LEARNING OBJECTIVES

Introduction

Continuous EEG (cEEG) monitoring is a key component of the neurologic assessment of critically ill children. The neurologic examination of children in the intensive care unit is often limited by the presence of coma and by the administration of sedative and neuromuscular blocking agents. Electrographic seizures (ES) are common in critically ill children, and in the absence of cEEG monitoring, these would remain largely undetected, as a significant proportion of seizures in critically ill children are subclinical or electrographic only.^1–3 The contribution of seizures to ongoing cerebral injury remains unclear; however, there is an association between outcome (i.e., in-hospital mortality, poor neurodevelopmental outcome, and epilepsy) and electrographic seizures in subpopulations of critically ill children.^1–4 In this chapter, we will define ES and electrographic status epilepticus (ESE), identify patients at high risk for ES, and define the minimum duration of continuous EEG monitoring recommended to identify seizures in critically ill children.

Definition of Electrographic Seizure and Electrographic Status Epilepticus

An electrographic seizure is defined by the presence of evolution, which manifests as at least two sequential changes in frequency, morphology, or location.⁵ ES has been defined as a paroxysmal event different from the background, lasting longer than 10 seconds (or shorter if associated with a clinical change) with temporo-spatial evolution in morphology, frequency, and amplitude, and with a plausible electrographic field (Figures 4.1–4.4).¹ Unequivocal ES is further defined as generalized spike-wave discharges at 3/second or faster (Figure 4.5) and clearly evolving discharges of any type that reach a frequency >4/second whether focal or generalized (Figure 4.6).⁵ ES, in turn, may be either electrographic only, without a discernible clinical correlate, or electroclinical, in which case a clinical manifestation can be identified. Electroclinical seizures may be subtle or clinically quite obvious.

Electrographic status epilepticus refers to a single seizure lasting 30 minutes or electrographic seizure activity that when summated occupies more than 50% of a one-hour epoch (Figure 4.7). This definition applies to both convulsive (CSE) and nonconvulsive status epilepticus (NCSE), with the caveat that NCSE is electrographic only.⁶ The International League Against Epilepsy (ILAE) offers a more conceptual definition of status epilepticus focusing on two time points: t1 (the time beyond which a seizure is considered abnormally prolonged and unlikely to stop on its own) and t2 (the time beyond which there is a risk of long-term consequences).⁷ t1 and t2 vary based on the type of status epilepticus under consideration; for generalized convulsive status epilepticus, t1 = 5 minutes and t2 = 30 minutes. t1 and t2 are longer for focal status epilepticus and absence status. This definition does not apply to neonates, which are beyond the scope of this chapter.

Guidelines and Consensus Statements

American Clinical Neurophysiology Society

The American Clinical Neurophysiology Society’s consensus statement on continuous EEG in critically ill adults and children outlines indications for the use of cEEG, with the most common being the detection of electrographic-only seizures in high-risk populations and characterization of clinical events concerning for seizure.⁸ Indications for cEEG are listed in Table 4.1. The guideline recommends a minimum of 24 hours of monitoring to identify 114electrographic seizures in high-risk populations and 24 hours of cEEG from the last electrographic seizure to monitor for recurrence.⁸

Neurocritical Care Society

The Neurocritical Care Society recommends that cEEG be initiated within 1 hour of status epilepticus if continued seizure activity is suspected. The guideline recommends 48 hours of EEG monitoring in comatose patients to identify seizures.⁹

Seizure Prevalence in Critically Ill Children

Pediatric Intensive Care Unit

Electrographic seizures have been reported in 10% to 40% of children who undergo a clinically indicated EEG in the pediatric intensive care unit.^{1–4,10–22} In a multicenter retrospective cohort, 30% of patients undergoing video EEG monitoring were found to have ES, 38% of whom had electrographic status epilepticus (ESE) and 36% of whom had exclusively nonconvulsive 115seizures.¹ Within this cohort, the following subgroups had a seizure prevalence of 30% or more: sepsis (58%) (Figure 4.8), epilepsy (48%) (Figures 4.9 and 4.10), brain malformations (38%) (Figure 4.11), stroke (33%) (Figures 4.12 and 4.13), autoimmune/inflammatory disorders (33%) (Figure 4.14), and traumatic brain injury (30%) (Figures 4.15 and 4.16).¹ Electrographic seizures were more likely to occur in younger children (median age of 23 months in those with seizure compared to 42 months in those without seizure), suggesting younger age as a risk factor for seizure occurrence.¹ In a recent study of 719 critically ill children with acute encephalopathy undergoing cEEG in the pediatric intensive care unit, variables associated with increased ES risk included age, etiology of encephalopathy, clinical seizures prior to cEEG initiation, EEG background category, and the presence of epileptiform discharges.²³ In a study of 98 children presenting with CSE, risk factors for ES after CSE were found to include a prior diagnosis of epilepsy and the presence of interictal epileptiform discharges.²⁴ Other background features found to correlate with development of ES include lateralized periodic discharges and absence of background reactivity.¹¹ The majority of studies of seizure prevalence and risk factors have included patient cohorts that are heterogeneous in terms of etiology. Many studies to date have also been limited by selection bias, in that in most cases, a clinical concern for seizure prompts cEEG, which may artificially inflate prevalence estimates in those populations that have not been monitored in a systematic, unbiased fashion. A review of the prevalence of ES in individual ICU cohorts is presented in Table 4.2.

Specific High-Risk Populations

CONGENITAL HEART DISEASE

Neonates and infants who undergo repair of congenital heart disease are at risk for ES (Figure 4.17). The incidence of postoperative clinical seizure activity varied widely among early studies, ranging from 5% to 19%.^25,26 An early prospective study using both clinical and EEG criteria for seizure diagnosis found that among patients who underwent surgery for D-transposition of the great arteries, 6% had clinical seizures and 20% had electrographic seizures in the postoperative period.²⁷ Seizures were more common in patients who underwent deep hypothermic circulatory arrest as compared to low flow bypass.²⁷ A later prospective study using EEG criteria for seizure diagnosis identified seizures in the postoperative period in 11% of patients with single ventricle physiology and 18% of patients with hypoplastic left heart syndrome.²⁸ This study also showed that deep hypothermic circulatory arrest for greater than 40 minutes was associated with both clinical and electrographic seizures.²⁸ Similarly, a prospective study using cEEG criteria for the diagnosis of seizure in patients with all types of congenital heart disease during the first 48 hours after surgery identified seizures in 12% of patients.²⁹ A study focused on neonates found that 8% of patients monitored using cEEG had ES after congenital heart disease surgery, with the majority being electrographic-only; there was also a high rate of ESE in this study.³⁰ Recent studies have also indicated that seizures comprise 80% of acute neurologic complications observed after surgery for congenital heart disease.³¹

ECMO

Patients on ECMO have also been found to be at risk of electrographic seizures (Figure 4.18). The initial study focused on this population noted that 21% of children being treated with extracorporeal cardiac life support had seizures on continuous EEG monitoring, with half having nonconvulsive status epilepticus.³² Subsequent studies have shown that 18% to 23% of patients on ECMO have electrographic seizures; these studies have additionally shown an association between seizures and mortality in this population.^33,34

Recommended Duration of EEG Monitoring in Critically Ill Children

Timing of seizure onset from the start of cEEG is similar across pediatric subpopulations with a few notable exceptions. In patients with acute encephalopathy monitored on video EEG and found to have seizures, 97% had their first electrographic seizure within the first 24 hours of recording.²⁰ Similarly, in a cohort of children with nonconvulsive status epilepticus, 100% of seizures were detected within the first 24 hours of continuous EEG monitoring.²¹ When including all indications for EEG monitoring in the neonatal and pediatric ICU, 80% of seizures occur within the first 24 hours and 87% within the first 48 hours of the initiation of cEEG monitoring.¹¹

As noted above, children undergoing repair of congenital heart disease are at high risk for seizure in the postoperative period. The first electrographic seizures have been reported to occur between 10 and 36 hours postoperatively 117with a mean time of onset of 21 ± 6 hours.²⁹ This suggests that strict adherence to the ACNS guideline for a minimum duration of 24 hours of cEEG to identify seizures may not be sufficient in this population.

The initial study of electrographic seizures in patients on ECMO found that all seizures were captured within the first 24 hours of cEEG, with 50% of those being captured within the first hour of recording.³² In contrast, Lin et al. found that the median time to first seizure was 15 hours (IQR 6–24),³³ while Okochi et al. found that only 50% of seizures were captured in the first 24 hours,³⁴ suggesting that this population may also benefit from 24 to 48 hours of cEEG to identify seizures.

Impact of Seizures on Outcome

Pediatric Intensive Care Unit

One study examining ES and outcome found that in 200 patients with acute encephalopathy monitored by continuous EEG, ES alone were not associated with in-hospital mortality or worsening of Pediatric Cerebral Performance Category (PCPC) score (the PCPC is a measure of overall neurologic function).² However, ESE had a clear association with in-hospital mortality and PCPC worsening.² A follow-up study in previously healthy children assessed long-term outcome and found that electrographic status epilepticus, but not electrographic seizures, was associated with an increased risk of subsequent epilepsy, poorer quality of life, and unfavorable global outcomes after adjustment for diagnosis, EEG background category, and age.⁴ Similarly, a prospective observational study assessing the relationship between electrographic seizure burden and short-term neurologic outcome found that patients with posthospitalization neurologic decline were found to have a mean seizure burden of 15.7% per hour compared to only 1.8% per hour for those without a decline in neurologic function.³ Seizure burden was calculated from the 1-hour epoch of EEG with maximum seizure duration, and neurologic outcome was based on PCPC scores.³ The probability and severity of neurologic decline rose sharply above a maximum seizure burden of 20% (12 minutes/hour), suggesting that seizure burden, or maximum percentage per hour of seizure activity, was independently associated with poor short-term neurologic outcome.³ While further studies are needed, this information has been used to advocate for early treatment of ES to attempt to minimize seizure burden. Of note, studies to date have not shown that early treatment of ES improves outcome in critically ill children.

Cardiac Intensive Care Unit

The impact of seizures on long-term neurodevelopmental outcome has been studied in children with congenital heart disease. Infants with transposition of the great arteries enrolled in the Boston Circulatory Arrest study who had clinical or subclinical seizures postoperatively were found to have lower motor function scores and an increased likelihood of having abnormalities on MRI at 1 year of age.³⁵ By 4 years of age, patients who had had seizures in the postoperative period had a lower mean IQ score and a higher risk of neurologic abnormalities.³⁶ By the time this cohort reached adolescence, at 16 years of age, seizures in the postoperative period were associated with lower scores on math and reading composites, executive function, general memory index, and visual spatial tasks.³⁷ A large prospective study at the Children’s Hospital of Philadelphia assessing patients with complex congenital heart disease, including hypoplastic left heart syndrome and variants, found that at 1 year of age patients with seizures were more likely to have an abnormal neuromuscular examination.³⁸ There was no association with seizures and worse developmental outcome; however, when assessing the location of seizures, those with frontal lobe seizures had lower scores on a psychomotor development index when compared to patients with seizures coming from other areas.³⁸ At 4 years of age, the cohort with postoperative seizures had deficits in executive function and social interactions/restrictive behaviors.³⁹

Electrographic Seizure Patterns in the Intensive Care Unit

The electrographic description of seizures in the intensive care unit varies among studies. One cohort found spike wave discharges alone to be the most common EEG signature (39%), followed by a combination of spike wave, sharp wave, and/or polyspike wave (31%).⁴⁰ The mean duration of seizures was 159 seconds with a range of 10 seconds to 11 minutes. Other ictal patterns include low-voltage fast activity leading to rhythmic sharp waves, accelerating spike-slow wave discharges, and sharp rhythmic theta with duration of 10 seconds to 2 minutes and maximum seizure duration of 45 minutes.¹⁴ Typical and atypical spike and wave discharges and rhythmic sharp delta have also been described, with frequencies from 0.5 to 15 Hz; the majority of seizures in this cohort were focal.¹³ While most studies document brief seizures, one cohort showed that 51% were greater than 5 minutes, 27% of which were greater than 30 minutes.¹⁶

Conclusion

Electrographic seizures are common in critically ill children. High seizure burden and status epilepticus have been associated with poor outcome, ranging from in-hospital mortality to worse neurodevelopmental outcome and epilepsy. While prompt identification and treatment of ES may improve outcome, this remains to be demonstrated.

Related posts:

Templates and Treatment Algorithms Cases Basics of EEG Monitoring in Pediatric Critical Care Questions and Answers by Chapter Encephalopathy, Coma Patterns, and Other Abnormalities of the EEG Background in Critically Ill Children Convulsive and Nonconvulsive Status Epilepticus in Critically Ill Children

Stay updated, free articles. Join our Telegram channel

Join