Abstract Details

Diagnosis of non-convulsive or subclinical seizures is greatly aided by electroencephalography (EEG), but access to conventional EEG is often delayed by hours. It has been suggested that when physicians suspect subclinical seizures, they empirically treat patients with anti-seizure medications (ASMs) before EEG arrives. A recent multi-center clinical trial suggests that the Ceribell Rapid Response EEG (Rapid-EEG) can acquire EEG within minutes and improve physicians’ diagnostic decisions. In this study, we assessed whether Rapid-EEG diagnoses would correlate with changes in ASM usage.

In this study, we tested the hypothesis that diagnoses from a Rapid Response EEG during the initial neurological evaluation of patients with suspected subclinical seizures correlate with changes in anti-seizure medication (ASM) usage.

We performed a retrospective chart-review of 100 consecutive patients at an academic medical center who underwent Rapid-EEG monitoring. Cases were excluded if clinical course information or EEG reports were not available in the electronic medical record. EEG interpretation by the on-call EEG fellow and attending was used as the gold standard. We categorized EEG diagnoses as seizures (SZ), highly epileptiform patterns (HEP), diffuse or focal slowing, or normal.

Nineteen patients were diagnosed with SZ (N=5) or HEP (N=14) based on Rapid-EEG, while in 81 patients, seizures and epileptiform discharges were ruled out. In total, 46% of patients had already been treated with ASMs prior to Rapid-EEG. Patients were less likely to be continued or started on ASMs if the Rapid-EEG diagnosis was normal or slow compared to SZ/HEP (51% vs. 84.2% ; χ² (1, N = 100) = 7.086, p = .0078).

Rapid-EEG helped identify patients with ongoing non-convulsive/subclinical seizures or highly epileptiform patterns. Patients in these high seizure-risk groups were more likely to be treated with ASMs after Rapid-EEG. By contrast, rapidly ruling out seizures correlated with a significant reduction in ASM treatment.