Abstracts

Rationale: Video EEG (VEEG) monitoring is expensive and sometimes a typical event is not recorded so a definitive diagnosis is not made after one prolonged admission. The primary purpose of this study is to determine the utility of repeat epilepsy monitoring unit (EMU) admissions for spell characterization after an initial ? 5 day nondiagnostic evaluation. Previous studies found that 48-72 hours of monitoring was usually sufficient for diagnosis in most patients with only a minority of cases requiring more than one week for establishing a diagnosis. Methods: On retrospective review of 4663 admissions between1998-2010, we identified 43 patients who were readmitted for characterization of their spells after an initial nondiagnostic ? 5 day evaluation. A patient s evaluation was considered nondiagnostic when a typical spell was not captured and there was no objective evidence of epilepsy based on video EEG monitoring and the clinical history. Patients with interictal spikes were included if these cases were still considered nondiagnostic when the history and interictal EEG did not correlate. Results: Of 43 patients included in this study, 18 (41%) had diagnostic repeat evaluations. Interictal spikes occurred in 6 of 43 during the initial evaluation. On repeat 2-8 day evaluations of these patients, 3(50%) were diagnosed with epilepsy and 3(50%) remained nondiagnostic. Out of 37 with no interictal discharges during the initial admission, 15(40.5%) had a diagnostic repeat evaluation. Of these, 9 were diagnosed with psychogenic nonepileptic seizures (PNES), 4 with epileptic seizures and 2 had interictal spikes and only PNES. In these 37 cases, mean length of monitoring for initial evaluation was 8.2 days (range 5-16), and for second evaluation it was 6 days (range 2-15). Three admissions were performed in 9 patients with length of monitoring ranging form 2-20 days. We found that 32/43 (74%) of the study cases included were female. Compared to the entire database population at our center, 640/902 (71%) of patients with nonepileptic seizures were female and in the patients with proven epilepsy 880/1615 (54%) were female. Conclusions: These data justify a repeat, and sometimes prolonged, VEEG evaluation in patients with an initial nondiagnostic EMU evaluation for 5 or more days. Repeat admissions allowed 41% of cases to have diagnostic findings to characterize seizures or spells. A high proportion of patients with an initial ? 5 day EMU evaluation are female. When a diagnosis is made on subsequent admissions in patients without spikes in the first admission, most have PNES.