Neurologists' and obstetricians' perspectives and current practices in breastfeeding counseling in women with epilepsy.

OBJECTIVES: In this study, we aimed to characterize practice patterns of neurologists and obstetricians in breastfeeding (BF) counseling in women with epilepsy (WWE) and explore factors that may influence physician counseling behaviors. METHODS: We conducted a cross-sectional study of neurologists and obstetricians via an anonymous survey from September 2021 until November 2021. A survey was developed to explore the following areas in WWE: current physicians' BF counseling patterns, physician-specific factors affecting BF counseling, and patient-specific factors and their impact on BF counseling. Descriptive statistics were generated for each survey question. Responses from neurologists and obstetricians were compared. Odds ratios (OR) and confidence intervals (CI) were calculated to assess factors that influence BF counseling in WWE. RESULTS: A total of 185 physicians participated in the study and consisted of 91 (49.2 %) neurologists, 83 (44.8 %) obstetricians, and 11 (6 %) participants from other specialties. Ninety-four percent (94 %) of neurologists and 92 % of obstetricians indicated that they provide BF safety counseling to WWE primarily during preconception and occasionally during pregnancy. Fifty-six percent of obstetricians reported being very comfortable with BF counseling in WWE, compared to 68 % of neurologists. Both groups rated research and clinical practice guidelines as two factors that have major impact on BF counseling; however, less than half (45 %) of neurologists are very familiar with the current literature and only a quarter (24 %) of obstetricians are very familiar with current literature regarding safety of BF in WWE. Regarding barriers to BF counseling, relative to neurologists, obstetricians believe that delivery of conflicting opinions among medical specialists about BF safety is a barrier that may impede effective BF counseling in WWE [OR = 2.78 (95 % CI: 1.30,5.95), adjusted p value (P = 0.008)]. SIGNIFICANCE: Variable knowledge of current literature in BF in WWE and low comfort levels in BF counseling among various specialists, as well as perceived inadequate data and clinical practice guidelines, may contribute to suboptimal BF counseling and impact health outcomes in WWE and their children.

Individuals with chronic epilepsy have elevated P-wave heterogeneity comparable to patients with atrial fibrillation.

OBJECTIVE: Identification of epilepsy patients with elevated risk for atrial fibrillation (AF) is critical given the heightened morbidity and premature mortality associated with this arrhythmia. Epilepsy is a worldwide health problem affecting nearly 3.4 million people in the United States alone. The potential for increased risk for AF in patients with epilepsy is not well appreciated, despite recent evidence from a national survey of 1.4 million hospitalizations indicating that AF is the most common arrhythmia in people with epilepsy. METHODS: We analyzed inter-lead heterogeneity of P-wave morphology, a marker reflecting arrhythmogenic nonuniformities of activation/conduction in atrial tissue. The study groups consisted of 96 patients with epilepsy and 44 consecutive patients with AF in sinus rhythm before clinically indicated ablation. Individuals without cardiovascular or neurological conditions (n = 77) were also assessed. We calculated P-wave heterogeneity (PWH) by second central moment analysis of simultaneous beats from leads II, III, and aVR ("atrial dedicated leads") from standard 12-lead electrocardiography (ECG) recordings from admission day to the epilepsy monitoring unit (EMU). RESULTS: Female patients composed 62.5%, 59.6%, and 57.1% of the epilepsy, AF, and control subjects, respectively. The AF cohort was older (66 ± 1.1 years) than the epilepsy group (44 ± 1.8 years, p < .001). The level of PWH was greater in the epilepsy group than in the control group (67 ± 2.6 vs. 57 ± 2.5 µV, p = .046) and reached levels observed in AF patients (67 ± 2.6 vs. 68 ± 4.9 µV, p = .99). In multiple linear regression analysis, PWH levels in individuals with epilepsy were mainly correlated with the PR interval and could be related to sympathetic tone. Epilepsy remained associated with PWH after adjustments for cardiac risk factors, age, and sex. SIGNIFICANCE: Patients with chronic epilepsy have increased PWH comparable to levels observed in patients with AF, while being ~20 years younger, suggesting an acceleration in structural change and/or cardiac electrical instability. These observations are consistent with emerging evidence of an "epileptic heart" condition.

Breastfeeding recommendations for women taking anti-seizure medications.

The literature regarding breastfeeding and effects of anti-seizure medication (ASM) exposure on the breastfed infant has been evolving rapidly over the last decade as new studies advance our understanding of the extent of medication exposure via breastfeeding and the long-term developmental outcomes of breastfed infants. Currently, strong evidence supports the safety of breastfeeding for women with epilepsy (WWE) taking most prescribed ASMs. In this review, we present a comprehensive overview of the data regarding ASM exposure in breastfed infants and neurodevelopmental outcomes in breastfed infants of mothers taking various ASMs. In addition, we present current breastfeeding recommendations and the reported adverse effects of various ASMs to facilitate decision making in the clinical care of WWE.

Epileptic seizures and Epilepsy Monitoring Unit admission disclose latent cardiac electrical instability.

BACKGROUND: Sudden cardiac arrest results from cardiac electrical instability and is 3-fold more frequent in patients with chronic epilepsy than in the general population. We hypothesized that focal to bilateral tonic-clonic seizures (FTBTCS) would acutely impact T-wave alternans (TWA), a marker of cardiac electrical instability linked to an elevated risk for sudden cardiac death, more than focal seizures (FS) [focal aware seizures (FAS) and focal with impaired awareness seizures (FIAS)], due to their greater sympathetic stimulation of the heart. Since stress has been shown to cause significant TWA elevations in patients with heart disease, we also hypothesized that the early days of an inpatient admission to an epilepsy monitoring unit (EMU) would be associated with higher TWA levels compared to later hospital days in patients with chronic epilepsy, presumably due to stress. DESIGN/METHODS: We analyzed the acute effects of seizures [FAS, FIAS, FTBTCS, and nonepileptic seizures (NES)] and day of hospital stay on TWA in 18 patients admitted to the EMU using high-resolution wireless electrocardiographic (ECG) patch monitors. RESULTS: A total of 5 patients had FTBTCS, 7 patients had FS (2 FAS, 5 FIAS), and 3 patients had NES only during the index hospital stay. Four patients did not have any electroclinical seizures or NES. FTBTCS resulted in marked acute increases in ictal TWA from baseline (2 ± 0.3 µV) to ictal maximum (70 ± 6.1 µV, p < 0.0001), the latter exceeding the 60 µV cut point defined as severely abnormal. By comparison, while FAS and FIAS also provoked significant increases in TWA (from 2 ± 0.5 µV to 30 ± 3.3 µV, p < 0.0001), maximum ictal TWA levels did not reach the 47 µV cut point defined as abnormal. Heart rate increases during FTBTCS from baseline (62 ± 5.8 beats/min) to ictal maximum (134 ± 8.6 beats/min, an increase of 72 ± 7.2 beats/min, p < 0.02) were also greater (p = 0.014) than heart rate increases during FS (from 70 ± 5.2 beats/min to 118 ± 6.2 beats/min, an increase of 48 ± 2.6 beats/min, p < 0.03). In 3 patients with NES, TWA rose mildly during the patients' typical episodes (from 2 ± 0.6 µV to 14 ± 2.6 µV, p < 0.0004), well below the cut point of abnormality, while heart rate increases were observed (from 75 ± 1.3 to 112 ± 8.7 beats/min, an increase of 37 ± 8.9 beats/min, p = 0.03). Patients with EEG-confirmed electroclinical seizures recorded while in the EMU exhibited significantly elevated interictal TWA maxima (61 ± 3.4 µV) on EMU admission day which were similar in magnitude to ictal maxima seen during FTBTCS (70 ± 6.1 µV, p = 0.21). During subsequent days of hospitalization, daily interictal TWA maxima showed gradual habituation in patients with both FS and FTBTCS but not in patients with NES only. CONCLUSIONS: This is the first study to our knowledge demonstrating that FTBTCS acutely provoke highly significant increases in TWA to levels that have been associated with heightened risk for sudden cardiac death in other patient populations. We speculate that mortality temporally associated with FTBTCS may, in some cases, be due to sudden cardiac death rather than respiratory failure. In patients with EEG-confirmed epilepsy, hospital admission is associated with interictal TWA maxima that approach those seen during FTBTCS, presumably related to stress during the early phase of hospitalization compared to later in the hospitalization, indicating cardiac electrical instability and potential vulnerability to sudden cardiac death related to stress independent of temporal relationships to seizures. The elevated heart rates observed acutely with seizures and on hospital Day 1 are consistent with a hyperadrenergic state and the effect of elevated sympathetic output on a vulnerable cardiac substrate, a phenomenon termed "the Epileptic Heart."

T-wave heterogeneity crescendo in the surface EKG is superior to heart rate acceleration for seizure prediction.

We examined whether T-wave heterogeneity (TWH) on the surface electrocardiographic (EKG) could predict epileptic seizure onset. Patients with electroencephalography-confirmed generalized tonic-clonic seizures (GTCS) (n = 6) exhibited abnormal elevations in TWH (>80 µV) at baseline (105 ±â€¯20.4 µV), which increased from 30 min prior to seizure without heart rate increases > 2 beats/min until 10 min pre-seizure. Specifically, TWH on 3-lead surface EKG patch recordings increased from 1-hour baseline to 30 min (<0.05), 20 min (p < 0.002), 10 min (p = 0.01), and 1 min (p = 0.01) before seizure onset. At 10 min following GTCS, TWH returned to 110 ±â€¯20.3 µV, similar to baseline (p = 0.54). This pre-ictal TWH warning pattern was not present in patients with psychogenic nonepileptic seizures (PNES) (n = 3), as TWH did not increase until PNES and returned to baseline within 10 min after PNES. Acute elevations in TWH may predict impending GTCS and may discriminate patients with GTCS from those with behaviorally similar PNES.

Epileptic heart: A clinical syndromic approach.

Prevention of premature death in patients with chronic epilepsy remains a major challenge. Multiple pathophysiologic factors have been implicated, with intense investigation of cardiorespiratory mechanisms. Up to four in five patients with chronic epilepsy exhibit cardiovascular comorbidities. These findings led us to propose the concept of an "epileptic heart," defined as "a heart and coronary vasculature damaged by chronic epilepsy as a result of repeated surges in catecholamines and hypoxemia leading to electrical and mechanical dysfunction." Among the most prominent changes documented in the literature are high incidence of myocardial infarction and arrhythmia, altered autonomic tone, diastolic dysfunction, hyperlipidemia, and accelerated atherosclerosis. This suite of pathologic changes prompted us to propose for the first time in this review a syndromic approach for improved clinical detection of the epileptic heart condition. In this review, we discuss the key pathophysiologic mechanisms underlying the candidate criteria along with standard and novel techniques that permit evaluation of each of these factors. Specifically, we present evidence of the utility of standard 12-lead, ambulatory, and multiday patch-based electrocardiograms, along with measures of cardiac electrical instability, including T-wave alternans, heart rate variability to detect altered autonomic tone, echocardiography to detect diastolic dysfunction, and plasma biomarkers for assessing hyperlipidemia and accelerated atherosclerosis. Ultimately, the proposed clinical syndromic approach is intended to improve monitoring and evaluation of cardiac risk in patients with chronic epilepsy to foster improved therapeutic strategies to reduce premature cardiac death.

Factors affecting breastfeeding in women with epilepsy.

OBJECTIVE: This study aimed to compare rates of breastfeeding initiation and maintenance in women with epilepsy (WWE) with those of the general population and to identify factors affecting breastfeeding patterns in WWE. METHODS: We retrospectively reviewed data for the following variables in pregnant WWE (n = 102) and healthy women without epilepsy (n = 112): demographic characteristics (age, race, ethnicity), epilepsy type (focal or generalized onset), antiseizure medication(s), psychiatric comorbidities, postpartum seizure control, breastfeeding counseling, and lactation consultation. Fisher exact test and logistic regression analyses were performed to compare the rates of breastfeeding initiation and continuation in pregnant WWE with those of healthy pregnant women and to determine factors associated with rates of breastfeeding initiation and maintenance. RESULTS: The rate of breastfeeding initiation in WWE was significantly lower than in healthy women without epilepsy (50.9% vs. 87.6%), and WWE were less likely to maintain breastfeeding at 6 weeks (38.2%) and 3 months (36%) postpartum. Nearly half (53%) of WWE received antepartum breastfeeding counseling by their neurologists, and these women had higher odds of breastfeeding initiation and continuation (odds ratio [OR] = 2.53, 95% confidence interval [CI] = 1.14-5.72, p = .02). Postpartum consultation with a board-certified lactation consultant was associated with higher odds of breastfeeding continuation at 6 weeks (OR = 5.43, 95% CI = 1.39-27.23, p = .02) and at 3 months (OR = 4.9, 95% CI = 1.34-20.87, p = .019). Women taking levetiracetam were more likely to initiate and continue breastfeeding than those taking lamotrigine (OR = 6.22, 95% CI = 2.15-20.20, p = .001). SIGNIFICANCE: The initiation rate and duration of breastfeeding were significantly lower in WWE than in healthy women without epilepsy and were significantly associated with several factors. Identification of potential barriers to breastfeeding in WWE may lead to development of intervention strategies that can improve breastfeeding rates in WWE to maximize positive health outcomes for WWE and their infants.

Diagnosis and Management of Status Epilepticus.

Seizures are among the most common neurological presentations to the emergency room. They present on a spectrum of severity from isolated new-onset seizures to acute repetitive seizures and, in severe cases, status epilepticus. The latter is the most serious, as it is associated with high morbidity and mortality. Prompt recognition and treatment of both seizure activity and associated acute systemic complications are essential to improve the overall outcome of these patients. The purpose of this review is to provide the current viewpoint on the diagnostic evaluation and pharmacological management of patients presenting with status epilepticus, and the common associated systemic complications.

Seizures and status epilepticus may be risk factor for cardiac arrhythmia or cardiac arrest across multiple time frames.

OBJECTIVE: To determine if Emergency Department (ED) or inpatient encounters for epilepsy or status epilepticus are associated with increased odds of cardiac arrhythmia or cardiac arrest over successively longer time frames. METHODS: The State Inpatient and ED Databases (from New York, Florida, and California) are statewide datasets containing data on 97% of hospitalizations and ED encounters from these states. In this retrospective, case-crossover study, we used International Classification of Diseases, Ninth Revision, Clinical Modification codes to identify index cardiac arrhythmia encounters. Exposures were inpatient or ED encounters for epilepsy or status epilepticus. The case-crossover analysis tested whether an epilepsy or status epilepticus encounter within various case periods (1, 3, 7, 30, 60, 90, and 180 days prior to index encounter) was associated with subsequent ED or inpatient encounter for cardiac arrhythmia, as compared to control periods of equal length one year prior. RESULTS: The odds ratio (OR) for cardiac arrhythmia after an epilepsy encounter was significant at all time intervals (OR range 2.37-3.36), and highest at 1 day after epilepsy encounter (OR 3.63, 95% confidence interval [CI] 1.66-7.93, p = 0.0013). The OR after status epilepticus was significant at 7- to 180-day intervals (OR range 2.25-2.74), and highest at 60 days (OR 2.74, CI 2.09-3.61, p < 0.0001). SIGNIFICANCE: Epilepsy and status epilepticus events are associated with increased odds of subsequent cardiac arrhythmia or cardiac arrest over multiple chronic timeframes. Increased cardiac surveillance may be warranted to minimize morbidity and mortality in patients with epilepsy.

The Epileptic Heart: Concept and clinical evidence.

Sudden unexpected death in epilepsy (SUDEP) is generally considered to result from a seizure, typically convulsive and usually but not always occurring during sleep, followed by a sequence of events in the postictal period starting with respiratory distress and progressing to eventual cardiac asystole and death. Yet, recent community-based studies indicate a 3-fold greater incidence of sudden cardiac death in patients with chronic epilepsy than in the general population, and that in 66% of cases, the cardiac arrest occurred during routine daily activity and without a temporal relationship with a typical seizure. To distinguish a primarily cardiac cause of death in patients with epilepsy from the above description of SUDEP, we propose the concept of the "Epileptic Heart" as "a heart and coronary vasculature damaged by chronic epilepsy as a result of repeated surges in catecholamines and hypoxemia leading to electrical and mechanical dysfunction." This review starts with an overview of the pathophysiological and other lines of evidence supporting the biological plausibility of the Epileptic Heart, followed by a description of tools that have been used to generate new electrocardiogram (EKG)-derived data in patients with epilepsy that strongly support the Epileptic Heart concept and its propensity to cause sudden cardiac death in patients with epilepsy independent of an immediately preceding seizure.

Depot medroxyprogesterone acetate concentrations in patients with and without the use of antiseizure medications.

OBJECTIVES: To measure plasma concentrations of medroxyprogesterone acetate (MPA) in users with epilepsy treated with antiseizure medications and compare these to MPA concentrations in those without epilepsy. STUDY DESIGN: For this multisite cross-sectional study, we obtained a single blood sample from those with epilepsy treated with various antiseizure medications (n = 18) within the week before their next depot medroxyprogesterone injection. Among the participants without epilepsy (n = 20), 10 similarly were scheduled within the week prior to the next injection, and 10 were scheduled at earlier intervals to attempt to balance the time intervals between groups. MPA concentrations were determined by a validated assay. RESULTS: MPA concentrations were similar among those with epilepsy and controls and between groups with and without the use of enzyme-inducing medications. The lowest MPA concentrations, under 0.07 ng/mL, were observed among two of eight using enzyme-inducing antiseizure medications, one of 10 using noninducing medications, and one of 19 controls had concentrations below 0.2 ng/mL. CONCLUSIONS: In this exploratory study, lower MPA concentrations in some participants using enzyme-inducing antiseizure medications suggest a potential interaction that could reduce depot medroxyprogesterone efficacy.

Neurophysiology State Dynamics Underlying Acute Neurologic Recovery After Cardiac Arrest.

BACKGROUND AND OBJECTIVES: Epileptiform activity and burst suppression are neurophysiology signatures reflective of severe brain injury after cardiac arrest. We aimed to delineate the evolution of coma neurophysiology feature ensembles associated with recovery from coma after cardiac arrest. METHODS: Adults in acute coma after cardiac arrest were included in a retrospective database involving 7 hospitals. The combination of 3 quantitative EEG features (burst suppression ratio [BSup], spike frequency [SpF], and Shannon entropy [En]) was used to define 5 distinct neurophysiology states: epileptiform high entropy (EHE: SpF ≥4 per minute and En ≥5); epileptiform low entropy (ELE: SpF ≥4 per minute and <5 En); nonepileptiform high entropy (NEHE: SpF <4 per minute and ≥5 En); nonepileptiform low entropy (NELE: SpF <4 per minute and <5 En), and burst suppression (BSup ≥50% and SpF <4 per minute). State transitions were measured at consecutive 6-hour blocks between 6 and 84 hours after return of spontaneous circulation. Good neurologic outcome was defined as best cerebral performance category 1-2 at 3-6 months. RESULTS: One thousand thirty-eight individuals were included (50,224 hours of EEG), and 373 (36%) had good outcome. Individuals with EHE state had a 29% rate of good outcome, while those with ELE had 11%. Transitions out of an EHE or BSup state to an NEHE state were associated with good outcome (45% and 20%, respectively). No individuals with ELE state lasting >15 hours had good recovery. DISCUSSION: Transition to high entropy states is associated with an increased likelihood of good outcome despite preceding epileptiform or burst suppression states. High entropy may reflect mechanisms of resilience to hypoxic-ischemic brain injury.

Treatment of nonconvulsive status epilepticus.

OPINION STATEMENT: Status epilepticus is a neurologic emergency with significant potential morbidity and even mortality. It should be diagnosed and treated as quickly as possible, with concurrent rapid identification of the underlying cause. Nonconvulsive status epilepticus can be difficult to diagnose due to its protean manifestations, so a high index of suspicion is vital to prompt diagnosis and effective management. There are several different types and syndromes of nonconvulsive status epilepticus, and appropriate treatment varies according to those types. Some are relatively benign. Others, particularly the nonconvulsive status encountered in sick, hospitalized patients, can be life-threatening. The latter should be treated much more rapidly and aggressively than the former, sometimes even requiring coma-inducing anesthetic agents.

The Epileptic Heart and the Case for Routine Use of the Electrocardiogram in Patients with Chronic Epilepsy.

A dire complication associated with chronic epilepsy is abrupt premature death, currently referred to as sudden unexpected death in epilepsy (SUDEP). Although the traditional view has been that SUDEP is due primarily to peri-ictal respiratory failure leading to cardiac asystole, mounting evidence implicates accelerated heart disease, leading to an "epileptic heart" condition, especially after age 40, as another potential cause of abrupt premature death, although cardiac death is specifically excluded by the standard definition of SUDEP. Sudden cardiac death in epilepsy carries a 2.8-fold greater risk than in the general population and is 4.5 times more frequent than SUDEP. This review will discuss the rationale for routine use of electrocardiograms to assess cardiac risk in patients with epilepsy and the impact of epilepsy treatments, namely antiseizure medications and chronic vagus nerve stimulation.

Predicting Neurological Outcome From Electroencephalogram Dynamics in Comatose Patients After Cardiac Arrest With Deep Learning.

OBJECTIVE: Most cardiac arrest patients who are successfully resuscitated are initially comatose due to hypoxic-ischemic brain injury. Quantitative electroencephalography (EEG) provides valuable prognostic information. However, prior approaches largely rely on snapshots of the EEG, without taking advantage of temporal information. METHODS: We present a recurrent deep neural network with the goal of capturing temporal dynamics from longitudinal EEG data to predict long-term neurological outcomes. We utilized a large international dataset of continuous EEG recordings from 1,038 cardiac arrest patients from seven hospitals in Europe and the US. Poor outcome was defined as a Cerebral Performance Category (CPC) score of 3-5, and good outcome as CPC score 0-2 at 3 to 6-months after cardiac arrest. Model performance is evaluated using 5-fold cross validation. RESULTS: The proposed approach provides predictions which improve over time, beginning from an area under the receiver operating characteristic curve (AUC-ROC) of 0.78 (95% CI: 0.72-0.81) at 12 hours, and reaching 0.88 (95% CI: 0.85-0.91) by 66 h after cardiac arrest. At 66 h, (sensitivity, specificity) points of interest on the ROC curve for predicting poor outcomes were (32,99)%, (55,95)%, and (62,90)%, (99,23)%, (95,47)%, and (90,62)%; whereas for predicting good outcome, the corresponding operating points were (17,99)%, (47,95)%, (62,90)%, (99,19)%, (95,48)%, (70,90)%. Moreover, the model provides predicted probabilities that closely match the observed frequencies of good and poor outcomes (calibration error 0.04). CONCLUSIONS AND SIGNIFICANCE: These findings suggest that accounting for EEG trend information can substantially improve prediction of neurologic outcomes for patients with coma following cardiac arrest.

How much time is enough? Establishing an optimal duration of recording for ambulatory video EEG.

OBJECTIVE: Ambulatory video EEG allows for extended recording of EEG in the comfort of a patient's home. However, the optimal duration of recording to capture clinical events is yet to be established. The current study uses retrospective analyses to identify an optimal recording duration for at-home video EEG. METHODS: A retrospective review was performed utilizing an anonymized database of ambulatory video EEG recordings performed between March and September 2020 with a national in-home EEG provider. Only completed assessments with neurologists' reads of raw data were reviewed, resulting in 3644 unique studies divided into three age cohorts: pediatrics (n = 941), adult (n = 2020), and geriatric (n = 683). Cohorts were characterized by assessment yield and time to first typical clinical event, as well as subsequent typical events over duration of recording. RESULTS: Frequency distributions reveal over half of first events are captured within 12 hours, but longer recording durations capture a much wider majority of both first typical events, as well as the mean number of subsequent events (5 clinical events). In 72 hours, over 97% of first events were observed in adult and geriatric patients, as well as over 95% of the mean number of subsequent events. In children, time to first event was significantly earlier than either adult or geriatric samples, with 98% of first events, and 92.8% of the mean number of subsequent events being observed in 48 hours. SIGNIFICANCE: These results from a large-scale, national dataset of patients using in-home EEG monitoring suggests recording at least 48 hours in duration for children, and at least 72 hours in duration for adult and geriatric samples, is optimal to maximize the likelihood of observing typical clinical events to facilitate diagnosis.

Predicting neurological outcome in comatose patients after cardiac arrest with multiscale deep neural networks.

OBJECTIVE: Electroencephalography (EEG) is an important tool for neurological outcome prediction after cardiac arrest. However, the complexity of continuous EEG data limits timely and accurate interpretation by clinicians. We develop a deep neural network (DNN) model to leverage complex EEG trends for early and accurate assessment of cardiac arrest coma recovery likelihood. METHODS: We developed a multiscale DNN combining convolutional neural networks (CNN) and recurrent neural networks (long short-term memory [LSTM]) using EEG and demographic information (age, gender, shockable rhythm) from a multicenter cohort of 1,038 cardiac arrest patients. The CNN learns EEG feature representations while the multiscale LSTM captures short-term and long-term EEG dynamics on multiple time scales. Poor outcome is defined as a Cerebral Performance Category (CPC) score of 3-5 and good outcome as CPC score 1-2 at 3-6 months after cardiac arrest. Performance is evaluated using area under the receiver operating characteristic curve (AUC) and calibration error. RESULTS: Model performance increased with EEG duration, with AUC increasing from 0.83 (95% Confidence Interval [CI] 0.79-0.87 at 12h to 0.91 (95%CI 0.88-0.93) at 66h. Sensitivity of good and poor outcome prediction was 77% and 75% at a specificity of 90%, respectively. Sensitivity of poor outcome was 50% at a specificity of 99%. Predicted probability was well matched to the observation frequency of poor outcomes, with a calibration error of 0.11 [0.09-0.14]. CONCLUSIONS: These results demonstrate that incorporating EEG evolution over time improves the accuracy of neurologic outcome prediction for patients with coma after cardiac arrest.