed at stimulating a negative fluid balance initially with diuretics rather than peritoneal dialysis.
In addition to neonatal patients, we also examined younger pediatric patients. While KDIGO recommends employing the adult KDIGO staging system in this age group,18 to the best of our knowledge, these values have not been validated in any pediatric population except neonates using UO criteria. Despite the expected tubular renal maturity in younger children, the proportionally greater water body compared with adults necessitated an outcome validation of such thresholds, thereby reinforcing the likelihood that adult KDIGO UO thresholds are applicable across all age groups except neonates.
Our study has several limitations, but we believe that the two most important limitations have already been discussed: first, whether evaluating post-cardiac surgery neonates and younger children could limit data validation to other critically ill populations, and second, assessing AKI staging only up to 48 hours after surgery. These limitations were necessary to ensure accurate measurement of UO. Precise measurement of UO in critically ill children, mainly in neonates, is challenging because this population can be more prone to infection, and the size limitations of both the patients and the catheters restrict the use of indwelling catheters.21,22 Our data suggest that, for at least short periods, the use of indwelling urinary catheters is safe. Another limitation is that our study was conducted at only one center, and multicenter validation of the proposed thresholds is necessary.
In conclusion, we conducted prospective data collection involving two large cohorts of neonates and younger children (1 month to 2 years) after cardiac surgery, with precise measurements of UO up to 48 hours after surgery. Using hard outcomes, we validated adult KDIGO urine thresholds on a 24-hour basis in younger children (1 month to 2 years) and proposed higher UO thresholds for all AKI stages in neonates.
Supplementary Material
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/CJN/B985.
Funding
A.B. Libório: Conselho Nacional de Desenvolvimento Científico e Tecnológico (306377/2022-5).
Author Contributions
Conceptualization: Alexandre Braga Libório, Candice Torres de Melo Bezerra Cavalcante.
Data curation: Andrea Consuelo de Oliveira Teles, Klebia Magalhães Pereira Castello Branco, Candice Torres de Melo Bezerra Cavalcante, Adriana Torres de Melo Bezerra Girão.
Formal analysis: Alexandre Braga Libório, Klebia Magalhães Pereira Castello Branco, Adriana Torres de Melo Bezerra Girão.
Investigation: Candice Torres de Melo Bezerra Cavalcante.
Methodology: Andrea Consuelo de Oliveira Teles, Alexandre Braga Libório, Adriana Torres de Melo Bezerra Girão.
Writing – original draft: Alexandre Braga Libório.
Writing – review & editing: Andrea Consuelo de Oliveira Teles, Alexandre Braga Libório, Klebia Magalhães Pereira Castello Branco, Candice Torres de Melo Bezerra Cavalcante.
Data Sharing Statement
All data are included in the manuscript and/or supporting information; partial restrictions to the data and/or materials apply. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Supplemental Material
This article contains the following supplemental material online at http://links.lww.com/CJN/B984.
Supplemental Table 1. Individual and composite outcomes by age group.
Supplemental Table 2. Variables included in the final model for predicting the composite outcome.
Supplemental Table 3. Area under the receiver operating characteristic curve (AUC-ROC) for both criteria for each individual outcome according to age group.
Supplemental Table 4. Sensitivity, specificity, and positive and negative predictive values for the composite outcome of the actual and modified AKI systems by age group considering both urine output and serum creatinine.
Supplemental Figure 1. Area under the curve-receiver operating characteristic (AUC-ROC) curve for lowest urine output as a continuous variable and modified (A) neonatal and (B) younger children AKI urine output criteria. The AUC-ROC curve was used to predict the composite outcome.
References
- 1.Jetton JG Boohaker LJ Sethi SK, et al. Incidence and outcomes of neonatal acute kidney injury (AWAKEN): a multicentre, multinational, observational cohort study. Lancet Child Adolesc Health. 2017;1(3):184–194. doi: 10.1016/S2352-4642(17)30069-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Stoops C Stone S Evans E, et al. Baby NINJA (nephrotoxic injury negated by just-in-time action): reduction of nephrotoxic medication-associated acute kidney injury in the neonatal intensive care unit. J Pediatr. 2019;215:223–228.e6. doi: 10.1016/j.jpeds.2019.08.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Charlton JR Boohaker L Askenazi D, et al. Incidence and risk factors of early onset neonatal AKI. Clin J Am Soc Nephrol. 2019;14(2):184–195. doi: 10.2215/CJN.03670318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Harer MW Askenazi DJ Boohaker LJ, et al. Association between early caffeine citrate administration and risk of acute kidney injury in preterm neonates: results from the AWAKEN study. JAMA Pediatr. 2018;172(6):e180322. doi: 10.1001/jamapediatrics.2018.0322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Alten JA Cooper DS Blinder JJ, et al. Epidemiology of acute kidney injury after neonatal cardiac surgery: a report from the multicenter neonatal and pediatric heart and renal outcomes Network. Crit Care Med. 2021;49(10):e941–e951. doi: 10.1097/CCM.0000000000005165 [DOI] [PubMed] [Google Scholar]
- 6.Zappitelli M Ambalavanan N Askenazi DJ, et al. Developing a neonatal acute kidney injury research definition: a report from the NIDDK neonatal AKI workshop. Pediatr Res. 2017;82(4):569–573. doi: 10.1038/pr.2017.136 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bezerra CTdM, Vaz Cunha LC, Libório AB. Defining reduced urine output in neonatal ICU: importance for mortality and acute kidney injury classification. Nephrol Dial Transplant. 2013;28(4):901–909. doi: 10.1093/ndt/gfs604 [DOI] [PubMed] [Google Scholar]
- 8.De Mul A Parvex P Héneau A, et al. Urine output monitoring for the diagnosis of early-onset acute kidney injury in very preterm infants. Clin J Am Soc Nephrol. 2022;17(7):949–956. doi: 10.2215/CJN.15231121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Harer MW, Charlton JR. Urine or you’re out? Clin J Am Soc Nephrol. 2022;17(7):939–941. doi: 10.2215/CJN.06010522 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hasson DC Alten JA Bertrandt RA, et al. Persistent acute kidney injury and fluid accumulation with outcomes after the Norwood procedure: report from NEPHRON. Pediatr Nephrol. 2024;39(5):1627–1637. doi: 10.1007/s00467-023-06235-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Selewski DT Cornell TT Heung M, et al. Validation of the KDIGO acute kidney injury criteria in a pediatric critical care population. Intensive Care Med. 2014;40(10):1481–1488. doi: 10.1007/s00134-014-3391-8 [DOI] [PubMed] [Google Scholar]
- 12.Sutherland SM Byrnes JJ Kothari M, et al. AKI in hospitalized children: comparing the pRIFLE, AKIN, and KDIGO definitions. Clin J Am Soc Nephrol. 2015;10(4):554–561. doi: 10.2215/CJN.01900214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Macedo E Cerdá J Hingorani S, et al. Recognition and management of acute kidney injury in children: the ISN 0by25 Global Snapshot study. PLoS One. 2018;13(5):e0196586. doi: 10.1371/journal.pone.0196586 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jenkins KJ Castañeda AR Cherian KM, et al. Reducing mortality and infections after congenital heart surgery in the developing world. Pediatrics. 2014;134(5):e1422–e1430. doi: 10.1542/peds.2014-0356 [DOI] [PubMed] [Google Scholar]
- 15.Schwartz GJ Muñoz A Schneider MF, et al. New equations to estimate GFR in children with CKD. J Am Soc Nephrol. 2009;20(3):629–637. doi: 10.1681/ASN.2008030287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jacobs JP O’Brien SM Pasquali SK, et al. Variation in outcomes for risk-stratified pediatric cardiac surgical operations: an analysis of the STS congenital heart surgery database. Ann Thorac Surg. 2012;94(2):564–572. doi: 10.1016/j.athoracsur.2012.01.105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Favia I, Vitale V, Ricci Z. The vasoactive-inotropic score and levosimendan: time for LVIS?. J Cardiothorac Vasc Anesth. 2013;27(2):e15–e16. doi: 10.1053/j.jvca.2012.11.009 [DOI] [PubMed] [Google Scholar]
- 18.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–c184. doi: 10.1159/000339789 [DOI] [PubMed] [Google Scholar]
- 19.Cook NR. Statistical evaluation of prognostic versus diagnostic models: beyond the ROC curve. Clin Chem. 2008;54(1):17–23. doi: 10.1373/clinchem.2007.096529 [DOI] [PubMed] [Google Scholar]
- 20.Han SS Kang KJ Kwon SJ, et al. Additional role of urine output criterion in defining acute kidney injury. Nephrol Dial Transplant. 2012;27(1):161–165. doi: 10.1093/ndt/gfr312 [DOI] [PubMed] [Google Scholar]
- 21.Hsu HE Mathew R Wang R, et al. Health care–associated infections among critically ill children in the US, 2013-2018. JAMA Pediatr. 2020;174(12):1176–1183. doi: 10.1001/jamapediatrics.2020.3223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Crigger C, Kuzbel J, Al-Omar O. Choosing the right catheter for pediatric procedures: patient considerations and preference. Res Rep Urol. 2021;13:185–195. doi: 10.2147/RRU.S282654 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data are included in the manuscript and/or supporting information; partial restrictions to the data and/or materials apply. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.