Study Design and Population

This prospective, single-center, longitudinal study was done at a government medical college of Northern India. The patients were enrolled after getting written informed consent from their parents. The study was conducted over 21 months from January 2017 to October 2018. A convenient sample size of 200 patients was taken.

Inclusion/Exclusion Criteria

The parents of all patients with CSE between 1 and 16 years of age presenting consecutively to the pediatric emergency or pediatric intensive care unit were asked to participate in the study. The known cases of cerebral palsy or degenerative brain disorder were excluded. Those who refused consent were also excluded.

Study Procedure and Data Collection

All new patients with CSE, fulfilling inclusion criteria during data collection, were enrolled.

SE definition was taken as per the International League Against Epilepsy 2015 guidelines and t ₂ (status lasting > 30 minutes) was taken as the time point, since long-term neuronal damage is supposed to occur after this point.13 Only patients with CSE were included. Refractory SE was defined as prolonged seizures that did not stop after administration of two antiseizure medications with different mechanisms of action or that need continuous administration of medication to abort seizures, regardless of seizure duration. The following definitions for CSE were considered: prolonged febrile seizures (PFS): SE in a previously neurologically normal child in the age group of 6 months to 5 years during a febrile (temperature > 38°C) illness and no evidence of definite intracranial infection2; acute symptomatic CSE: SE in a patient with no previous neurological abnormality, within 7 days of an identified acute neurologic insult, including pyogenic meningitis, viral encephalitis, metabolic abnormalities, drug-related effects, head trauma, hypoxic or anoxic event, or cerebrovascular disease2; remote symptomatic CSE: SE without any evidence of an identifiable acute insult but with a history of a preexisting central nervous system abnormality more than 7 days before an episode of CSE2; and Unknown: SE not classifiable into any other group.2 13

Clinico-demographic details were recorded on the prespecified case record form. The baseline evaluation at the time of enrolment was done by a physician/study investigator. A structured institutional treatment regime to treat CSE was followed. After taking care of airway, breathing, and circulation, blood sugar levels were checked. Patients with seizure > 5 minutes duration were given intravenous 0.2 mg/kg diazepam as bolus dose. If seizure persisted after diazepam, intravenous phenytoin 20 mg/kg was given in a slow intravenous infusion over 15 to 20 minutes; if there was no response, second loading with phenytoin 10 mg/kg was given 10 minutes after the first phenytoin loading. Patients in whom, status persisted, loading of 30 mg/kg valproate was given. If still unresponsive, intravenous levetiracetam 30 mg/kg was given. In patients with liver dysfunction, valproate was not given. If seizures persisted after this step, either loading of phenobarbitone 20 mg/kg was given or midazolam bolus 0.2 mg/kg followed by 2 μg/kg/min to maximum of 10 μg/kg/min. After seizures were controlled, the infusion of midazolam was continued for at least 24 hours and then gradually tapered off. The patients with Glasgow Coma Scale (GCS) score < 8 or breathing difficulty were intubated.

After stabilization, details of developmental history, any coexisting medical diseases, epilepsy, antiepileptic drug (AED) use, and seizure semiology were taken. A detailed physical examination was performed. The laboratory tests including hemogram, random blood sugar levels, serum calcium, serum electrolytes and renal function tests, cerebrospinal fluid analysis, electroencephalography, neuroimaging (computed tomographic scan, magnetic resonance imaging of brain), and screening for metabolic disorders were conducted as per clinician's discretion. During the hospital stay, the outcome, critical care needs, and total stay in hospital were noted. The patients were followed throughout the hospital stay.

Methods of neurological disability assessment: Glasgow Outcome Score (GOS) was used to assess outcome at three points: at discharge, at 1 month, and at 3 months.

At follow-up, patients were called to the hospital; those who refused to come to the hospital were assessed by telephonic interview.

Outcome Measures

The primary outcome measures were taken as proportion of patients with CSE having poor outcome (GOS I–IV) at discharge, at 1 month, and 3 months postdischarge, and correlation of GOS score to clinical profile was analyzed.

Statistical Analysis

Data analysis was done on SPSS version 18. Descriptive data for all study variables were done as mean ± standard deviation (SD), numbers and percentages, and median and interquartile ranges (IQRs). The categorical and continuous variables were analyzed by chi-square test and independent t-test. The mortality/morbidity in the subgroups was reported as the proportion of the sample. The potential influence of confounding variables for poor outcome was looked by a multivariate binomial logistic regression model and results were reported as odds ratios with 95% confidence intervals. A p-value of < 0.05 was taken as statistically significant.

Ethics

The study was approved by the institutional ethics committee (Ethical approval number: EC/2016/0054).

Clinical and Demographic Features

The mean (SD) age of the patients was 7 (2.9) years and 58% were males (Table 1). Children under 5 years constituted 41%. A very high number of children (both genders) were malnourished (p 0.97) (Table 2). Majority (87%) were immunized as per national immunization schedule. Among the presenting complaints, apart from seizures, fever (49.5%) and altered sensorium (16.5%) were main complaints. Sixty-five percent patients had generalized seizures and 39% had seizure duration more than 1 hour at presentation. Higher percentage (47%) of girls had SE lasting for > 1 hour (p 0.0001) (Table 2). Only 24% patients received nasal/parenteral anticonvulsant at home, from a private practitioner, or from a primary health care center.

At hospital admission, mean (SD) GCS score was 13.75 (2.58). Eleven percent patients had a GCS score < 8 at presentation (out of them 64% were girls), 12.5% had features of raised intracranial pressure, 7.5% respiratory failure, and 5% circulatory shock.

Twenty-two percent had normal neuroimaging; 18% patients had neurocysticercosis, 7.5% tuberculoma, and 6% had features of old insult (gliosis, porencephalic cyst), whereas 2.5% patients had temporal lobe involvement suggestive of herpes simplex encephalitis.

Management

During hospital stay, 15% required intubation, 17% needed respiratory support, 14% neurocritical care, and 13% inotropic support. More girls as compared with boys (26% vs. 12%) required critical care. Note that 24.5% patients had remote symptomatic etiology (RSE). Mean (IQR) duration of hospital stay was 6.81 (2–30) days (Table 1).

Etiology

It was assigned as per the clinical and investigational findings. Out of 200 patients, acute symptomatic etiology (63.5%) was the most common etiology followed by remote symptomatic in 9%, PFS in 16.5%, and unknown in 10% cases. Among acute symptomatic etiology, acute intracranial infections were the most common (95%) (Supplementary Table S1, available in the online version only). There was no significant difference between etiological distributions between genders (Table 2).

Outcome

The patients were assessed for outcome by GOS at three assessment points, at discharge, at 1 month, and at 3 months.

At discharge, out of 200 patients, 25 (12.5%) died during hospital stay. Rest of the patients was assessed at discharge for neurological sequelae. Eighty percent had good recovery, 3% mild disability, 3% moderate disability, and 1.5% had persistent vegetative state (Table 3).

Predictive Factors for Poor Outcome at Discharge (Table 4)

We grouped GOS 1 to 4 as poor outcome and GOS 5 as good outcome. Patients aged > 5 years had 1.5 times chances of poor outcome, also girls had more chances of poor outcome (p 0.0009). Patients with generalized seizures and seizure duration > 60 minutes had poor outcome (p 0.0008 and < 0.000). RSE was likely to cause 15 times more chances of poor outcome in patients as compared with patients without RSE. Patients presenting with GCS < 8 had 79 times more chances of having poor outcome than who had higher GCS. Requirement of critical care (assisted ventilation/inotropic support/management of raised intracranial pressure) reflecting seriousness of clinical condition was associated with very high chances of poor outcome (p < 0.0000). Patients with acute symptomatic etiology had higher chances of mortality and morbidity as compared with all other etiologies clubbed together.

Follow-up: Only 85% (149/175) patients could be followed up at 1 month and 71% (123/173) at 3 months (Table 3).

One-month follow-up: Two patients died before the first follow-up, cause was probably aspiration in one and seizures in another (Table 5). Ninety percent patients had a good recovery, 4% had moderate and severe disability each, and one patient had a persistent vegetative state (Table 3).

Three-month follow-up: Out of 123 contactable patients, one patient was reported dead, the cause of death was probably repeat seizure, 3.3% patients had severe disability, 9.7% moderate disability, and 86% good recovery. Seven patients (4 males) deteriorated from GOS 5 to 4 (as compared with follow-up at 1 month), three had acute symptomatic etiology of SE, two unknown etiology, and two patients PFS. One of the PFS patients developed mesial temporal sclerosis (MTS) on repeat neuroimaging at 1-month follow-up. None of these seven patients reported repeat seizures. Five of them were on anticonvulsants (Table 6).