Acute Neurologic Injury in Children Admitted to the Cardiac Intensive Care Unit

Interoperable Models for Identifying Critically Ill Children at Risk of Neurologic Morbidity

Importance

Decreasing mortality in the field of pediatric critical care medicine has shifted practicing clinicians' attention to preserving patients' neurodevelopmental potential as a main objective. Earlier identification of critically ill children at risk for incurring neurologic morbidity would facilitate heightened surveillance that could lead to timelier clinical detection, earlier interventions, and preserved neurodevelopmental trajectory.

Objectives

To develop machine-learning models for identifying acquired neurologic morbidity in hospitalized pediatric patients with critical illness and assess correlation with contemporary serum-based, brain injury-derived biomarkers.

Design, Setting, and Participants

This prognostic study used data from all children admitted to a quaternary pediatric intensive care unit in a large, freestanding children's hospital in Western Pennsylvania between January 1, 2010, and December 31, 2022. External model validation used data from children admitted between January 1, 2018, and December 31, 2023, to a quaternary pediatric intensive care unit in a large, freestanding children's hospital that serves as a referral center for the 5-state region of Washington, Wyoming, Alaska, Montana, and Idaho.

Exposures

Critical illness.

Main Outcomes and Measures

The outcome was neurologic morbidity, defined according to a computable, composite definition at the development site or an order for neurocritical care consultation at the validation site. Models were developed using varying time windows for temporal feature engineering and varying censored time horizons between the last feature and the identified neurologic morbidity. A generalizable model created at the development site was optimized and assessed at an external validation site. Correlation was assessed between development site model predictions and measurements of brain biomarkers from a convenience cohort.

Results

After exclusions, there were 18 568 encounters from 2010 to 2022 in the development site generalizable model cohort (median age, 70 [IQR, 18-161] months; 8325 [45%] female). There were 6825 encounters from 2018 to 2021 at the external validation site (median age, 96 [IQR 18-171] months; 3159 [46%] female). A generalizable extreme gradient boosted model with a 24-hour time horizon and 48-hour feature engineering window demonstrated an F1 score of 0.37 (95% CI, 0.33-0.40), area under the receiver operating characteristics curve of 0.81 (95% CI, 0.78-0.83), and number needed to alert of 4 at the validation site. After recalibration at the validation site, the Brier score was 0.04. Serum levels of the brain injury biomarker glial fibrillary acidic protein significantly correlated with model output (rs = 0.34; P = .007).

Conclusions and Relevance

This prognostic study of prediction models for detecting neurologic morbidity in critically ill children demonstrated a well-performing ensemble of models with biomolecular corroboration. Prospective assessment and refinement of biomarker-coupled risk models in pediatric critical illness are warranted.

Neuromonitoring and follow-up in patients with congenital heart disease in Spain

INTRODUCTION

There is evidence of the high incidence of neurological abnormalities in patients with congenital heart disease (CHD). Despite this, perioperative neuromonitoring strategies and long-term follow-up protocols are not standardized in Spain.

OBJECTIVE

The aim of our study was to describe current clinical practice in neuromonitoring, neuroimaging and neurodevelopmental follow-up in patients with CHD in Spanish hospitals that perform paediatric cardiac surgery (PCS).

MATERIAL AND METHOD

We conducted a survey by adapting a questionnaire originally developed by the European Association Brain and Congenital Heart Disease Consortium to collect data on aspects such as the implementation of perioperative neuromonitoring and the type of neuroimaging techniques and neurological follow-up performed. The questionnaire was distributed to the 19 Spanish hospitals that perform PCS.

RESULTS

We received responses from 17 centres. Eighty-eight percent performed some type of preoperative neuroimaging and 81% postoperative monitoring. The most widely used technique was transfontanellar sonography. Fifty-six percent of the centres used some form of intraoperative neuromonitoring, most frequently near-infrared spectroscopy. Nineteen percent had an established protocol for the follow-up of these patients and 13% were in the process of developing it.

CONCLUSIONS

There is considerable heterogeneity in neuromonitoring, neuroimaging and neurologic follow-up practices in the management of patients with CHD in hospitals that perform PCS in Spain. These findings highlight the need to pursue a consensus in order to standardise neuromonitoring and neurologic follow-up strategies in children with CHD in Spain.

Development, External Validation, and Biomolecular Corroboration of Interoperable Models for Identifying Critically Ill Children at Risk of Neurologic Morbidity

Importance

Declining mortality in the field of pediatric critical care medicine has shifted practicing clinicians' attention to preserving patients' neurodevelopmental potential as a main objective. Earlier identification of critically ill children at risk for incurring neurologic morbidity would facilitate heightened surveillance that could lead to timelier clinical detection, earlier interventions, and preserved neurodevelopmental trajectory.

Objective

Develop machine-learning models for identifying acquired neurologic morbidity while hospitalized with critical illness and assess correlation with contemporary serum-based, brain injury-derived biomarkers.

Design

Retrospective cohort study.

Setting

Two large, quaternary children's hospitals.

Exposures

Critical illness.

Main Outcomes and Measures

The outcome was neurologic morbidity, defined according to a computable, composite definition at the development site or an order for neurocritical care consultation at the validation site. Models were developed using varying time windows for temporal feature engineering and varying censored time horizons prior to identified neurologic morbidity. Optimal models were selected based on F1 scores, cohort sizes, calibration, and data availability for eventual deployment. A generalizable created at the development site was assessed at an external validation site and optimized with spline recalibration. Correlation was assessed between development site model predictions and measurements of brain biomarkers from a convenience cohort.

Results

After exclusions there were 14,222-25,171 encounters from 2010-2022 in the development site cohorts and 6,280-6,373 from 2018-2021 in the validation site cohort. At the development site, an extreme gradient boosted model (XGBoost) with a 12-hour time horizon and 48-hour feature engineering window had an F1-score of 0.54, area under the receiver operating characteristics curve (AUROC) of 0.82, and a number needed to alert (NNA) of 2. A generalizable XGBoost model with a 24-hour time horizon and 48-hour feature engineering window demonstrated an F1-score of 0.37, AUROC of 0.81, AUPRC of 0.51, and NNA of 4 at the validation site. After recalibration at the validation site, the Brier score was 0.04. Serum levels of the brain injury biomarker glial fibrillary acidic protein measurements significantly correlated with model output (rs=0.34; P=0.007).

Conclusions and Relevance

We demonstrate a well-performing ensemble of models for predicting neurologic morbidity in children with biomolecular corroboration. Prospective assessment and refinement of biomarker-coupled risk models in pediatric critical illness is warranted.

Cerebral autoregulation monitoring using the cerebral oximetry index after neonatal cardiac surgery: A single-center retrospective cohort study

OBJECTIVE

To investigate whether cerebral autoregulation is impaired after neonatal cardiac surgery and whether changes in autoregulation metrics are associated with different congenital heart defects or the incidence of postoperative neurologic events.

METHODS

This is a retrospective observational study of neonates undergoing monitoring during the first 72 hours after cardiac surgery. Archived data were processed to calculate the cerebral oximetry index (COx) and derived metrics. Acute neurologic events were identified by an electronic medical record review. The Skillings-Mack test and the Wilcoxon signed-rank test were used to analyze the evolution of autoregulation metrics over time; the Mann-Whitney U test was used for comparison between groups.

RESULTS

We included 28 neonates, 7 (25%) with hypoplastic left heart syndrome and 21 (75%) with transposition of the great arteries. Overall, the median percentage of time spent with impaired autoregulation, defined as percentage of time with a COx >0.3, was 31.6% (interquartile range, 21.1%-38.3%). No differences in autoregulation metrics between different cardiac defects subgroups were observed. Seven patients (25%) experienced a postoperative acute neurologic event. Compared to the neonates without an acute neurologic event, those with an acute neurologic event had a higher COx (0.16 vs 0.07; P = .035), a higher percentage of time with a COx >0.3 (39.4% vs 29.2%; P = .017), and a higher percentage of time with a mean arterial pressure below the lower limit of autoregulation (13.3% vs 6.9%; P = .048).

CONCLUSIONS

COx monitoring after cardiac surgery allowed for the detection of impaired cerebral autoregulation, which was more frequent in neonates with postoperative acute neurologic events.

Neurodevelopmental outcomes in congenital heart disease: Usefulness of biomarkers of brain injury

INTRODUCTION

At present, neurodevelopmental abnormalities are the most frequent type of complication in school-aged children with congenital heart disease (CHD). We analysed the incidence of acute neurologic events (ANEs) in patients with operated CHD and the usefulness of neuromarkers for the prediction of neurodevelopment outcomes.

METHODS

Prospective observational study in infants with a prenatal diagnosis of CHD who underwent cardiac surgery in the first year of life. We assessed the following variables: (1) serum biomarkers of brain injury (S100B, neuron-specific enolase) in cord blood and preoperative blood samples; (2) clinical and laboratory data from the immediate postnatal and perioperative periods; (3) treatments and complications; (4) neurodevelopment (Bayley-III scale) at age 2 years.

RESULTS

the study included 84 infants with a prenatal diagnosis of CHD who underwent cardiac surgery in the first year of life. Seventeen had univentricular heart, 20 left ventricular outflow obstruction and 10 genetic syndromes. The postoperative mortality was 5.9% (5/84) and 10.7% (9/84) patients experienced ANEs. The mean overall Bayley-III scores were within the normal range, but 31% of patients had abnormal scores in the cognitive, motor or language domains. Patients with genetic syndromes, ANEs and univentricular heart had poorer neurodevelopmental outcomes. Elevation of S100B in the immediate postoperative period was associated with poorer scores.

CONCLUSIONS

children with a history of cardiac surgery for CHD in the first year of life are at risk of adverse neurodevelopmental outcomes. Patients with genetic syndromes, ANEs or univentricular heart had poorer outcomes. Postoperative ANEs may contribute to poorer outcomes. Elevation of S100B levels in the postoperative period was associated with poorer neurodevelopmental outcomes at 2 years. Studies with larger samples and longer follow-ups are needed to define the role of these biomarkers of brain injury in the prediction of neurodevelopmental outcomes in patients who undergo surgery for management of CHD.

Early Functional Status Change After Cardiopulmonary Resuscitation in a Pediatric Heart Center: A Single-Center Retrospective Study

Children with cardiac disease are at significantly higher risk for in-hospital cardiac arrest (CA) compared with those admitted without cardiac disease. CA occurs in 2-6% of patients admitted to a pediatric intensive care unit (ICU) and 4-6% of children admitted to the pediatric cardiac-ICU. Treatment of in-hospital CA with cardiopulmonary resuscitation (CPR) results in return of spontaneous circulation in 43-64% of patients and survival rate that varies from 20 to 51%. We aimed to investigate the change in functional status of survivors who experienced an in-hospital CA using the functional status scale (FSS) in our heart center by conducting a retrospective study of all patients 0-18 years who experienced CA between June 2015 and December 2020 in a free-standing university-affiliated quaternary children's hospital. Of the 165 CA patients, 61% (n = 100) survived to hospital discharge. The non-survivors had longer length from admission to CA, higher serum lactate levels peri-CA, and received higher number of epinephrine doses. Using FSS, of the survivors, 26% developed new morbidity, and 9% developed unfavorable outcomes. There was an association of unfavorable outcomes with longer CICU-LOS and number of epinephrine doses given. Sixty-one-percent of CA patients survived to hospital discharge. Of the survivors, 26% developed new morbidity and 91% had favorable outcomes. Future multicenter studies are needed to help better identify modifiable risk factors for development of poor outcomes and help improve outcomes of this fragile patient population.

Incidence of Acute Neurological Events in Neonates and Infants Undergoing Cardiac Surgery Using a High-Hematocrit/ High-Flow Bypass Strategy

Background: The incidence of new acute neurological injury occurring in neonates and infants during cardiac surgery utilizing cardiopulmonary bypass is reportedly 3% to 5%. In 2013, we adopted a high flow rate, and high hematocrit bypass strategy, and sought to assess the incidence of early neurological injuries associated with this strategy. Methods: Neonates and infants undergoing cardiopulmonary bypass between January 2013 and December 2019 (n  =  714) comprise the study. Adverse neurological events (ANEs) were defined as any abnormality of pupils, delayed awakening, seizures, focal neurological deficits, concerns prompting neurological consultation, or any abnormality on neurological imaging in the postoperative period. Our bypass strategy included a high flow rate (150-200 mL/kg/min), without reduction of flow rates during cooling and maintaining a target hematocrit on bypass > 32% with a terminal hematocrit of > 42%. Results: Median weight at the time of the procedure was 4.6 kg (IQR 3.6-6.1 kg) with the smallest patient weighing 1.36 kg. There were 46 premature patients (6.4%). There were 149 patients (20.9%) patients who underwent deep hypothermic circulatory arrest with a median time of 26 min (IQR 21-41 min). Hospital mortality was 3.5% (24/714, 95% CI: 2.28-5.13). The incidence of neurological events as defined above was 0.84% (6/714, 95% CI: 0.31-1.82). Neurological imaging identified ischemic injury in 4 patients and intraventricular hemorrhage in 2. Conclusions: High flow/high hematocrit bypass strategy was associated with a low incidence of ANE in this vulnerable population.

Variation in hospital costs and resource utilisation after congenital heart surgery

BACKGROUND

Children undergoing cardiac surgery have overall improving survival, though they consume substantial resources. Nationwide inpatient cost estimates and costs at longitudinal follow-up are lacking.

METHODS

Retrospective cohort study of children <19 years of age admitted to Pediatric Health Information System administrative database with an International Classification of Diseases diagnosis code undergoing cardiac surgery. Patients were grouped into neonates (≤30 days of age), infants (31-365 days of age), and children (>1 year) at index procedure. Primary and secondary outcomes included hospital stay and hospital costs at index surgical admission and 1- and 5-year follow-up.

RESULTS

Of the 99,670 cohort patients, neonates comprised 27% and had the highest total hospital costs, though daily hospital costs were lower. Mortality declined (5.6% in 2004 versus 2.5% in 2015, p < 0.0001) while inpatient costs rose (5% increase/year, p < 0.0001). Neonates had greater index diagnosis complexity, greater inpatient costs, required the greatest ICU resources, pharmacotherapy, and respiratory therapy. We found no relationship between hospital surgical volume, mortality, and hospital costs. Neonates had higher cumulative hospital costs at 1- and 5-year follow-up compared to infants and children.

CONCLUSIONS

Inpatient hospital costs rose during the study period, driven primarily by longer stay. Neonates had greater complexity index diagnosis, required greater hospital resources, and have higher hospital costs at 1 and 5 years compared to older children. Surgical volume and in-hospital mortality were not associated with costs. Further analyses comprising merged clinical and administrative data are necessary to identify longer stay and cost drivers after paediatric cardiac surgery.

Association Between Early Change in Arterial Carbon Dioxide Tension and Outcomes in Neonates Treated by Extracorporeal Membrane Oxygenation

The primary objective was to investigate the association between partial pressure of carbon dioxide (PaCO2) change after extracorporeal membrane oxygenation (ECMO) initiation and neurologic outcome in neonates treated for respiratory failure. A retrospective analysis of the Extracorporeal Life Support Organization (ELSO) database including newborns supported by ECMO for respiratory indication during 2015-2020. The closest Pre-ECMO (Pre-ECMO PaCO2) and at 24 hours after ECMO initiation (H24 PaCO2) PaCO2 values allowed to calculate the relative change in PaCO2 (Rel Δ PaCO2 = [H24 PaCO2 - Pre-ECMO PaCO2]/Pre-ECMO PaCO2). The primary outcome was the onset of any acute neurologic event (ANE), defined as cerebral bleeding, ischemic stroke, clinical or electrical seizure, or brain death during ECMO. We included 3,583 newborns (median age 1 day [interquartile range {IQR}, 1-3], median weight 3.2 kg [IQR, 2.8-3.6]) from 198 ELSO centers. The median Rel Δ PaCO2 value was -29.9% [IQR, -46.2 to -8.5]. Six hundred nine (17%) of them had ANE (405 cerebral bleedings, 111 ischemic strokes, 225 seizures, and 6 brain deaths). Patients with a decrease of PaCO2 > 50% were more likely to develop ANE than others (odds ratio [OR] 1.78, 95% confidence interval [CI], 1.31-2.42, p < 0.001). This was still observed after adjustment for all clinically relevant confounding factors (adjusted OR 1.94, 95% CI, 1.29-2.92, p = 0.001). A significant decrease in PaCO2 after ECMO start is associated with ANE among neonates requiring ECMO for respiratory failure. Cautious PaCO2 decrease should be considered after start of ECMO therapy.

Impact of Comorbid Prematurity and Congenital Anomalies: A Review

Preterm infants are more likely to be born with congenital anomalies than those who are born at full-term. Conversely, neonates born with congenital anomalies are also more likely to be born preterm than those without congenital anomalies. Moreover, the comorbid impact of prematurity and congenital anomalies is more than cumulative. Multiple common factors increase the risk of brain injury and neurodevelopmental impairment in both preterm babies and those born with congenital anomalies. These include prolonged hospital length of stay, feeding difficulties, nutritional deficits, pain exposure and administration of medications including sedatives and analgesics. Congenital heart disease provides a well-studied example of the impact of comorbid disease with prematurity. Impaired brain growth and maturity is well described in the third trimester in this population; the immature brain is subsequently more vulnerable to further injury. There is a colinear relationship between degree of prematurity and outcome both in terms of mortality and neurological morbidity. Both prematurity and relative brain immaturity independently increase the risk of subsequent neurodevelopmental impairment in infants with CHD. Non-cardiac surgery also poses a greater risk to preterm infants despite the expectation of normal in utero brain growth. Esophageal atresia, diaphragmatic hernia and abdominal wall defects provide examples of congenital anomalies which have been shown to have poorer neurodevelopmental outcomes in the face of prematurity, with associated increased surgical complexity, higher relative cumulative doses of medications, longer hospital and intensive care stay and increased rates of feeding difficulties, compared with infants who experience either prematurity or congenital anomalies alone.

Preoperative Care of Neonates With Congenital Heart Disease

Congenital heart disease (CHD) is one of the most common types of birth defects, with 40,000 newborns diagnosed yearly in the United States. This article describes: (1) four common heart defects seen in neonatal intensive care units, (2) the typical medical/nursing care of these neonates, and (3) common surgical management for the defects. Hypoplastic left heart syndrome, dextro-transposition of the great arteries, tetralogy of Fallot, and pulmonary atresia with intact ventricular septum are four common types of CHD requiring NICU admission. Knowledge of these defects will help nurses to appropriately manage and treat neonates with these types of CHD.

Early Neurologic Complications and Long-term Neurologic Outcomes of Extracorporeal Membrane Oxygenation Performed in Children

BACKGROUND

We aimed at evaluating acute neurologic complications (ANC) and clinical outcome at a 2-year follow-up in children after extracorporeal membrane oxygenation (ECMO).

METHODS

We conducted a single-center, retrospective review of our patient cohort aged between 1 month and 18 years at the time of ECMO support (between June 2014 to January 2017). Outcome analysis included ANC and their clinical consequences.The Pediatric Overall Performance Category (POPC) and Pediatric Cerebral Performance Category (PCPC) were used for neurologic assessment performed at discharge from the hospital and at 2nd year follow-up.

RESULTS

There were 35 children who required ECMO. The median ECMO time was 9 days (range 2-32 days). Decannulation from ECMO was achieved in 68.6% of patients, and overall, 42.8% survived (15 patients), The incidence of ANC in the surviving patients was 40% (6 children). ANC were intracranial hemorrhage, seizures, cerebral infarction, which occurred in one, two and three of the 15 surviving patients respectively (6.6, 13.3 and 20%). A higher rate of organ failure was related to death (p=0.043), whereas duration on ECMO was a risk factor for the development of ANC (p<0.05). At hospital discharge, the 14 patients evaluated had normal development or -mild disability in 73.2%, and at the 2-year follow-up, 93.4% had these scores.

CONCLUSION

Children who receive ECMO have a risk to develop ANC, which was related to the length of ECMO treatment, while survival was related to less organ failure, Long-term neurological outcome was good in our patient cohort.

Mechanisms of perioperative brain damage in children with congenital heart disease

Congenital heart disease, particularly cyanotic congenital heart disease (CCHD), may lead to a neurodevelopmental delay through central nervous system injury, more unstable central nervous system development, and increased vulnerability of the nervous system. Neurodevelopmental disease is the most serious disorder of childhood, affecting the quality of life of children and their families. Therefore, the monitoring and optimization of nerve damage treatments are important. The factors contributing to neurodevelopmental disease are primarily related to preoperative, intraoperative, postoperative, genetic, and environmental causes, with intraoperative causes being the most influential. Nevertheless, few studies have examined these factors, particularly the influencing factors during early postoperative care. Children with congenital heart disease may experience brain damage during early heart intensive care due to unstable haemodynamics and total body oxygen transfer, particularly early postoperative inflammatory reactions in the brain, blood glucose levels, and other factors that potentially influence long-term neural development. This study analyses the forms of structural and functional brain damage in the early postoperative period, along with the recent evolution of research on its contributing factors.

Continuous Monitoring of Cerebral Autoregulation in Children Supported by Extracorporeal Membrane Oxygenation: A Pilot Study

OBJECTIVE

Cerebral autoregulation (CA) impairment may pose a risk factor for neurological complications among children supported by extracorporeal membrane oxygenation (ECMO). Our first objective was to investigate the feasibility of CA continuous monitoring during ECMO treatment and to describe its evolution over time. The second objective was to analyze the association between CA impairment and neurological outcome.

DESIGN

Observational prospective study.

PATIENTS AND SETTING

Twenty-nine children treated with veno-arterial or veno-venous ECMO in the PICU of Nantes University Hospital, France, and the PICU of the IRCCS Giannina Gaslini Institute in Genoa, Italy.

MEASUREMENTS

A correlation coefficient between the variations of regional cerebral oxygen saturation and the variations of mean arterial blood pressure (MAP) was calculated as an index of CA (cerebral oxygenation reactivity index, COx). A COx > 0.3 was considered as indicative of autoregulation impairment. COx-MAP plots were investigated allowing determining optimal MAP (MAPopt) and limits of autoregulation: lower (LLA) and upper (ULA). Neurological outcome was assessed by the onset of an acute neurological event (ANE) after ECMO start.

RESULTS

We included 29 children (median age 84 days, weight 4.8 kg). MAPopt, LLA, and ULA were detected in 90.8% (84.3-93.3) of monitoring time. Mean COx was significantly higher during day 1 of ECMO compared to day 2 [0.1 (0.02-0.15) vs. 0.01 (- 0.05 to 0.1), p = 0.002]. Twelve children experienced ANE (34.5%). The mean COx and the percentage of time spent with a COx > 0.3 were significantly higher among ANE+ compared to ANE- patients [0.09 (0.01-0.23) vs. 0.04 (- 0.02 to 0.06), p = 0.04 and 33.3% (24.8-62.1) vs. 20.8% (17.3-23.7) p = 0.001]. ANE+ patients spent significantly more time with MAP below LLA [17.2% (6.5-32.9) vs. 5.6% (3.6-9.9), p = 0.02] and above ULA [13% (5.3-38.4) vs. 4.2% (2.7-7.4), p = 0.004], respectively.

CONCLUSION

CA assessment is feasible in pediatric ECMO. The first 24 h following ECMO represents the most critical period regarding CA. Impaired autoregulation is significantly more severe among patients who experience ANE.

The Congenital Heart Disease Brain: Prenatal Considerations for Perioperative Neurocritical Care

Altered brain development has been highlighted as an important contributor to adverse neurodevelopmental outcomes in children with congenital heart disease. Abnormalities begin prenatally and include micro- and macrostructural disturbances that lead to an altered trajectory of brain growth throughout gestation. Recent progress in fetal imaging has improved understanding of the neurobiological mechanisms and risk factors for impaired fetal brain development. The impact of the prenatal environment on postnatal neurological care has also gained increased focus. This review summarizes current data on the timing and pattern of altered prenatal brain development in congenital heart disease, the potential mechanisms of these abnormalities, and the association with perioperative neurological complications.

Electrographic Seizures and Brain Injury in Children Requiring Extracorporeal Membrane Oxygenation

BACKGROUND

Single-center studies suggest that up to 30% of children undergoing extracorporeal membrane oxygenation have electrographic seizures. The aim of this study was to characterize seizure prevalence, seizure risk factors, and brain injury prevalence in the pediatric extracorporeal membrane oxygenation population at a tertiary care children's hospital.

METHODS

We performed a retrospective systematic review of medical records for 86 consecutive children (neonates to age 21 years) who received Neurology consults and continuous video electroencephalography while undergoing extracorporeal membrane oxygenation from November 2015 to September 2018.

RESULTS

Continuous video electroencephalography was initiated in 86 of 170 children who required extracorporeal membrane oxygenation (51%); median duration of continuous vodeo electroencephalography was four days. Nineteen of 86 had electroencephalography-confirmed seizures (22%). Sixteen of 19 had seizures within the first 48 hours on continuous video electroencephalography. Interictal epileptiform discharges were a significant risk factor for seizures; 89% of those with seizures versus 46% of those without had interictal epileptiform discharges (P < 0.001, Fisher's exact test). Children with seizures also had higher pericannulation lactate (median 6.7, interquartile range of 4.3 to 19.0 for those with, and median 4.0, interquartile range of 2.0 to 7.3 for those without; P = 0.02, Mann-Whitney U test). Seizures were associated with hemorrhage on neuroimaging (68% of children with seizures had intracranial hemorrhage versus 34% of those without, P = 0.01, chi-square test).

CONCLUSION

Approximately half the children undergoing extracorporeal membrane oxygenation received continuous video electroencephalography during the study period, and 22% had seizures. Interictal epileptiform discharges and elevated pre-extracorporeal membrane oxygenation lactate levels were risk factors for seizures; seizures were associated with intracranial hemorrhage.