Abstracts

Non-convulsive status epilepticus (NCSE) is believed to be common in comatose patients and is presumed to increase morbidity and mortality. While the electroencephalogram (EEG) remains the only method of diagnosis, considerable disagreement exists regarding the definition of EEG criteria for NCSE. In addition, substantial resources are required to maintain 24-hour EEG coverage. The objective of this project was to determine whether EEG findings influenced acute management or were predictive of outcome in comatose patients. EEGs and clinical charts were reviewed in 86 consecutive patients for whom EEGs were ordered during admission to a tertiary-care ICU. Patient outcomes were correlated with EEG findings, historical factors, and acute therapies. Of the 86 subjects: 9 (10.5%) had Periodic Generalized Epileptiform Discharges (PGEDs) or triphasic waves, 25 (29.1%) had interictal epileptic discharges, 1 (0.1%) had Periodic Lateralized Epileptiform Discharges (PLEDs), 7 (8.1%) had burst suppression pattern, and the remaining 44 (51.2%) had background disturbance without paroxysmal activity. Three (3.5%) of 86 had seizures recorded during the EEG. All 3 of these patients had clinical convulsions during their admission to hospital; 2 presented initially with clinical status epilepticus. Two of the 3 patients had both clinical and electrographic seizures during the EEG. None of the patients were found to have continuous seizure activity (NCSE). Thirty-three (38.4%) patients died, 35 (40.7%) were transferred to subacute hospitals or nursing homes and 18 (20.9%) were discharged home. Medical management was modified in 14 (16.3%) patients based on EEG findings. EEG findings and use of antiepileptic drugs (AEDs) did not correlate with patient outcome. Worse outcome was correlated with history of cardiac arrest and the presence of motor activity other than clonic seizures. Better outcome was correlated with convulsive status epilepticus and clinical evidence of convulsive seizures. Our finding of improved outcome in comatose patients with clinical evidence of seizures as compared to those without was unexpected based on the assumption that seizures are detrimental. These observations suggest that while patients presenting with convulsive seizures are likely to respond to AEDs, EEG changes and motor activity resulting from severe brain injury such as cardiac arrest may be manifestations of different underlying pathophysiology, which are unlikely to respond to AEDs.