Abstracts

Rationale: Non-convulsive status epilepticus (NCSE) has been associated with poor outcome in critically ill patients. The reasons for this are thought to be multifactorial including underlying etiology, delay in diagnosis and failure to escalate treatment rapidly amongst other factors. Achieving burst-suppression pattern on EEG is thought to reflect termination of status epilepticus and potentially facilitate better outcomes. We sought to determine if those cases who went into burst suppression during treatment of NSCE had better prognosis. Methods: Retrospective analysis of all adult patients undergoing more than 24 hours of Continuous EEG monitoring(CEEG) at our institution in 2006 for evaluation of coma. Using EEG reports and clinical data from our electronic medical record, each patient’s clinical presentation, course, and EEG findings were compared. Results: 39 patients were monitored during the one year period. Twenty-six patients (67%) were found to be in NCSE. There was an overall 58% mortality rate, which was slightly higher in those with hypoxia (73%) and in the elderly (73%). Seizure control was defined as 24 hours without electrographic seizure or, if CEEG discontinued, clinical parameters supportive of absence of NCSE. Burst-suppression was noted as an electrographic endpoint during treatment in 7 of 26 cases. Five of these seven patients died, of which death occurred in four (80%) of those after a decision to withdraw care for multiple comorbidities. One expired from herniation, in another seizures recurred a day later with fatality, and in a third the patient’s seizures were well-treated but died of a surgical complication. The mean time to onset of burst suppression in these 7 patients was 56 hours, compared to a mean time of 34 hours to achieve “seizure control” in the overall group. In those patients in whom there was no burst suppression, mortality was 10/19 (53%). The overall mean time to seizure control in all 26 patients with NCSE was 34 hours. In the 6 patients continuing with NCSE, this pattern continued until electrocerebral silence occurred. Conclusions: In our small sample size, achieving burst suppression does not result in reduced mortality. Several observations can be made from this retrospective analysis: 1) Burst suppression was achieved in only a fraction of cases of status epilepticus at our institution, with an average time of 2-3 days from seizure onset. 2) Because of multiple comorbidities and subsequent withdrawal of care in a majority of these critically ill patients, the efficacy of intervention with pharmacologic coma and burst suppression cannot be evaluated precisely. 3) As in other larger studies of similar patients, advanced age and hypoxia are associated with higher mortality within the NCSE population. Further study in NCSE to achieve rapid pharmacological induction of burst suppression to assess this EEG endpoint for morbidity and mortality outcomes is needed.