Abstracts

Rationale: Generalized convulsive status epilepticus is a medical emergency and requires aggressive treatment. However, it is less clear if frequent or continuous nonconvulsive seizures carry a similarly poor prognosis. We sought to determine if certain characteristics of the EEG could be used to determine a seizure burden score which could be used to predict prognosis.Methods: In this retrospective study, we included patients who had undergone prolonged EEG monitoring at Duke University Hospital from October 2009 to May 2010 who had electrographic seizures within the first 4 hours of the recording. EEGs or EEG reports were reviewed to determine the frequency of seizures (seizures/hour), duration of individual seizures (seconds), and spatial extent of seizures ( focal, hemispheric, bilateral but not generalized, and generalized). The medical record was reviewed to obtain patient demographics, medical history (including etiology of seizures, comorbidities), treatments and outcome (death or hospice placement, skilled nursing facility placement, rehabilitation placement or home). The seizure burden was calculated as the product of seizure time per hour (average seizure duration times seizure frequency), average seizure duration, and spatial extent of the seizure. Patients were then categorized according to whether or not their seizure burden score was above or below the median value for the group. Outcomes were compared between the groups whose seizure burden score fell at or below the median and those above the median value. Results: Forty-one patients were included in the analysis. For the 21 patients whose seizure burden was at or below the median, 13 had a good outcome (10 to home and 3 to rehabilitation) and 8 had a poor outcome (4 to skilled nursing and 4 to hospice/death). For the 20 patients whose seizure burden score was above the median, 7 had a good outcome (5 to home and 2 to rehabilitation) and 13 had a poor outcome (2 to skilled nursing and 11 to hospice/death).Conclusions: In this retrospective analysis, a seizure burden score based on the first 4 hours of video EEG data suggests a potential for providing prognostic information in patients with frequent seizures on prolonged EEG. The seizure burden score may be a useful way of quantifying the severity of recurrent seizures on a prolonged EEG. In combination with clinical factors, it may eventually prove useful for determining prognosis and aiding in determining how aggressively to treat patients.