Abstracts

Rationale: The value of detecting and treating seizures in the neuroscience intensive care unit (ICU) in terms of improving outcomes has not been clearly shown. We sought to determine whether effective seizure treatment was associated with survival to hospital discharge and ICU length of stay (LOS).Methods: The hospital course of all patients treated in the neuroscience ICU for an 18 month period from 5/2013-10/2014 was retrospectively reviewed for clinical features, use of long-term electroencephalography (EEG) monitoring for at least 24 hours, seizure occurrence and detailed documentation of seizure treatment. Statistics used were descriptive, Chi-square and linear regression for the association of hours to seizure cessation after initial detection (categorical: by 24, by 48 or >48 hours) with ICU LOS and survival. The effect of age, initial Glasgow Coma Scale (GCS) score, etiology for ICU admission and number of seizures within the first 24 hours were also assessed for their contribution to the outcomes.Results: Two hundred and sixty three patients underwent EEG monitoring and 25 had seizures (9.5%), 16 of whom had only subclinical seizures (64%). Age was from 18-73 years, with a median age of 71; ten were female. Eleven patients died. Survival was highly associated with time to seizure cessation (p=0.007). No patients died whose seizures stopped within 48 hours after initial seizure detection (Table 1). Survival was not associated with age (p=0.11), initial GCS score (p=0.12), etiology for ICU admission (approximately 50% subdural hematoma and 50% intracranial hemorrhage) (p=0.26) and number of seizures within the first 24 hours (p=0.82). ICU LOS was not associated with any clinical factors. Twenty-one patients were receiving antiepileptic drugs (AEDs) at the time of seizure detection including seven who were receiving sedative infusions. All but one patient had AEDs added or increased in response to seizure detection; one had a sedative added.Conclusions: These data indicate that seizure cessation within 48 hours of seizure detection in the neuroscience ICU is independently associated with improved survival. More aggressive AED treatment after seizure detection occurred in nearly all subjects. In our population, seizure cessation was more strongly associated with survival than other clinical factors.