Interobserver reproducibility of electroencephalogram interpretation in critically ill children

Interictal High Frequency Oscillations Detected with Simultaneous Magnetoencephalography and Electroencephalography as Biomarker of Pediatric Epilepsy

Crucial to the success of epilepsy surgery is the availability of a robust biomarker that identifies the Epileptogenic Zone (EZ). High Frequency Oscillations (HFOs) have emerged as potential presurgical biomarkers for the identification of the EZ in addition to Interictal Epileptiform Discharges (IEDs) and ictal activity. Although they are promising to localize the EZ, they are not yet suited for the diagnosis or monitoring of epilepsy in clinical practice. Primary barriers remain: the lack of a formal and global definition for HFOs; the consequent heterogeneity of methodological approaches used for their study; and the practical difficulties to detect and localize them noninvasively from scalp recordings. Here, we present a methodology for the recording, detection, and localization of interictal HFOs from pediatric patients with refractory epilepsy. We report representative data of HFOs detected noninvasively from interictal scalp EEG and MEG from two children undergoing surgery.

The underlying generators of HFOs were localized by solving the inverse problem and their localization was compared to the Seizure Onset Zone (SOZ) as this was defined by the epileptologists. For both patients, Interictal Epileptogenic Discharges (IEDs) and HFOs were localized with source imaging at concordant locations. For one patient, intracranial EEG (iEEG) data were also available. For this patient, we found that the HFOs localization was concordant between noninvasive and invasive methods. The comparison of iEEG with the results from scalp recordings served to validate these findings. To our best knowledge, this is the first study that presents the source localization of scalp HFOs from simultaneous EEG and MEG recordings comparing the results with invasive recordings. These findings suggest that HFOs can be reliably detected and localized noninvasively with scalp EEG and MEG. We conclude that the noninvasive localization of interictal HFOs could significantly improve the presurgical evaluation for pediatric patients with epilepsy.

Measuring the Severity of Neonatal Seizures: Temporal-Spatial Burden

PURPOSE

The American Clinical Neurophysiology Society recommends measuring neonatal seizures' severity by their frequency (number of seizures-anywhere per hour), burden (percentage of time with seizures-anywhere), or on a region-by-region, temporal-spatial basis. This study compares two reduced-channel montages for temporal-spatial seizure burden analyses and examines the agreement of seizures' quantification among these three methodologies.

METHODS

A convenience sample of 10 neonatal electroencephalograms was annotated for the beginnings and ends of seizures, which appeared anywhere in the full neonatal montage, then repeated on a more precise, region-by-region basis using 2 reduced-channel montages A and B. Seizure severity was measured by seizures-anywhere frequency, seizures-anywhere burden, and temporal-spatial seizure burdens using montages A and B. The results were compared by measuring their correlation and by linear regression modeling.

RESULTS

Seizures-anywhere frequency was correlated with seizures-anywhere burden (ρ = 0.77). However, a narrow range of seizures-anywhere frequencies corresponded with a broad range of seizures-anywhere burdens. Although there was high correlation between seizures-anywhere burdens and temporal-spatial seizure burdens (ρ = 0.92 montage A, ρ = 0.90 montage B), seizures-anywhere burdens were insensitive to variations in the spatial aspects of seizures, which were highly prevalent even in this small sample set. After adjusting for intrareader variability, the temporal-spatial seizure burdens measured by montages A and B were not significantly different (P = 0.56).

CONCLUSIONS

The severity of neonatal seizures is poorly represented by simple measures such as seizures-anywhere frequencies or burdens. The use of temporal-spatial seizure burden measurements is supported in work where great precision in quantifying neonatal seizures is required.

Early Electroencephalographic Findings Correlate With Neurologic Outcome in Children Following Cardiac Arrest

OBJECTIVES

To determine the clinical and electroencephalographic findings associated with prognosis in nonneonate children following cardiac arrest.

DESIGN

Retrospective observational study.

SETTING

PICU and cardiac ICU.

PATIENTS

Nonneonate children with a history of cardiac arrest more than 2 minutes.

INTERVENTIONS

Electroencephalographic monitoring within 72 hours of return of spontaneous circulation.

MEASUREMENTS AND MAIN RESULTS

Clinical and features, neurophysiologic data, and Pediatric Cerebral Performance Category scores were collected. Electroencephalographic traces were reviewed in a blinded manner, all seizures and electroencephalographic findings noted, and the electroencephalography was scored at 1 hour, 24 hours, and continuous electroencephalographic end. Discrete data regarding specific characteristics of the electroencephalographic background and seizures were studied. Univariate and multivariate analyses were performed to identify associations between clinical variables, electroencephalographic findings, and Pediatric Cerebral Performance Category score at hospital discharge. Multivariate analysis of 73 children revealed duration of cardiac arrest less than 20 minutes or continuous electroencephalographic background activity within 12 hours postreturn of spontaneous circulation were associated with good short term neurologic outcome. Change in electroencephalographic background score over time and electroencephalographic data collected after the initial hour were not associated with outcome.

CONCLUSIONS

Following pediatric cardiac arrest, an initially normal electroencephalography or generalized slowing of the electroencephalographic background was associated with good neurologic outcome at hospital discharge.

Early Electroencephalographic Background Features Predict Outcomes in Children Resuscitated From Cardiac Arrest

OBJECTIVES

To determine 1) whether early electroencephalographic background features were associated with survival and neurologic outcomes among children resuscitated from cardiac arrest and not treated with therapeutic hypothermia and 2) if addition of electroencephalographic background to commonly used clinical criteria is more predictive of outcome than clinical criteria alone.

DESIGN

Retrospective study.

SETTING

PICU and Cardiac ICUs of a tertiary children's hospital.

PATIENTS

Patients resuscitated from in-hospital or out-of-hospital cardiac arrest who underwent clinically indicated electroencephalographic monitoring and were not treated with therapeutic hypothermia.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

One-hundred twenty-eight patients underwent electroencephalographic monitoring within 1 day of return of spontaneous circulation. Background category was normal in four subjects (3%), slow-disorganized in 58 subjects (45%), discontinuous-burst suppression in 24 subjects (19%) and attenuated-flat in 42 subjects (33%). Forty-six subjects (36%) had a reactive electroencephalography. Twenty subjects (15%) had a seizure during electroencephalographic monitoring. Absence of reactivity (p < 0.001) and seizures (p = 0.04) were associated with worse electroencephalographic background category. After controlling for covariates, for each incrementally worse background score, the odds of death was 3.63 (95% CI, 2.18-6.0; p < 0.001) and the odds of unfavorable neurologic outcome was 4.38 (95% CI, 2.51-7.17; p = 0.001).

CONCLUSIONS

Worse electroencephalographic background early after resuscitation from both in-hospital and out-of-hospital cardiac arrest is associated with increased odds of death and unfavorable neurologic outcomes at hospital discharge. These electroencephalographic background patterns may be used in addition to clinical criteria to support prognostic decision making.

Medical management of epileptic seizures: challenges and solutions

Epilepsy is one of the most common neurologic illnesses. This condition afflicts 2.9 million adults and children in the US, leading to an economic impact amounting to $15.5 billion. Despite the significant burden epilepsy places on the population, it is not very well understood. As this understanding continues to evolve, it is important for clinicians to stay up to date with the latest advances to provide the best care for patients. In the last 20 years, the US Food and Drug Administration has approved 15 new antiepileptic drugs (AEDs), with many more currently in development. Other advances have been achieved in terms of diagnostic modalities like electroencephalography technology, treatment devices like vagal nerve and deep-brain stimulators, novel alternate routes of drug administration, and improvement in surgical techniques. Specific patient populations, such as the pregnant, elderly, those with HIV/AIDS, and those with psychiatric illness, present their own unique challenges, with AED side effects, drug interactions, and medical-psychiatric comorbidities adding to the conundrum. The purpose of this article is to review the latest literature guiding the management of acute epileptic seizures, focusing on the current challenges across different practice settings, and it discusses studies in various patient populations, including the pregnant, geriatric, those with HIV/AIDS, comatose, psychiatric, and "pseudoseizure" patients, and offers possible evidence-based solutions or the expert opinion of the authors. Also included is information on newer AEDs, routes of administration, and significant AED-related drug-interaction tables. This review has tried to address only some of these issues that any practitioner who deals with the acute management of seizures may encounter. The document also highlights the numerous avenues for new research that would help practitioners optimize epilepsy management.

Interobserver agreement for neonatal seizure detection using multichannel EEG

Objective

To determine the interobserver agreement (IOA) of neonatal seizure detection using the gold standard of conventional, multichannel EEG.

Methods

A cohort of full‐term neonates at risk of acute encephalopathy was included in this prospective study. The EEG recordings of these neonates were independently reviewed for seizures by three international experts. The IOA was estimated using statistical measures including Fleiss' kappa and percentage agreement assessed over seizure events (event basis) and seizure duration (temporal basis).

Results

A total of 4066 h of EEG recordings from 70 neonates were reviewed with an average of 2555 seizures detected. The IOA was high with temporal assessment resulting in a kappa of 0.827 (95% CI: 0.769–0.865; n = 70). The median agreement was 83.0% (interquartile range [IQR]: 76.6–89.5%; n = 33) for seizure and 99.7% (IQR: 98.9–99.8%; n = 70) for nonseizure EEG. Analysis of events showed a median agreement of 83.0% (IQR: 72.9–86.6%; n = 33) for seizures with 0.018 disagreements per hour (IQR: 0.000–0.090 per hour; n = 70). Observers were more likely to disagree when a seizure was less than 30 sec. Overall, 33 neonates were diagnosed with seizures and 28 neonates were not, by all three observers. Of the remaining nine neonates with contradictory EEG detections, seven presented with low total seizure burden.

Interpretation

The IOA is high among experts for the detection of neonatal seizures using conventional, multichannel EEG. Agreement is reduced when seizures are rare or have short duration. These findings support EEG‐based decision making in the neonatal intensive care unit, inform EEG interpretation guidelines, and provide benchmarks for seizure detection algorithms.

Outcomes in pediatric patients with nonconvulsive status epilepticus

Recognition of nonconvulsive status epilepticus (NCSE) is gaining increasing attention in the assessment and evaluation of critically ill pediatric patients. The underlying cause of NCSE is often the most important factor in determining outcome. However, there is a growing body of literature suggesting that electrical seizure burden in NCSE also contributes to unfavorable outcomes. Determination of impact of NCSE on outcome based on current evidence involves consideration of heterogeneous study settings, study populations, and process of care and outcome measures. In addition, the lack of data on neurocognitive function prior to episodes of NCSE as well as limited long-term neurocognitive assessment data confines precise conclusions about neurocognitive changes. This article is part of a Special Issue entitled "Status Epilepticus".

Quantitative EEG is an objective, sensitive, and reliable indicator of transient anesthetic effects during Wada tests

The intracarotid amobarbital or Wada procedure is a component of the presurgical evaluation for refractory epilepsy, during which monitoring the onset and offset of transient anesthetic effects is critical. In this study, the authors characterized changes of 8 quantitative measures during 26 Wada tests, which included alpha, beta, theta, and delta powers, alpha/delta power ratio, beta/delta power ratio, median amplitude-integrated EEG, and 90% spectral edge frequency (SEF90), and correlated them with contralateral hemiplegia. The authors found that on the side of injection, delta and theta powers, alpha/delta power ratio, beta/delta power ratio, and SEF90 peaked within 1 minute after injection of 70 to 150 mg amobarbital or 4 to 7 mg methohexital. When contralateral arm strength returned to 3/5, delta power and amplitude-integrated EEG decayed on average 24% and 19%, respectively, for amobarbital, similar to that of methohexital (27% and 18%). Because delta power resolution most closely mirrored that of the hemiplegia and amplitude-integrated EEG had the highest signal/noise ratio, these quantitative values appear to be the best measures for decay of anesthetic effects. Increase in alpha power persisted longest, and therefore may be the best measure of late residual anesthetic effects.

Development and validation of a seizure prediction model in critically ill children

PURPOSE

Electrographic seizures are common in encephalopathic critically ill children, but identification requires continuous EEG monitoring (CEEG). Development of a seizure prediction model would enable more efficient use of limited CEEG resources. We aimed to develop and validate a seizure prediction model for use among encephalopathic critically ill children.

METHOD

We developed a seizure prediction model using a retrospectively acquired multi-center database of children with acute encephalopathy without an epilepsy diagnosis, who underwent clinically indicated CEEG. We performed model validation using a separate prospectively acquired single center database. Predictor variables were chosen to be readily available to clinicians prior to the onset of CEEG and included: age, etiology category, clinical seizures prior to CEEG, initial EEG background category, and inter-ictal discharge category.

RESULTS

The model has fair to good discrimination ability and overall performance. At the optimal cut-off point in the validation dataset, the model has a sensitivity of 59% and a specificity of 81%. Varied cut-off points could be chosen to optimize sensitivity or specificity depending on available CEEG resources.

CONCLUSION

Despite inherent variability between centers, a model developed using multi-center CEEG data and few readily available variables could guide the use of limited CEEG resources when applied at a single center. Depending on CEEG resources, centers could choose lower cut-off points to maximize identification of all patients with seizures (but with more patients monitored) or higher cut-off points to reduce resource utilization by reducing monitoring of lower risk patients (but with failure to identify some patients with seizures).

The standardization debate: A conflation trap in critical care electroencephalography

PURPOSE

Persistent uncertainty over the clinical significance of various pathological continuous electroencephalography (cEEG) findings in the intensive care unit (ICU) has prompted efforts to standardize ICU cEEG terminology and an ensuing debate. We set out to understand the reasons for, and a satisfactory resolution to, this debate.

METHOD

We review the positions for and against standardization, and examine their deeper philosophical basis.

RESULTS

We find that the positions for and against standardization are not fundamentally irreconcilable. Rather, both positions stem from conflating the three cardinal steps in the classic approach to EEG, which we term "description", "interpretation", and "prescription". Using real-world examples we show how this conflation yields muddled clinical reasoning and unproductive debate among electroencephalographers that is translated into confusion among treating clinicians. We propose a middle way that judiciously uses both standardized terminology and clinical reasoning to disentangle these critical steps and apply them in proper sequence.

CONCLUSION

The systematic approach to ICU cEEG findings presented herein not only resolves the standardization debate but also clarifies clinical reasoning by helping electroencephalographers assign appropriate weights to cEEG findings in the face of uncertainty.

Electroencephalography (EEG) for neurological prognostication after cardiac arrest and targeted temperature management; rationale and study design

BACKGROUND

Electroencephalography (EEG) is widely used to assess neurological prognosis in patients who are comatose after cardiac arrest, but its value is limited by varying definitions of pathological patterns and by inter-rater variability. The American Clinical Neurophysiology Society (ACNS) has recently proposed a standardized EEG-terminology for critical care to address these limitations.

METHODS/DESIGN

In the TTM-trial, 399 post cardiac arrest patients who remained comatose after rewarming underwent a routine EEG. The presence of clinical seizures, use of sedatives and antiepileptic drugs during the EEG-registration were prospectively documented.

DISCUSSION

A well-defined terminology for interpreting post cardiac arrest EEGs is critical for the use of EEG as a prognostic tool.

TRIAL REGISTRATION

The TTM-trial is registered at ClinicalTrials.gov (NCT01020916).

EEG interpretation reliability and interpreter confidence: a large single-center study

The intrarater and interrater reliability (I&IR) of EEG interpretation has significant implications for the value of EEG as a diagnostic tool. We measured both the intrarater reliability and the interrater reliability of EEG interpretation based on the interpretation of complete EEGs into standard diagnostic categories and rater confidence in their interpretations and investigated sources of variance in EEG interpretations. During two distinct time intervals, six board-certified clinical neurophysiologists classified 300 EEGs into one or more of seven diagnostic categories and assigned a subjective confidence to their interpretations. Each EEG was read by three readers. Each reader interpreted 150 unique studies, and 50 studies were re-interpreted to generate intrarater data. A generalizability study assessed the contribution of subjects, readers, and the interaction between subjects and readers to interpretation variance. Five of the six readers had a median confidence of ≥99%, and the upper quartile of confidence values was 100% for all six readers. Intrarater Cohen's kappa (κc) ranged from 0.33 to 0.73 with an aggregated value of 0.59. Cohen's kappa ranged from 0.29 to 0.62 for the 15 reader pairs, with an aggregated Fleiss kappa of 0.44 for interrater agreement. Cohen's kappa was not significantly different across rater pairs (chi-square=17.3, df=14, p=0.24). Variance due to subjects (i.e., EEGs) was 65.3%, due to readers was 3.9%, and due to the interaction between readers and subjects was 30.8%. Experienced epileptologists have very high confidence in their EEG interpretations and low to moderate I&IR, a common paradox in clinical medicine. A necessary, but insufficient, condition to improve EEG interpretation accuracy is to increase intrarater and interrater reliability. This goal could be accomplished, for instance, with an automated online application integrated into a continuing medical education module that measures and reports EEG I&IR to individual users.

Best-case kappa scores calculated retrospectively from EEG report databases

PURPOSE

The most popular metric for interrater reliability in electroencephalography is the kappa (κ) score. κ calculation is laborious, requiring EEG readers to read the same EEG studies. We introduce a method to determine the best-case κ score (κBEST) for measuring interrater reliability between EEG readers, retrospectively.

METHODS

We incorporated 1 year of EEG reports read by four adult EEG readers at our institution. We used SQL queries to determine EEG findings for subsequent analysis. We generated logistic regression models for particular EEG findings, dependent on patient age, location acuity, and EEG reader. We derived a novel measure, the κBEST statistic, from the logistic regression coefficients.

RESULTS

Increasing patient age and location acuity were associated with decreased sleep and increased diffuse abnormalities. For certain findings, EEG readers exhibited the dominant influence, manifesting directly as lower between-reader κBEST scores for certain EEG findings. Within-reader κBEST control scores were higher than between-reader scores, suggesting internal consistency.

CONCLUSIONS

The κBEST metric can measure significant interrater reliability differences between any number of EEG readers and reports, retrospectively, and is generalizable to other domains (e.g., pathology or radiology reporting). We suggest using this metric as a guide or starting point for focused quality control efforts.

Automated analysis of background EEG and reactivity during therapeutic hypothermia in comatose patients after cardiac arrest

Visual analysis of electroencephalography (EEG) background and reactivity during therapeutic hypothermia provides important outcome information, but is time-consuming and not always consistent between reviewers. Automated EEG analysis may help quantify the brain damage. Forty-six comatose patients in therapeutic hypothermia, after cardiac arrest, were included in the study. EEG background was quantified with burst-suppression ratio (BSR) and approximate entropy, both used to monitor anesthesia. Reactivity was detected through change in the power spectrum of signal before and after stimulation. Automatic results obtained almost perfect agreement (discontinuity) to substantial agreement (background reactivity) with a visual score from EEG-certified neurologists. Burst-suppression ratio was more suited to distinguish continuous EEG background from burst-suppression than approximate entropy in this specific population. Automatic EEG background and reactivity measures were significantly related to good and poor outcome. We conclude that quantitative EEG measurements can provide promising information regarding current state of the patient and clinical outcome, but further work is needed before routine application in a clinical setting.

Density spectral array for seizure identification in critically ill children

PURPOSE

We evaluated the validity and interrater reliability of encephalographer interpretation of color density spectral array EEG for seizure identification was evaluated in critically ill children and explored predictors of accurate seizure identification.

METHODS

Conventional EEG tracings from 21 consecutive critically ill children were scored for electrographic seizures. Four 2-hour long segments from each subject were converted to 8-channel color density spectral array displays, yielding 84 images. Eight encephalographers received color density spectral array training and circled elements thought to represent seizures. Images were reviewed in random order (Group A) or with information regarding seizure presence in the initial 30 minutes and with subject images in order (Group B). Sensitivity, specificity, and interrater reliability were calculated. Factors associated with color density spectral array seizure identification were assessed.

RESULTS

Seizure prevalence was 43% on conventional EEG. Specificity was significantly higher for Group A than Group B (92.3% vs. 78.2%, P < 0.00). Sensitivity was not significantly different between Groups A and B (64.8% vs. 75%, P = 0.22). Interrater reliability was moderate in both groups. Ten percent of images were falsely classified as containing a seizure. Seizure duration ≥2 minutes predicted identification (P < 0.001).

CONCLUSIONS

Color density spectral array may be a useful screening tool for seizure identification by encephalographers, but it does not identify all seizures and false positives occur.

Subclinical early posttraumatic seizures detected by continuous EEG monitoring in a consecutive pediatric cohort

PURPOSE

Traumatic brain injury (TBI) is an important cause of morbidity and mortality in children, and early posttraumatic seizures (EPTS) are a contributing factor to ongoing acute damage. Continuous video-EEG monitoring (cEEG) was utilized to assess the burden of clinical and electrographic EPTS.

METHODS

Eighty-seven consecutive, unselected (mild - severe), acute TBI patients requiring pediatric intensive care unit (PICU) admission at two academic centers were monitored prospectively with cEEG per established clinical TBI protocols. Clinical and subclinical seizures and status epilepticus (SE, clinical and subclinical) were assessed for their relation to clinical risk factors and short-term outcome measures.

KEY FINDINGS

Of all patients, 42.5% (37/87) had seizures. Younger age (p = 0.002) and injury mechanism (abusive head trauma - AHT, p < 0.001) were significant risk factors. Subclinical seizures occurred in 16.1% (14/87), while 6.9% (6/87) had only subclinical seizures. Risk factors for subclinical seizures included younger age (p < 0.001), AHT (p < 0.001), and intraaxial bleed (p < 0.001). SE occurred in 18.4% (16/87) with risk factors including younger age (p < 0.001), AHT (p < 0.001), and intraaxial bleed (p = 0.002). Subclinical SE was detected in 13.8% (12/87) with significant risk factors including younger age (p < 0.001), AHT (p = 0.001), and intraaxial bleed (p = 0.004). Subclinical seizures were associated with lower discharge King's Outcome Scale for Childhood Head Injury (KOSCHI) score (p = 0.002). SE and subclinical SE were associated with increased hospital length of stay (p = 0.017 and p = 0.041, respectively) and lower hospital discharge KOSCHI (p = 0.007 and p = 0.040, respectively).

SIGNIFICANCE

cEEG monitoring significantly improves detection of seizures/SE and is the only way to detect subclinical seizures/SE. cEEG may be indicated after pediatric TBI, particularly in younger children, AHT cases, and those with intraaxial blood on computerized tomography (CT).

Nonconvulsive electrographic seizures are common in children with abusive head trauma*

OBJECTIVE

To determine the prevalence of nonconvulsive seizures in children with abusive head trauma.

DESIGN

Retrospective study of children with abusive head trauma undergoing clinically indicated continuous electroencephalographic monitoring.

SETTING

PICU of a tertiary care hospital.

SUBJECTS

Children less than or equal to 2 years old with evidence of abusive head trauma determined by neuroimaging, physical examination, and determination of abuse by the Child Protection Team.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Thirty-two children with abusive head trauma were identified with a median age of 4 months (interquartile range 3, 5.5 months). Twenty-one of 32 children (66%) underwent electroencephalographic monitoring. Those monitored were more likely to have a lower admission Glasgow Coma Scale (8 vs 15, p = 0.05) and be intubated (16 vs 2, p = 0.002). Electrographic seizures occurred in 12 of 21 children (57%) and constituted electrographic status epilepticus in 8 of 12 children (67%). Electrographic seizures were entirely nonconvulsive in 8 of 12 children (67%). Electroencephalographic background category (discontinuous and slow-disorganized) (p = 0.02) and neuroimaging evidence of ischemia were associated with the presence of electrographic seizures (p = 0.05). Subjects who had electrographic seizures were no more likely to have clinical seizures at admission (67% electrographic seizures vs 33% none, p = 0.6), parenchymal imaging abnormalities (61% electrographic seizures vs 39% none, p = 0.40), or extra-axial imaging abnormalities (56% electrographic seizures vs 44% none, p = 0.72). Four of 21 (19%) children died prior to discharge; none had electrographic seizures, but all had attenuated-featureless electroencephalographic backgrounds. Follow-up outcome data were available for 16 of 17 survivors at a median duration of 9.5 months following PICU admission, and the presence of electrographic seizures or electrographic status epilepticus was not associated with the Glasgow Outcome Scale score (p = 0.10).

CONCLUSIONS

Electrographic seizures and electrographic status epilepticus are common in children with abusive head trauma. Most seizures have no clinical correlate. Further study is needed to determine whether seizure identification and management improves outcome.

Electroencephalography monitoring in critically ill children: current practice and implications for future study design

PURPOSE

Survey data indicate that continuous electroencephalography (EEG) (CEEG) monitoring is used with increasing frequency to identify electrographic seizures in critically ill children, but studies of current CEEG practice have not been conducted. We aimed to describe the clinical utilization of CEEG in critically ill children at tertiary care hospitals with a particular focus on variables essential for designing feasible prospective multicenter studies evaluating the impact of electrographic seizures on outcome.

METHODS

Eleven North American centers retrospectively enrolled 550 consecutive critically ill children who underwent CEEG. We collected data regarding subject characteristics, CEEG indications, and CEEG findings.

KEY FINDINGS

CEEG indications were encephalopathy with possible seizures in 67% of subjects, event characterization in 38% of subjects, and management of refractory status epilepticus in 11% of subjects. CEEG was initiated outside routine work hours in 47% of subjects. CEEG duration was <12 h in 16%, 12-24 h in 34%, and >24 h in 48%. Substantial variability existed among sites in CEEG indications and neurologic diagnoses, yet within each acute neurologic diagnosis category a similar proportion of subjects at each site had electrographic seizures. Electrographic seizure characteristics including distribution and duration varied across sites and neurologic diagnoses.

SIGNIFICANCE

These data provide a systematic assessment of recent CEEG use in critically ill children and indicate variability in practice. The results suggest that multicenter studies are feasible if CEEG monitoring pathways can be standardized. However, the data also indicate that electrographic seizure variability must be considered when designing studies that address the impact of electrographic seizures on outcome.

Electroencephalographic monitoring in the pediatric intensive care unit

Continuous electroencephalographic (CEEG) monitoring is used with increasing frequency in critically ill children to provide insight into brain function and to identify electrographic seizures. CEEG monitoring use often impacts clinical management, most often by identifying electrographic seizures and status epilepticus. Most electrographic seizures have no clinical correlate, and thus would not be identified without CEEG monitoring. There are increasing data showing that electrographic seizures and electrographic status epilepticus are associated with worse outcome. Seizure identification efficiency may be improved by further development of quantitative electroencephalography trends. This review describes the clinical impact of CEEG data, the epidemiology of electrographic seizures and status epilepticus, the impact of electrographic seizures on outcome, the utility of quantitative electroencephalographic trends for seizure identification, and practical considerations regarding CEEG monitoring.

Automated quantification of spikes

Methods for rapid and objective quantification of interictal spikes in raw, unprocessed electroencephalogram (EEG) samples are scarce. We evaluated the accuracy of a tailored automated spike quantification algorithm. The automated quantification was compared with the quantification by two board-certified clinical neurophysiologists (gold-standard) in five steps: 1) accuracy in a single EEG channel (5 EEG samples), 2) accuracy in multiple EEG channels and across different stages of the sleep-wake cycles (75 EEG samples), 3) capacity to detect lateralization of spikes (6 EEG samples), 4) accuracy after application of a machine-learning mechanism (11 EEG samples), and 5) accuracy during wakefulness only (8 EEG samples). Our method was accurate during all stages of the sleep-wake cycle and improved after the application of the machine-learning mechanism. Spikes were correctly lateralized in all cases. Our automated method was accurate in quantifying and detecting the lateralization of interictal spikes in raw unprocessed EEG samples.

Automated identification of multiple seizure-related and interictal epileptiform event types in the EEG of mice

Visual scoring of murine EEG signals is time-consuming and subject to low inter-observer reproducibility. The Racine scale for behavioral seizure severity does not provide information about interictal or sub-clinical epileptiform activity. An automated algorithm for murine EEG analysis was developed using total signal variation and wavelet decomposition to identify spike, seizure, and other abnormal signal types in single-channel EEG collected from kainic acid-treated mice. The algorithm was validated on multi-channel EEG collected from γ-butyrolacetone-treated mice experiencing absence seizures. The algorithm identified epileptiform activity with high fidelity compared to visual scoring, correctly classifying spikes and seizures with 99% accuracy and 91% precision. The algorithm correctly identifed a spike-wave discharge focus in an absence-type seizure recorded by 36 cortical electrodes. The algorithm provides a reliable and automated method for quantification of multiple classes of epileptiform activity within the murine EEG and is tunable to a variety of event types and seizure categories.

Recording and analysis techniques for high-frequency oscillations

In recent years, new recording technologies have advanced such that, at high temporal and spatial resolutions, high-frequency oscillations (HFO) can be recorded in human partial epilepsy. However, because of the deluge of multichannel data generated by these experiments, achieving the full potential of parallel neuronal recordings depends on the development of new data mining techniques to extract meaningful information relating to time, frequency and space. Here, we aim to bridge this gap by focusing on up-to-date recording techniques for measurement of HFO and new analysis tools for their quantitative assessment. In particular, we emphasize how these methods can be applied, what property might be inferred from neuronal signals, and potentially productive future directions.

Behavioral measures and EEG monitoring using the Brain Symmetry Index during the Wada test in children

EEG monitoring is used routinely during the Wada test in children. We quantified EEG asymmetry using the Brain Symmetry Index (BSI) to reduce subjectivity of EEG interpretation. Clinical and procedural variables were obtained and EEG data were retrieved from 46 patients with a total of 89 injections. The BSI, the absolute value of the relative difference of the average spectral density of the right and left hemisphere, was calculated over time for all EEGs. Lateralized slowing was correctly identified in all procedures. Asymmetry was minimal at baseline (BSI 0.16) and increased with injection of amobarbital (BSI 0.49). Various patterns of the BSI were seen in distinct clinical and procedural scenarios. In this retrospective analysis, the BSI could not predict an unsuccessful Wada procedure. Our results suggest application of the BSI during the Wada test in children is feasible. Real-time calculation of the BSI during EEG monitoring in the angiography suite is warranted for further validation.

Automatic burst detection for the EEG of the preterm infant

To aid with prognosis and stratification of clinical treatment for preterm infants, a method for automated detection of bursts, interburst-intervals (IBIs) and continuous patterns in the electroencephalogram (EEG) is developed. Results are evaluated for preterm infants with normal neurological follow-up at 2 years. The detection algorithm (MATLAB®) for burst, IBI and continuous pattern is based on selection by amplitude, time span, number of channels and numbers of active electrodes. Annotations of two neurophysiologists were used to determine threshold values. The training set consisted of EEG recordings of four preterm infants with postmenstrual age (PMA, gestational age + postnatal age) of 29-34 weeks. Optimal threshold values were based on overall highest sensitivity. For evaluation, both observers verified detections in an independent dataset of four EEG recordings with comparable PMA. Algorithm performance was assessed by calculation of sensitivity and positive predictive value. The results of algorithm evaluation are as follows: sensitivity values of 90% ± 6%, 80% ± 9% and 97% ± 5% for burst, IBI and continuous patterns, respectively. Corresponding positive predictive values were 88% ± 8%, 96% ± 3% and 85% ± 15%, respectively. In conclusion, the algorithm showed high sensitivity and positive predictive values for bursts, IBIs and continuous patterns in preterm EEG. Computer-assisted analysis of EEG may allow objective and reproducible analysis for clinical treatment.

EEG monitoring during therapeutic hypothermia in neonates, children, and adults

Therapeutic hypothermia is being utilized as a neuroprotective strategy in neonates, children, and adults. The most common indications are hypoxic ischemic encephalopathy in neonates and post cardiac arrest in adults. Electroencephalographic monitoring use is increasing in critical care units, and is sometimes a component of therapeutic hypothermia clinical pathways. Monitoring may detect non-convulsive seizures or non-convulsive status epilepticus, and it may provide prognostic information. We review data regarding indications for therapeutic hypothermia and electroencephalographic monitoring in neonatal, pediatric, and adult critical care units, and discuss technical aspects related to such monitoring.

What is the present-day EEG evidence for a preictal state?

EEG-based seizure prediction has undergone phases of optimism when analyses based on limited EEG samples suggested high sensitivity and specificity for several algorithms extracting features from raw preictal EEG data. When using long-term recordings, a more realistic view emerged which suggests that statistically significant predictions might be possible from surface and intracranial EEG, but no algorithm has yet demonstrated performance allowing for clinical application. Here, progress in EEG recording techniques, EEG analysis, and requirements for proper statistical validation of results are reported and discussed as they pertain to clinical implementation.

Short-term outcome prediction by electroencephalographic features in children treated with therapeutic hypothermia after cardiac arrest

BACKGROUND

Electroencephalographic (EEG) features may provide objective data regarding prognosis in children resuscitated from cardiac arrest (CA), but therapeutic hypothermia (TH) may impact its predictive value. We aimed to determine whether specific EEG features were predictive of short-term outcome in children treated with TH after CA, both during hypothermia and after return to normothermia.

METHODS

Thirty-five children managed with a standard clinical TH algorithm after CA were prospectively enrolled. EEG recordings were scored in a standardized manner and categorized. EEG category 1 consisted of continuous and reactive tracings. EEG category 2 consisted of continuous but unreactive tracings. EEG category 3 included those with any degree of discontinuity, burst suppression, or lack of cerebral activity. The primary outcome was unfavorable short-term outcome defined as Pediatric Cerebral Performance Category score of 4-6 (severe disability, vegetative, death) at hospital discharge. Univariate analyses of the association between EEG category and outcome was performed using logistic regression.

RESULTS

For tracings obtained during hypothermia, patients with EEGs in categories 2 or 3 were far more likely to have poor outcome than those in category 1 (OR 10.7, P = 0.023 and OR 35, P = 0.004, respectively). Similarly, for tracings obtained during normothermia, patients with EEGs in categories 2 or 3 were far more likely to have poor outcomes than those in category 1 (OR 27, P = 0.006 and OR 18, P = 0.02, respectively).

CONCLUSIONS

A simple EEG classification scheme has predictive value for short-term outcome in children undergoing TH after CA.

Nonconvulsive seizures are common in critically ill children

BACKGROUND

Retrospective studies have reported the occurrence of nonconvulsive seizures in critically ill children. We aimed to prospectively determine the incidence and risk factors of nonconvulsive seizures in critically ill children using predetermined EEG monitoring indications and EEG interpretation terminology.

METHODS

Critically ill children (non-neonates) with acute encephalopathy underwent continuous EEG monitoring if they met institutional clinical practice criteria. Study enrollment and data collection were prospective. Logistic regression analysis was utilized to identify risk factors for seizure occurrence.

RESULTS

One hundred children were evaluated. Electrographic seizures occurred in 46 and electrographic status epilepticus occurred in 19. Seizures were exclusively nonconvulsive in 32. The only clinical risk factor for seizure occurrence was younger age (p=0.03). Of patients with seizures, only 52% had seizures detected in the first hour of monitoring, while 87% were detected within 24 hours.

CONCLUSIONS

Seizures were common in critically ill children with acute encephalopathy. Most were nonconvulsive. Clinical features had little predictive value for seizure occurrence. Further study is needed to confirm these data in independent high-risk populations, to clarify which children are at highest risk for seizures so limited monitoring resources can be allocated optimally, and to determine whether seizure detection and management improves outcome.