Abstracts

Rationale: Video electroencephalogram (EEG) monitoring (VEM) is a valuable tool for the diagnosis and treatment of both epileptic seizures (ES) and nonepileptic seizures (NES). There is currently no consensus on the required duration of monitoring. The utility of prolonged VEM has not been robustly explored. We sought to determine the benefits of prolonged length of stay, specifically querying whether there was a point at which VEM became futile at yielding a diagnosis.Methods: We retrospectively reviewed the medical records of all patients admitted to our institution for VEM between 1/2004 and 12/2008. All patients underwent scalp EEG monitoring for classification/localization of ES and NES. We recorded the reason for admission, length of stay, and discharge diagnosis. A discharge diagnosis of nondiagnostic was assigned if patients had none of their habitual seizures during the admission. For patients having >1 admission during the study period, only the first admission was analyzed. We progressively analyzed lengths of stay until we discovered significant differences in the rates of nondiagnostic admissions for stays exceeding specific limits, ranging from >= 4 days to 14 days. We utilized chi-square analysis (Fisher s Exact Test, 2 sided), with p-values <0.05 considered statistically significant.Results: Five hundred ninety six patients were admitted for VEM. The majority (333, 55.9%) were admitted for a presurgical evaluation with presumed ES. The remaining patients were admitted for differential diagnosis of either presumed NES (150, 25.2%) or spells of other, unknown etiology (113, 19%). Only 89/596 admissions (14.9%) were nondiagnostic. Patients admitted for differential diagnosis of presumed NES were significantly more likely to have a nondiagnostic admission (31/150, 20.7%) compared to all others (58/446, 13%, p=0.033). Amongst all patients, there was no significant difference in the likelihood of having a nondiagnostic admission if monitoring was continued for any duration, including 5 or more days (62/428 patients, 14.5%) compared to less than 5 days (27/168 patients, 16.1%, p=0.61). This continued to be true if only patients admitted for a presurgical evaluation with presumed ES were analyzed (23/283, 8.1% versus 6/50, 12%, p=0.41). However, for patients admitted with presumed NES, a length of stay >