Retrospective Validation of a Computerized Physiologic Equation to Predict Minute Ventilation Needs in Critically Ill Children.

OBJECTIVES: Mechanical ventilation (MV) is pervasive among critically ill children. We sought to validate a computerized physiologic equation to predict minute ventilation requirements in children and test its performance against clinician actions in an in silico trial. DESIGN: Retrospective, electronic medical record linkage, cohort study. SETTING: Quaternary PICU. PATIENTS: Patients undergoing invasive MV, serial arterial blood gas (ABG) analysis within 1-6 hours, and pharmacologic neuromuscular blockade (NMB). MEASUREMENTS AND MAIN RESULTS: ABG values were filtered to those occurring during periods of NMB. Simultaneous ABG and minute ventilation data were linked to predict serial Pa co2 and pH values using previously published physiologic equations. There were 15,121 included ABGs across 500 encounters among 484 patients, with a median (interquartile range [IQR]) of 20 (10-43) ABGs per encounter at a duration of 3.6 (2.1-4.2) hours. The median (IQR) Pa co2 prediction error was 0.00 (-3.07 to 3.00) mm Hg. In Bland-Altman analysis, the mean error was -0.10 mm Hg (95% CI, -0.21 to 0.01 mm Hg). A nested, in silico trial of ABGs meeting criteria for weaning (respiratory alkalosis) or escalation (respiratory acidosis), compared the performance of recommended ventilator changes versus clinician decisions. There were 1,499 of 15,121 ABGs (9.9%) among 278 of 644 (43.2%) encounters included in the trial. Calculated predictions were favorable to clinician actions in 1124 of 1499 ABGs (75.0%), equivalent to clinician choices in 26 of 1499 ABGs (1.7%), and worse than clinician decisions in 349 of 1499 ABGs (23.3%). Calculated recommendations were favorable to clinician decisions in sensitivity analyses limiting respiratory rate, analyzing only when clinicians made changes, excluding asthma, and excluding acute respiratory distress syndrome. CONCLUSIONS: A computerized equation to predict minute ventilation requirements outperformed clinicians' ventilator adjustments in 75% of ABGs from critically ill children in this retrospective analysis. Prospective validation studies are needed.

Characterization of Circulating Cold Shock Proteins FGF21 and RBM3 in a Multi-Center Study of Pediatric Cardiac Arrest.

Fibroblast Growth Factor 21 (FGF21) is a neuroprotective hormone induced by cold exposure that targets the ß-klotho co-receptor. ß-klotho is abundant in the newborn brain but decreases rapidly with age. RNA-Binding Motif 3 (RBM3) is a potent neuroprotectant upregulated by FGF21 in hypothermic conditions. We characterized serum FGF21 and RBM3 levels in patients enrolled in a prospective multi-center study of pediatric cardiac arrest (CA) via a secondary analysis of samples collected to evaluate brain injury biomarkers. Patients (n = 111) with remnant serum samples available from at least two of three available timepoints (0-24, 24-48 or 48-72 hours post-resuscitation) were included. Serum samples from 20 healthy controls were used for comparison. FGF21 was measured by Luminex and internally validated enzyme-linked immunoassay (ELISA). RBM3 was measured by internally validated ELISA. Of postarrest patients, 98 were managed with normothermia, while 13 were treated with therapeutic hypothermia (TH). FGF21 increased >20-fold in the first 24 hours postarrest versus controls (681 pg/mL [200-1864] vs. 29 pg/mL [15-51], n = 99 vs. 19, respectively, p < 0.0001, median [interquartile range]) with no difference in RBM3. FGF21 did not differ by sex, while RBM3 was increased in females versus males at 48-72 hours postarrest (1866 pg/mL [873-5176] vs. 1045 pg/mL [535-2728], n = 40 vs. 54, respectively, p < 0.05). Patients requiring extracorporeal membrane oxygenation (ECMO) postresuscitation had increased FGF21 versus those who did not at 48-72 hours (6550 pg/mL [1455-66,781] vs. 1213 pg/mL [480-3117], n = 7 vs 74, respectively, p < 0.05). FGF21 and RBM3 did not correlate (Spearman's rho = 0.004, p = 0.97). We conclude that in a multi-center study of pediatric CA patients where normothermic targeted temperature management was largely used, FGF21 was markedly increased postarrest versus control and highest in patients requiring ECMO postresuscitation. RBM3 was sex-dependent. We provide a framework for future studies examining the effect of TH on FGF21 or use of FGF21 therapy after pediatric CA.

N-Acetylcysteine and Probenecid Adjuvant Therapy for Traumatic Brain Injury.

N-Acetylcysteine (NAC) has shown promise as a putative neurotherapeutic for traumatic brain injury (TBI). Yet, many such promising compounds have limited ability to cross the blood-brain barrier (BBB), achieve therapeutic concentrations in brain, demonstrate target engagement, among other things, that have hampered successful translation. A pharmacologic strategy for overcoming poor BBB permeability and/or efflux out of the brain of organic acid-based, small molecule therapeutics such as NAC is co-administration with a targeted or nonselective membrane transporter inhibitor. Probenecid is a classic ATP-binding cassette and solute carrier inhibitor that blocks transport of organic acids, including NAC. Accordingly, combination therapy using probenecid as an adjuvant with NAC represents a logical neurotherapeutic strategy for treatment of TBI (and other CNS diseases). We have completed a proof-of-concept pilot study using this drug combination in children with severe TBI-the Pro-NAC Trial (ClinicalTrials.gov NCT01322009). In this review, we will discuss the background and rationale for combination therapy with probenecid and NAC in TBI, providing justification for further clinical investigation.

Interoperable and explainable machine learning models to predict morbidity and mortality in acute neurological injury in the pediatric intensive care unit: secondary analysis of the TOPICC study.

Background: Acute neurological injury is a leading cause of permanent disability and death in the pediatric intensive care unit (PICU). No predictive model has been validated for critically ill children with acute neurological injury. Objectives: We hypothesized that PICU patients with concern for acute neurological injury are at higher risk for morbidity and mortality, and advanced analytics would derive robust, explainable subgroup models. Methods: We performed a secondary subgroup analysis of the Trichotomous Outcomes in Pediatric Critical Care (TOPICC) study (2011-2013), predicting mortality and morbidity from admission physiology (lab values and vital signs in 6 h surrounding admission). We analyzed patients with suspected acute neurological injury using standard machine learning algorithms. Feature importance was analyzed using SHapley Additive exPlanations (SHAP). We created a Fast Healthcare Interoperability Resources (FHIR) application to demonstrate potential for interoperability using pragmatic data. Results: 1,860 patients had suspected acute neurological injury at PICU admission, with higher morbidity (8.2 vs. 3.4%) and mortality (6.2 vs. 1.9%) than those without similar concern. The ensemble regressor (containing Random Forest, Gradient Boosting, and Support Vector Machine learners) produced the best model, with Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.91 [95% CI (0.88, 0.94)] and Average Precision (AP) of 0.59 [0.51, 0.69] for mortality, and decreased performance predicting simultaneous mortality and morbidity (0.83 [0.80, 0.86] and 0.59 [0.51, 0.64]); at a set specificity of 0.995, positive predictive value (PPV) was 0.79 for mortality, and 0.88 for mortality and morbidity. By comparison, for mortality, the TOPICC logistic regression had AUROC of 0.90 [0.84, 0.93], but substantially inferior AP of 0.49 [0.35, 0.56] and PPV of 0.60 at specificity 0.995. Feature importance analysis showed that pupillary non-reactivity, Glasgow Coma Scale, and temperature were the most contributory vital signs, and acidosis and coagulopathy the most important laboratory values. The FHIR application provided a simulated demonstration of real-time health record query and model deployment. Conclusions: PICU patients with suspected acute neurological injury have higher mortality and morbidity. Our machine learning approach independently identified previously-known causes of secondary brain injury. Advanced modeling achieves improved positive predictive value in this important population compared to published models, providing a stepping stone in the path to deploying explainable models as interoperable bedside decision-support tools.

Clinical translation of noninvasive intracranial pressure sensing with diffuse correlation spectroscopy.

OBJECTIVE: Intracranial pressure (ICP) is an important therapeutic target in many critical neuropathologies. The current tools for ICP measurements are invasive; hence, these are only selectively applied in critical cases where the benefits surpass the risks. To address the need for low-risk ICP monitoring, the authors developed a noninvasive alternative. METHODS: The authors recently demonstrated noninvasive quantification of ICP in an animal model by using morphological analysis of microvascular cerebral blood flow (CBF) measured with diffuse correlation spectroscopy (DCS). The current prospective observational study expanded on this preclinical study by translating the method to pediatric patients. Here, the CBF features, along with mean arterial pressure (MAP) and heart rate (HR) data, were used to build a random decision forest, machine learning model for estimation of ICP; the results of this model were compared with those of invasive monitoring. RESULTS: Fifteen patients (mean age ± SD [range] 9.8 ± 5.1 [0.3-17.5] years; median age [interquartile range] 11 [7.4] years; 10 males and 5 females) who underwent invasive neuromonitoring for any purpose were enrolled. Estimated ICP (ICPest) very closely matched invasive ICP (ICPinv), with a root mean square error (RMSE) of 1.01 mm Hg and 95% limit of agreement of ≤ 1.99 mm Hg for ICPinv 0.01-41.25 mm Hg. When the ICP range (ICPinv 0.01-29.05 mm Hg) was narrowed on the basis of the sample population, both RMSE and limit of agreement improved to 0.81 mm Hg and ≤ 1.6 mm Hg, respectively. In addition, 0.3% of the test samples for ICPinv ≤ 20 mm Hg and 5.4% of the test samples for ICPinv > 20 mm Hg had a limit of agreement > 5 mm Hg, which may be considered the acceptable limit of agreement for clinical validity of ICP sensing. For the narrower case, 0.1% of test samples for ICPinv ≤ 20 mm Hg and 1.1% of the test samples for ICPinv > 20 mm Hg had a limit of agreement > 5 mm Hg. Although the CBF features were crucial, the best prediction accuracy was achieved when these features were combined with MAP and HR data. Lastly, preliminary leave-one-out analysis showed model accuracy with an RMSE of 6 mm Hg and limit of agreement of ≤ 7 mm Hg. CONCLUSIONS: The authors have shown that DCS may enable ICP monitoring with additional clinical validation. The lower risk of such monitoring would allow ICP to be estimated for a wide spectrum of indications, thereby both reducing the use of invasive monitors and increasing the types of patients who may benefit from ICP-directed therapies.

Detecting and Quantifying pADPr In Vivo.

Poly(ADP-ribose) polymerases (PARP) participate in diverse biological processes contributing to cellular homeostasis or exacerbating injury. PARP catalyzes the addition of ADP-ribose molecules (pADPr) to the target proteins, a process termed poly-ADP-ribosylation. Overactivation of PARP - reflected by increased poly-ADP-ribosylation and accumulation of pADPr-modified proteins or free pADPr - contributes to depletion of NAD+ and mitochondrial dysfunction, potentially leading to cell death. Thus, PARP overactivation and increases in free pADPr have been identified as key contributors to the pathobiology of many diseases. In stark contrast, PARP inhibitors are in clinical use in cancer patients where they potentiate cell death induced by chemotherapeutic agents. Accordingly, monitoring PARP-1 activation - responsible for up to 80-90% of cellular pADPr synthesis - by detecting and quantifying pADPr may provide valuable mechanistic insights as well as facilitating therapeutic drug monitoring for PARP inhibitors.Several non-isotopic immunodetection methods for quantifying pADPr are discussed: Western blotting of poly-ADP-ribosylated proteins, cellular localization of pADPr by immunohistochemistry, quantification of pADPr by enzyme-linked immunoassay, and small-scale two-dimensional gel electrophoresis.

Limbic pathway vulnerability associates with neurologic outcome in children after cardiac arrest.

AIM: To analyze whether brain connectivity sequences including diffusion tensor imaging (DTI) and resting state functional magnetic resonance imaging (rsfMRI) identify vulnerable brain regions and networks associated with neurologic outcome after pediatric cardiac arrest. METHODS: Children aged 2 d-17 y with cardiac arrest were enrolled in one of 2 parent studies at a single center. Clinically indicated brain MRI with DTI and rsfMRI and performed within 2 weeks after arrest were analyzed. Tract-wise fractional anisotropy (FA) and axial, radial, and mean diffusivity assessed DTI, and functional connectivity strength (FCS) assessed rsfMRI between outcome groups. Unfavorable neurologic outcome was defined as Pediatric Cerebral Performance Category score 4-6 or change > 1 between 6 months after arrest vs baseline. RESULTS: Among children with DTI (n = 28), 57% had unfavorable outcome. Mean, radial, axial diffusivity and FA of varying direction of magnitude in the limbic tracts, including the right cingulum parolfactory, left cingulum parahippocampal, corpus callosum forceps major, and corpus callosum forceps minor tracts, were associated with unfavorable neurologic outcome (p < 0.05). Among children with rsfMRI (n = 12), 67% had unfavorable outcome. Decreased FCS in the ventromedial and dorsolateral prefrontal cortex, insula, precentral gyrus, anterior cingulate, and inferior parietal lobule were correlated regionally with unfavorable neurologic outcome (p < 0.05 Family-Wise Error corrected). CONCLUSION: Decreased multimodal connectivity measures of paralimbic tracts were associated with unfavorable neurologic outcome after pediatric cardiac arrest. Longitudinal analysis correlating brain connectivity sequences with long term neuropsychological outcomes to identify the impact of pediatric cardiac arrest in vulnerable brain networks over time appears warranted.

Temporal Patterns in Brain Tissue and Systemic Oxygenation Associated with Mortality After Severe Traumatic Brain Injury in Children.

BACKGROUND: Brain tissue hypoxia is an independent risk factor for unfavorable outcomes in traumatic brain injury (TBI); however, systemic hyperoxemia encountered in the prevention and/or response to brain tissue hypoxia may also impact risk of mortality. We aimed to identify temporal patterns of partial pressure of oxygen in brain tissue (PbtO2), partial pressure of arterial oxygen (PaO2), and PbtO2/PaO2 ratio associated with mortality in children with severe TBI. METHODS: Data were extracted from the electronic medical record of a quaternary care children's hospital with a level I trauma center for patients ≤ 18 years old with severe TBI and the presence of PbtO2 and/or intracranial pressure monitors. Temporal analyses were performed for the first 5 days of hospitalization by using locally estimated scatterplot smoothing for less than 1,000 observations and generalized additive models with integrated smoothness estimation for more than 1,000 observations. RESULTS: A total of 138 intracranial pressure-monitored patients with TBI (median 5.0 [1.9-12.8] years; 65% boys; admission Glasgow Coma Scale score 4 [3-7]; mortality 18%), 71 with PbtO2 monitors and 67 without PbtO2 monitors were included. Distinct patterns in PbtO2, PaO2, and PbtO2/PaO2 were evident between survivors and nonsurvivors over the first 5 days of hospitalization. Time-series analyses showed lower PbtO2 values on day 1 and days 3-5 and lower PbtO2/PaO2 ratios on days 1, 2, and 5 among patients who died. Analysis of receiver operating characteristics curves using Youden's index identified a PbtO2 of 30 mm Hg and a PbtO2/PaO2 ratio of 0.12 as the cut points for discriminating between survivors and nonsurvivors. Univariate logistic regression identified PbtO2 < 30 mm Hg, hyperoxemia (PaO2 ≥ 300 mm Hg), and PbtO2/PaO2 ratio < 0.12 to be independently associated with mortality. CONCLUSIONS: Lower PbtO2, higher PaO2, and lower PbtO2/PaO2 ratio, consistent with impaired oxygen diffusion into brain tissue, were associated with mortality in this cohort of children with severe TBI. These results corroborate our prior work that suggests targeting a higher PbtO2 threshold than recommended in current guidelines and highlight the potential use of the PbtO2/PaO2 ratio in the management of severe pediatric TBI.

Endogenous Interleukin-17a Contributes to Normal Spatial Memory Retention but Does Not Affect Early Behavioral or Neuropathological Outcomes after Experimental Traumatic Brain Injury.

Interleukin-17 (IL-17) is a proinflammatory cytokine primarily secreted in the brain by inflammatory T lymphocytes and glial cells. IL-17+ T-helper (Th17) cells are increased in the ipsilateral hemisphere after experimental traumatic brain injury (TBI), and IL-17 levels are increased in serum and brain tissue. We hypothesized that il17a and related gene expression would be increased in brain tissue after TBI in mice and il17a-/- mice would demonstrate neuroprotection versus wild type. The controlled cortical impact (CCI) model of TBI in adult male C57BL6/J mice was used for all experiments. Data were analyzed by analysis of variance (ANOVA) or repeated-measures two-way ANOVA with the Bonferroni correction. A value of p < 0.05 determined significance. Expression of il17a was significantly reduced in the ipsilateral cortex and hippocampus by day 3 after TBI, and expression remained low at 28 days. There were no differences between il17a-/- and il17a+/+ mice in beam balance, Morris water maze performance, or lesion volume after CCI. Surprisingly, naïve il17a -/- mice performed significantly (p = 0.02) worse than naïve il17a+/+ mice on the probe trial. In conclusion, sustained depression of il17a gene expression was observed in brains after TBI in adult mice. Genetic knockout of IL-17 was not neuroprotective after TBI. IL-17a may be important for memory retention in naïve mice.

Healthcare Use in the Year Following Bronchiolitis Hospitalization.

OBJECTIVES: Healthcare utilization after bronchiolitis hospitalization is incompletely understood. We aimed to characterize readmissions and outpatient visits within 1 year after hospital discharge. METHODS: Retrospective multicenter observational cohort study of children under 24-months old admitted with bronchiolitis between January 1, 2010 and December 12, 2019 to the Pediatric Health Information Systems database. A single-center nested subset using linked electronic health records allowed analysis of outpatient visits. RESULTS: There were 308 306 admissions for bronchiolitis among 271 115 patients across 47 hospitals between 2010-2019. The percent of patients readmitted within 30 days after discharge was 6.0% (16 167 of 271 115), and 17.8% (48 332 of 271 115) of patients were readmitted within 1 year. 22.9% (16 919 of 74 001) of patients admitted to an ICU and 26.8% (7865 of 29 378) of patients undergoing mechanical ventilation were readmitted within 1 year. There were 1438 patients with outpatient healthcare data available. There were a median (interquartile range) of 9 (6-13) outpatient visits per patient within 1 year after discharge. Outpatient healthcare use increased for 4 months following bronchiolitis hospitalization compared with previously reported age-matched controls. Higher income, white race, commercial insurance, complex chronic conditions, ICU admission, and mechanical ventilation were associated with higher outpatient utilization. Higher quartiles of outpatient use were associated with readmission for bronchiolitis and all-cause readmissions. CONCLUSIONS: Readmissions in the year after bronchiolitis hospitalization are common, and outpatient healthcare use is increased for 4 months following discharge. Prospective study is needed to track long-term outcomes of infants with bronchiolitis.

Association of EEG and Blood-Based Brain Injury Biomarker Accuracy to Prognosticate Mortality After Pediatric Cardiac Arrest: An Exploratory Study.

BACKGROUND: Evaluate the accuracy of brain-based blood biomarkers neuron-specific enolase (NSE) and S100b and electroencephalography (EEG) features alone and in combination with prognosticate 6-month mortality after pediatric cardiac arrest. We hypothesized that the combination of blood brain-based biomarkers and EEG features would have superior classification accuracy of outcome versus either alone. METHODS: Children (n = 58) aged between 1 week and 17 years admitted to the ICU following cardiac arrest at a tertiary care children's hopital were eligible for this secondary study. Blood NSE and S100b were measured closest to 24 hours after return of spontaneous circulation (ROSC). EEGs closest to 24 hours (median 11, interquartile range [IQR] 6 to 16 h) post-ROSC were evaluated by two epileptologists. EEG grade was informed by background frequency, amplitude, and continuity. Sleep spindles were present or absent. Mortality was assessed at six months post-ROSC. Area under the receiver operator curve (AUC) was performed for individual and combined brain-based biomarkers and EEG features. RESULTS: Children were aged 2.6 (IQR 0.6 to 10.4) years, and 25 (43%) died. Children who died had increased blood NSE (49.7 [28.0 to 63.1] vs 18.2 [9.8 to 31.8] ng/mL) and S100b (0.118 [0.036 to 0.296] vs 0.012 [0.003 to 0.021] ng/mL) and poor (discontinuous or isoelectric) EEG grade (76% vs 33%) more frequently than survivors (P < 0.05). AUC for NSE to predict mortality was 0.789, and was 0.841 when combined with EEG grade and spindles. S100b AUC for mortality was 0.856 and was optimal alone. CONCLUSIONS: In this exploratory study, the combination of brain-based biomarkers and EEG features may provide more accurate prognostication than either test alone after pediatric cardiac arrest.

Association between pediatric TBI mortality and median family income in the United States: A retrospective cohort study.

BACKGROUND: There are regional disparities in pediatric traumatic brain injury (TBI) mortality across the United States, but the factors underlying these differences are unclear. METHODS: We performed a retrospective cross-sectional analysis of the Pediatric Health Information System database including inpatient hospital encounters for children less than 18 years old with a primary diagnosis of TBI between 2010-2019. FINDINGS: Lower median family income was associated with pediatric TBI mortality. Encounters from zip-codes with a median family income of <$20,000 had a 3.1% (29/950) mortality, as opposed to 1.3% (29/2,267) mortality for zip-codes with a median family income of >$80,000 (p = 0.00096). In multivariable logistic regression, every $10,000 of income was associated with an odds ratio of mortality of 0.94 (95% confidence interval 0.90 - 0.98). 82.5% (397/481) of ballistic TBI injuries were caused by a firearm. Lower income was associated with a higher proportion of ballistic TBI injuries (2.5% [24/950] for <$20,000 versus 0.3% [7/2,267] for >$80,000, p < 0.0001). In multivariable logistic regression, ballistic TBI injuries were associated with an odds ratio of mortality of 5.19 (95% confidence interval 4.00 - 6.73). United States regional variation in pediatric TBI mortality was linearly associated with the percentage of ballistic TBI (adjusted r-squared 0.59, p = 0.0097). INTERPRETATION: Children from lower income zip-codes are more likely to sustain a ballistic TBI, and more likely to die. Further work is necessary to determine causal factors underlying these associations and to design interventions that prevent these injuries and/or improve outcomes.

Ascorbate deficiency confers resistance to hippocampal neurodegeneration after asphyxial cardiac arrest in juvenile rats.

BACKGROUND: Asphyxial cardiac arrest (CA) is a significant cause of death and disability in children. Using juvenile Osteogenic disorder Shionogi (ODS) rats that, like humans, do not synthesize ascorbate, we tested the effect of ascorbate deficiency on functional and histological outcome after CA. METHODS: Postnatal day 16-18 milk-fed ODS and wild-type Wistar rats underwent 9-min asphyxial CA (n = 8/group) or sham surgery (n = 4/group). ODS mothers received ascorbate in drinking water to prevent scurvy. Levels of ascorbate and glutathione (GSH) were measured in plasma and hippocampus at baseline and after CA. Neurologic deficit score (NDS) was measured at 3, 24, and 48 h and hippocampal neuronal counts, neurodegeneration, and microglial activation were assessed at day 7. RESULTS: ODS rats showed depletion of plasma and hippocampal ascorbate, attenuated hippocampal neurodegeneration and microglial activation, and increased CA1 hippocampal neuron survival vs. Wistar rats while NDS were similar. Hippocampal GSH levels were higher in ODS vs. Wistar rats at baseline and 10 min, whereas hypoxia-inducible factor-1α levels were higher in Wistar vs. ODS rats at 24 , after CA. CONCLUSION: Ascorbate-deficient juvenile ODS rats appear resistant to neurodegeneration produced by asphyxia CA, possibly related to upregulation of the endogenous antioxidant GSH in brain. IMPACT: Like humans and unlike other rodents, osteogenic disorder Shionogi (ODS) rats do not synthesize ascorbate, and thus may serve as a useful model for studying the role of ascorbate in human disease. Conflicting evidence exists regarding ascorbate's protective versus detrimental effects in animal models and clinical studies. Ascorbate-deficient ODS rats are resistant to neurodegeneration after experimental cardiac arrest.

Serum levels of the cold stress hormones FGF21 and GDF-15 after cardiac arrest in infants and children enrolled in single center therapeutic hypothermia clinical trials.

OBJECTIVE: Fibroblast Growth Factor 21 (FGF21) and Growth Differentiation Factor-15 (GDF-15) are putative neuroprotective cold stress hormones (CSHs) provoked by cold exposure that may be age-dependent. We sought to characterize serum FGF21 and GDF-15 levels in pediatric cardiac arrest (CA) patients and their association with use of therapeutic hypothermia (TH). METHODS: Secondary analysis of serum samples from clinical trials. We measured FGF21 and GDF-15 levels in pediatric patients post-CA and compared levels to both pediatric intensive care (PICU) and healthy controls. Post-CA, we compared normothermia (NT) vs TH (33 °C for 72 h) treated cohorts at < 24 h, 24 h, 48 h, 72 h, and examined the change in CSHs over 72 h. We also assessed association between hospital mortality and initial levels. RESULTS: We assessed 144 samples from 68 patients (27 CA [14 TH, 13 NT], 9 PICU and 32 healthy controls). Median initial FGF21 levels were higher post-CA vs. healthy controls (392 vs. 40 pg/mL, respectively, P < 0.001). Median GDF-15 levels were higher post-CA vs. healthy controls (7,089 vs. 396 pg/mL, respectively, P < 0.001). In the CA group, the median change in FGF21 from PICU day 1-3 (after 72 h of temperature control), was higher in TH vs. NT (231 vs. -20 pg/mL, respectively, P < 0.05), with no difference in GDF-15 over time. Serum GDF-15 levels were higher in CA patients that died vs. survived (19,450 vs. 5,337 pg/mL, respectively, P < 0.05), whereas serum FGF21 levels were not associated with mortality. CONCLUSION: Serum levels of FGF21 and GDF-15 increased after pediatric CA, and FGF21 appears to be augmented by TH.

An Evaluation of Antimicrobial Prescribing and Risk-adjusted Mortality.

The Centers for Disease Control and Prevention recommends tracking risk-adjusted antimicrobial prescribing. Prior studies have used prescribing variation to drive quality improvement initiatives without adjusting for severity of illness. The present study aimed to determine the relationship between antimicrobial prescribing and risk-adjusted ICU mortality in the Pediatric Health Information Systems (PHIS) database, assessed by IBM-Watson risk of mortality. A nested analysis sought to assess an alternative risk model incorporating laboratory data from federated electronic health records. METHODS: Retrospective cohort study of pediatric ICU patients in PHIS between 1/1/2010 and 12/31/2019, excluding patients admitted to a neonatal ICU, and a nested study of PHIS+ from 1/1/2010 to 12/31/2012. Hospital antimicrobial prescription volumes were assessed for association with risk-adjusted mortality. RESULTS: The cohort included 953,821 ICU encounters (23,851 [2.7%] nonsurvivors). There was 4-fold center-level variability in antimicrobial use. ICU antimicrobial use was not correlated with risk-adjusted mortality assessed using IBM-Watson. A risk model incorporating laboratory data available in PHIS+ significantly outperformed IBM-Watson (c-statistic 0.940 [95% confidence interval 0.933-0.947] versus 0.891 [0.881-0.901]; P < 0.001, area under the precision recall curve 0.561 versus 0.297). Risk-adjusted mortality was inversely associated with antimicrobial prescribing in this smaller cohort using both the PHIS+ and Watson models (P = 0.05 and P < 0.01, respectively). CONCLUSIONS: Antimicrobial prescribing among pediatric ICUs in the PHIS database is variable and not associated with risk-adjusted mortality as assessed by IBM-Watson. Expanding existing administrative databases to include laboratory data can achieve more meaningful insights when assessing multicenter antibiotic prescribing practices.

An exploratory assessment of serum biomarkers of post-cardiac arrest syndrome in children.

AIM: We hypothesized that serum biomarkers of inflammation including chemokine, cytokine, pituitary hormones, and growth factors following cardiac arrest in children would independently associate with 6-month neurologic outcome. METHODS: In this prospective observational single center study of children with in-hospital and out-of-hospital cardiac arrest surviving to intensive care unit admission, serum was obtained twice per 24 h period between 0 h and 96 h and once at approximately 196 h post-cardiac arrest. Inflammatory mediators, hormones, and growth factors were analyzed by Luminex Multiplex Bead Immunoassay. We recorded demographics, resuscitation characteristics, and Pediatric Cerebral Performance Category (PCPC) at 6 months. We analyzed the association and area under the curve (AUC) of biomarker levels with favorable (PCPC 1-3) or unfavorable (PCPC 4-6, or >1 increase from baseline) outcome. RESULTS: Forty-two children (50% female; median age of 2.5 [IQR: 0.4-10.2]) were enrolled and 18 (42%) died prior to 6-month follow up. Receiver operator curves for initial levels of ciliary neurotrophic factor (CNTF, AUC 0.84, 95% CI 0.73-0.96, p < 0.001) and interleukin (IL-17, AUC 0.84, 95% CI 0.73-0.97, p < 0.001) best classified favorable versus unfavorable 6-month outcome. In multivariable analysis, initial levels of CNTF and IL-17 remained associated with 6-month PCPC. Peak levels of interferon-γ-inducible protein 10 (IP-10), CNTF, and hepatocyte growth factor (HGF) were also independently associated with outcome. CONCLUSION: Increased serum concentrations of CNTF and IL-17 associated with unfavorable 6-month neurologic outcome of children surviving cardiac arrest. Further investigation of the prognostic utility and roles of CNTF and IL-17 in the pathophysiology of post-cardiac arrest syndrome are warranted. This project is registered with clinicaltrials.gov (NCT00797680) as "Duration of Hypothermia for Neuroprotection after Pediatric Cardiac Arrest: A Randomized, Controlled Trial".

Sustained Dysbiosis and Decreased Fecal Short-Chain Fatty Acids after Traumatic Brain Injury and Impact on Neurologic Outcome.

Traumatic brain injury (TBI) alters microbial populations present in the gut, which may impact healing and tissue recovery. However, the duration and impact of these changes on outcome from TBI are unknown. Short-chain fatty acids (SCFAs), produced by bacterial fermentation of dietary fiber, are important signaling molecules in the microbiota gut-brain axis. We hypothesized that TBI would lead to a sustained reduction in SCFA producing bacteria, fecal SCFAs concentration, and administration of soluble SCFAs would improve functional outcome after TBI. Adult mice (n = 10) had the controlled cortical impact (CCI) model of TBI performed (6 m/sec, 2-mm depth, 50-msec dwell). Stool samples were collected serially until 28 days after CCI and analyzed for SCFA concentration by high-performance liquid chromatography-mass spectrometry/mass spectrometry and microbiome analyzed by 16S gene sequencing. In a separate experiment, mice (n = 10/group) were randomized 2 weeks before CCI to standard drinking water or water supplemented with the SCFAs acetate (67.5 mM), propionate (25.9 mM), and butyrate (40 mM). Morris water maze performance was assessed on post-injury Days 14-19. Alpha diversity remained stable until 72 h, at which point a decline in diversity was observed without recovery out to 28 days. The taxonomic composition of post-TBI fecal samples demonstrated depletion of bacteria from Lachnospiraceae, Ruminococcaceae, and Bacteroidaceae families, and enrichment of bacteria from the Verrucomicrobiaceae family. Analysis from paired fecal samples revealed a reduction in total SCFAs at 24 h and 28 days after TBI. Acetate, the most abundant SCFA detected in the fecal samples, was reduced at 7 days and 28 days after TBI. SCFA administration improved spatial learning after TBI versus standard drinking water. In conclusion, TBI is associated with reduced richness and diversity of commensal microbiota in the gut and a reduction in SCFAs detected in stool. Supplementation of soluble SCFAs improves spatial learning after TBI.

Neurological Complications Acquired During Pediatric Critical Illness: Exploratory "Mixed Graphical Modeling" Analysis Using Serum Biomarker Levels.

OBJECTIVES: Neurologic complications, consisting of the acute development of a neurologic disorder, that is, not present at admission but develops during the course of illness, can be difficult to detect in the PICU due to sedation, neuromuscular blockade, and young age. We evaluated the direct relationships of serum biomarkers and clinical variables to the development of neurologic complications. Analysis was performed using mixed graphical models, a machine learning approach that allows inference of cause-effect associations from continuous and discrete data. DESIGN: Secondary analysis of a previous prospective observational study. SETTING: PICU, single quaternary-care center. PATIENTS: Individuals admitted to the PICU, younger than18 years old, with intravascular access via an indwelling catheter. INTERVENTIONS: None. MEASUREMENTS: About 101 patients were included in this analysis. Serum (days 1-7) was analyzed for glial fibrillary acidic protein, ubiquitin C-terminal hydrolase-L1, and alpha-II spectrin breakdown product 150 utilizing enzyme-linked immunosorbent assays. Serum levels of neuron-specific enolase, myelin basic protein, and S100 calcium binding protein B used in these models were reported previously. Demographic data, use of selected clinical therapies, lengths of stay, and ancillary neurologic testing (head CT, brain MRI, and electroencephalogram) results were recorded. The Mixed Graphical Model-Fast-Causal Inference-Maximum algorithm was applied to the dataset. MAIN RESULTS: About 13 of 101 patients developed a neurologic complication during their critical illness. The mixed graphical model identified peak levels of the neuronal biomarker neuron-specific enolase and ubiquitin C-terminal hydrolase-L1, and the astrocyte biomarker glial fibrillary acidic protein to be the direct causal determinants for the development of a neurologic complication; in contrast, clinical variables including age, sex, length of stay, and primary neurologic diagnosis were not direct causal determinants. CONCLUSIONS: Graphical models that include biomarkers in addition to clinical data are promising methods to evaluate direct relationships in the development of neurologic complications in critically ill children. Future work is required to validate and refine these models further, to determine if they can be used to predict which patients are at risk for/or with early neurologic complications.