The First-Time Seizure Emergency Department Electroencephalogram Study

Abstract

Seizures account for 1.2% of all emergency department (ED) visits, with 24% of those representing first-time seizures. Our primary goal is to determine whether obtaining an electroencephalogram (EEG) in the ED after a first-time seizure can identify individuals appropriate for initiation of anticonvulsant therapy on ED discharge. Our secondary goals are to determine the association of historical and clinical seizure features with epileptic EEGs and to determine the interobserver agreement for the EEG interpretation. We conducted a prospective study including patients older than 17 years with either a first-time seizure or previous seizures without a previous EEG, all of whom were candidates for discharge home from the ED without antiepileptic drug treatment. We based seizure diagnosis on provider impression. We excluded patients with laboratory studies or neuroimaging deemed to be the seizure cause. EEG technicians performed a 30-minute EEG in the ED, which was immediately remotely interpreted by an epileptologist, who made a recommendation on antiepileptic drug initiation. We categorized EEGs as normal, abnormal but not epileptic, or epileptic. In accordance with duplicate EEG interpretation by a second, blinded epileptologist, we calculated interrater agreement for EEG interpretation and antiepileptic drug initiation. As a secondary analysis, according to questionnaires completed by patients and seizure observers, we explored the association of aura, focal symptoms, provocation, or historical risk factors with epilepsy. We enrolled 73 patients, 71 of whom had an EEG performed. All EEGs were performed within 11 hours of seizure, with an average of 3.85 hours. Twenty-four percent of patients (95% confidence interval 15% to 36%) received a diagnosis of epilepsy, and all began receiving antiepileptic drug therapy from the ED. Our final study sample size afforded only an exploratory analysis about an association between aura, focal onset, provocation, or historical risk factors with an epilepsy diagnosis. Weighted κ agreement for EEG interpretation was 0.69 (95% confidence interval 0.55 to 0.82). Of the 34 patients who followed up with an epileptologist, 9 had received a diagnosis of epilepsy in the ED, and none had antiepileptic drug medication stopped at initial follow-up. ED EEG performance in adults with first-time seizures results in a substantial yield of an epilepsy diagnosis and immediate initiation of antiepileptic drug treatment. A larger study is required to determine whether historical and clinical seizure features are associated with an ED epilepsy diagnosis.