Celeste G. Dixon

Background: KDIGO acute kidney injury (AKI) criteria are not specific to children. Pediatric Organ Dysfunction Information Update Mandate (PODIUM) renal dysfunction criteria were developed for children but have not been adopted. PODIUM defines renal dysfunction more strictly than KDIGO and adds fluid overload (FO) as an independent criterion. The impact of these changes on AKI diagnosis has not been evaluated. Methods: Retrospective analysis of pediatric intensive care unit (PICU) admissions at an academic children’s hospital from 06/2022–05/2025. Creatinine, urine output, fluid balance, and weight were extracted in 6-h intervals. Patients were classified by maximum KDIGO AKI stage and PODIUM AKI. Outcomes were mortality, PICU length of stay (LOS), and major adverse kidney events at 30 days (MAKE-30). Results: In 10,454 patients, 2,602 (25%) met KDIGO AKI criteria: 1,124 (11%) KDIGO 1, 579 (5%) KDIGO 2, and 899 (9%) KDIGO 3. By PODIUM, 1,599 (15%) met criteria. 105 (9%) KDIGO 1 patients and 16 (0.2%) patients without KDIGO AKI met PODIUM criteria; all KDIGO 2/3 met PODIUM. 110/1,599 (7%) of patients with PODIUM AKI met only FO criterion. Outcomes were similar between KDIGO 2/3 and PODIUM AKI (mortality 6–7%, median LOS 7–8.5 days, MAKE-30 12%). Patients meeting PODIUM by FO alone had worse outcomes than KDIGO Stage 1 or No AKI (mortality 4% vs. 1–2%, median LOS 13.8 vs. 1.5–5.5 days), but not different MAKE-30. Conclusions: PODIUM criteria capture all KDIGO 2/3 patients while excluding most KDIGO 1 patients. PODIUM criteria appear useful for identifying critically ill children with clinically significant AKI resulting in greater short-term illness severity, while KDIGO criteria capture a larger spectrum of AKI severity. Graphical A higher resolution version of the Graphical Abstract is available as Supplementary information.

IMPORTANCE: Patients with acute respiratory distress syndrome (ARDS) are at risk for higher morbidity and mortality with concurrent acute kidney injury (AKI). AKI diagnosis relies on serum creatinine, which is confounded by fluid overload. Use of fluid overload-adjusted creatinine can diagnose AKI (termed Cryptic AKI) that would otherwise be missed by the traditional Kidney Disease Improving Global Outcomes (KDIGO) definition. It is unknown whether Cryptic AKI is biochemically distinct from patients without AKI. OBJECTIVES: Determine whether pediatric ARDS patients with Cryptic AKI have a biochemical profile that is similar or distinct from patients with KDIGO-defined AKI or no AKI. DESIGN, SETTING, AND PARTICIPANTS: Observational study of 333 pediatric ARDS patients in a quaternary PICU. MAIN OUTCOMES AND MEASURES: Mortality, length of stay, and ventilator-free days were measured for all patients. Biomarkers indicative of tissue damage and inflammation were measured on days 1, 3, and 7 of ARDS. RESULTS: Cryptic AKI was associated with higher mortality and longer length of stay than no AKI, similar to KDIGO AKI. Biomarkers angiopoietin-2, pro-collagen III N-terminal peptide, heat shock protein 70, tumor necrosis factor receptor 1, and interleukin-8 were significantly higher in Cryptic and KDIGO AKI than in no AKI. CONCLUSIONS AND RELEVANCE: Cryptic AKI is clinically significant and represents true kidney injury, with a biomarker profile similar to KDIGO AKI and distinct from patients without AKI.

OBJECTIVE: Although efforts have been made to standardize fluid balance calculations in the ICU, there is a limited understanding of how different calculation methods relate to one another across an ICU admission. We quantified the agreement between the cumulative fluid balance calculated from fluid intake and output (CFBf) and cumulative fluid balance calculated from serial weights (CFBw) in critically ill children during the first week of ICU admission. DESIGN: Retrospective, multicenter, federated observational study. SETTING: Four pediatric medical-SICUs (PICU) and two pediatric cardiac ICUs (PCICU) from four tertiary care centers. PATIENTS: Analysis included 8,895 pediatric patients (younger than 19 yr) representing 12,388 ICU encounters from 2023 to 2024. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A patient’s anchor weight was the weight closest to ICU admission. CFBf and CFBw were calculated at the time of new weight measurements. We assessed agreement between CFBf and CFBw using Bland-Altman analyses, stratified by ICU day and patient subgroups (neonates, early anchor weights [weight on ICU day 0], and encounters with unmeasured urine occurrences). Across all units and subgroups, CFBf exceeded CFBw (mean difference: all patients = 4.7 %CFB, early anchor weight = 4.7 %CFB, neonates = 5.9 %CFB). The mean difference increased significantly over time (days 0–3: 2.7% vs. days 4–7: 8.1%, p < 0.05), with greater divergence in neonates and those with early anchor weights. CONCLUSIONS: CFBf consistently exceeded CFBw across all subgroups, with a greater divergence on ICU days 4–7. Clinicians should understand these differences, prioritizing early and frequent patient weights throughout ICU admission. Future studies should assess each method’s association with patient outcomes to identify the most clinically informative CFB method.

Objective Although efforts have been made to standardize fluid balance calculations in the intensive care unit (ICU), there is a limited understanding of how different calculation methods relate to one another across an ICU admission. We quantified the agreement between the cumulative fluid balance calculated from fluid intake and output (CFBf) and from serial weights (CFBw) in critically ill children during the first week of ICU admission. Design Retrospective, multicenter, federated observational study. Setting Four pediatric medical-surgical ICUs (PICU) and two pediatric cardiac ICUs (PCICU) from four tertiary care centers. Patients Analysis included 8,895 pediatric patients (Interventions None. Measurements and Main Results A patient’s anchor weight was the weight closest to ICU admission. CFBf and CFBw were calculated at the time of new weight measurements. We assessed agreement between CFBf and CFBw using Bland-Altman analyses, stratified by ICU day and patient subgroups (neonates, early anchor weights [weight on ICU day 0], and encounters with unmeasured urine occurrences). Across all units and subgroups, CFBf exceeded CFBw (mean difference: all patients = 4.7 %CFB, early anchor weight = 4.7 %CFB, neonates = 5.9 %CFB). The mean difference increased significantly over time (days 0–3: 2.7% vs. days 4–7: 8.1%, pConclusions CFBf consistently exceeded CFBw across all subgroups, with a greater divergence on ICU days 4-7. Clinicians should understand these differences, prioritizing early and frequent patient weights throughout ICU admission. Future studies should assess each method’s association with patient outcomes to identify the most clinically informative CFB method. Research in Context Cumulative fluid balance (CFB) is calculated by fluid (intake minus output) and weight measurements in pediatric ICU patients, but few studies have directly compared these methods across multiple institutions, unit types, and throughout PICU admission. CFB calculated using fluid measures consistently exceeded weight-based calculations, with divergence increasing later in admission and for patients with anchor weights recorded closer to PICU admissions and neonates. Divergence between CFB methods may reduce CFBf reliability later in the ICU stay, so accurate and frequent PICU weight recordings are essential for reliable fluid balance assessment. At the Bedside This multicenter study demonstrates a consistent positive divergence between CFB calculated by net fluid compared to weight-based calculations across diverse PICU and PCICU populations. Differences between methods diverged later in admission (e.g., after day 3), particularly among neonates and patients with earlier recorded anchor weights, while CFB calculated before the first unmeasured urine differed little from the overall cohort. Clinicians should understand these differences and the factors that influence them, prioritizing early and frequent patient weights throughout ICU admission to recognize when net fluid-based CFB diverges from a weight-based CFB.

Background: There are risks associated with excessive intravenous fluid (IVF) administration in critically ill children. Previous efforts have described opportunities to reduce positive cumulative fluid balance (CFB) in this population but have not been widely implemented. In the wake of Hurricane Helene, a national IVF shortage led to the implementation of IVF conservation guidelines. We sought to determine if this was associated with a reduction in IVF use and CFB. Methods: The present study is a four-site cohort study of critically ill children utilizing a federated data collection framework to extract patient age, sex, weight, and daily fluid intake/output for days 1–4 of all admissions 28 days prior to and 28 days after the implementation of IVF conservation guidelines. Guidelines were individualized per institution. Total fluid intake, total IVF intake, % intake from IVF, and % CFB were compared between pre- and post-IVF conservation groups. Results: All sites had similar conservation recommendations. There were 633 patients admitted pre- and 619 patients admitted post-IVF conservation guideline implementation, with similar age and weight distributions. There was no significant difference in IVF use pre- and post-IVF conservation; 29–35% of patients had > 5% CFB on day 1 pre-IVF conservation while 27–39% did post-conservation, with increasing numbers on day 2. Conclusions: Even in the setting of a national IVF shortage, simple recommendations without structured change were insufficient to change IVF administration practices. This indicates additional practices will be needed to reduce IVF intake and % CFB in this vulnerable population. Graphical

Olfactomedin-4 (OLFM4) was associated with higher mortality, higher probability of dialysis by day 7, and fewer ventilator-free days (VFDs) in a pediatric acute respiratory distress syndrome (ARDS) cohort. In vitro, OLFM4 increased following H3 exposure in undifferentiated neutrophils, which was partly mitigated by toll-like receptor (TLR) antagonism. OLFM4 appears to be a marker, and potentially a mediator, of pathological inflammation and end-organ damage in ARDS.

Introduction: Renal dysfunction is common in critically ill children and increases morbidity and mortality risk. Diagnosis and management of renal dysfunction relies on creatinine, a delayed marker of renal injury. We aimed to develop and validate a machine learning model using routinely collected clinical data to predict 24-hour creatinine change in critically ill children before change is observed clinically.Methods: Retrospective cohort study of 39,932 pediatric intensive care unit encounters in a national multicenter database from 2007 to 2022. A neural network was trained to predict <50% or ≥50% creatinine change in the next 24 h. Admission demographics, routinely measured vital signs, laboratory tests, and medication use variables were used as predictors for the model. Data set was randomly split at the encounter level into model development (80%) and test (20%) sets. Performance and clinical relevance was assessed in the test set by accuracy of prediction classification and confusion matrix metrics.Results: The cohort had a male predominance (53.8%), median age of 8.0 years (IQR 1.9−14.6), 21.0% incidence of acute kidney injury, and 2.3% mortality. The overall accuracy of the model for predicting change of <50% or ≥50% was 68.1% (95% CI 67.6%−68.7%). The accuracy of classification improved substantially with higher creatinine values from 29.9% (CI 28.9%−31.0%) in pairs with an admission creatinine <0.3 mg/dl to 90.0–96.3% in pairs with an admission creatinine of ≥0.6 mg/dl. The model had a negative predictive value of 97.2% and a positive predictive value of 7.1%. The number needed to evaluate to detect one true change ≥50% was 14.Discussion: 24-hour creatinine change consistent with acute kidney injury can be predicted using routine clinical data in a machine learning model, indicating risk of significant renal dysfunction before it is measured clinically. Positive predictive performance is limited by clinical reliance on creatinine.

IMPORTANCE: Right atrial (RA) dysfunction is associated with worse outcomes in some populations with pulmonary hypertension or respiratory failure but the prevalence and correlates of RA dysfunction in pediatric acute respiratory distress syndrome (PARDS) are unknown. OBJECTIVES: The aim of this study was to evaluate RA function by characterizing the prevalence and pattern of RA dysfunction within the first 24 hours of PARDS onset. We hypothesized that RA dysfunction would be common and correlate with the presence of pulmonary hypertension and right ventricular (RV) systolic dysfunction. DESIGN, SETTING, AND PARTICIPANTS: Retrospective, single-center cohort study at a tertiary care PICU of children (< 18 yr) with a clinically obtained echocardiogram within 24 hours following PARDS diagnosis and healthy controls without cardiopulmonary disease. MAIN OUTCOMES AND MEASURES: Echocardiograms were evaluated for conventional and speckle-tracking (or strain) echocardiographic measures of RA and RV systolic function. Nonparametric summary statistics, comparisons, and correlational analyses were completed. RESULTS: Ninety-two PARDS patients and 55 controls were included. Using a priori thresholds (> 2 sds of control values), 49% (n = 45) of PARDS patients demonstrated RA dysfunction in at least one RA functional metric. The maximal RA strain during the reservoir phase was reduced in PARDS compared with controls (median 40.2% vs. 53.7%; p < 0.001). Patients with echocardiographic evidence of pulmonary hypertension had lower maximal RA strain during the reservoir phase (31.7%) compared with patients without (40.5%; p < 0.05). Patients with higher brain-type natriuretic peptide plasma concentrations had worse RA function. RA function significantly correlated with conventional and strain measures of RV systolic function. CONCLUSIONS AND RELEVANCE: RA dysfunction is common within the first 24 hours of PARDS onset. RA dysfunction during the reservoir phase is associated with pulmonary hypertension and RV systolic dysfunction. Future studies investigating trajectories of RA function and their association with outcomes in PARDS patients are needed.

Introduction Fluid overload (FO) due to excessive fluid administration in critically ill children is associated with worse pediatric intensive care unit (PICU) outcomes. Damage to intravenous fluid (IVF) production plants due to Hurricane Helene resulted in a national IVF shortage, leading many hospitals to implement IVF conservation guidelines. We hypothesized that these guidelines would lead to a reduction in IVF use and a decrease in % cumulative fluid balance (CFB). Methods Four-site cohort study of critically ill children utilizing a federated data collection framework to extract patient age, sex, weight, and daily fluid intake/output for all admissions one month prior to and one month after the implementation of IVF conservation guidelines. Guidelines were individualized per institution. Total fluid intake, total IVF intake, % intake from IVF, and % CFB, as total intake minus total output divided by PICU admission weight, were compared between pre and post IVF conservation groups. Results There were 633 pre and 619 post IVF conservation encounters, with similar age and size distributions. There was no significant difference in IVF use pre and post IVF conservation. All sites had similar conservation recommendations. Conclusions Simple recommendations without structured change was insufficient to change IVF administration practices, indicating additional practices will be needed to reduce IVF intake and % CFB in this vulnerable population.

Introduction Fluid overload (FO), a state of pathologic positive cumulative fluid balance (CFB), is common in Pediatric Intensive Care Units (PICU) and associated with morbidity and mortality. Because different PICUs may have unique needs, barriers, and limitations to accurately report fluid balance (FB) and reduce FO, understanding the drivers of positive FB is needed. We hypothesize CFB >5% and >10% is common within initial days of PICU admission, but that reasons for high %CFB will vary across sites, as will barriers to accurate FB recording and opportunities to improve FB recording and management. Methods Concurrent mixed methods study utilizing a retrospective observational cohort design and prospective interview and survey design performed at four tertiary pediatric ICUs. FB data was extracted from the electronic health record. A federated data collection framework allowed for rapid data aggregation. The primary outcome was %CFB on PICU days 1 and 2, defined as total intake minus total output divided by PICU admission weight. Chi-square test and Wilcoxon rank sum tests compared results across and within sites. Results Amongst 3,071 PICU encounters, day 2 CFB >5% varied from 39% to 54% (p=0.03) and day 2 CFB >10% varied from 16% to 25% (p=0.04) across sites. Urine occurrence recordings and patients receiving >100% Holliday-Segar fluids on Day 1 differed across sites (p5% and >10% was common among PICU encounters but proportion of patients varied significantly across PICUs. Individual ICUs have different drivers of FO that must be targeted to improve FB management.