Abstracts

Rationale: Long term monitoring for epilepsy (LTME) is a valuable tool for diagnosis of seizure disorders and localization of epileptic lesions. Admission for LTME requires extensive use of medical resources, and the diagnostic utility of monitoring is not frequently assessed. We sought to evaluate at a small academic medical center the diagnostic efficacy of LTME, the duration of admission for epileptic seizure (ES) vs. non-epileptic seizure (NES), and the characteristics of patients with ES vs. NES diagnoses. Methods: After IRB approval, medical records for all patients admitted to the University of Toledo Medical Center (formerly Medical University of Ohio, Medical College of Ohio) Epilepsy Monitoring Unit between 9/1/06 and 12/31/07 were retrospectively reviewed. Baseline EEG results were coded as normal, slowing, epileptiform activity, or seizure/status epilepticus. LTME results were coded as ES, NES, both, non-diagnostic or other. Results: 154 LTME admissions were reviewed, representing 144 unique patients (9 had > 1 LTME). Of 144 patients, 84 (58.3%) were female; age at time of LTME was 41.8 ± 19 y (mean ± SD). The pre-LTME clinical diagnosis was ES in 67 (43.5%) and uncertain in 87 (56.5%). In persons with a pre-LTME diagnosis of ES, the median duration of prior seizures was 12 y, range 0.1 to 57 y (data unavailable for 39 (25%)). Duration of LTME was longer in patients with prior diagnosis of seizures (median 3 d, range 1 to 7) than with uncertain diagnosis (median = 2 d, range 1 to 7, p=0.015). Baseline EEG was normal in 63 (41.2%), focal or generalized slowing in 78 (51%), focal or generalized spike-and-wave in 34 (22%), electrographic seizure in 7 (4.5%), and other in 1 (0.7%). Diagnostic events (ES/NES) were recorded in 96 (62.3%) of 154 LTMEs. The post-LTME diagnosis was ES in 41 (26.6%), NES in 55 (35.7%), both ES and NES in 1 (0.7%), inconclusive in 50 (32.5%) and other physiologic condition in 8 (5.2%). Of 67 with prior diagnosis of ES, the post-LTME diagnosis was ES in 33, NES in 15, uncertain in 19. In the 87 with uncertain pre-LTME diagnosis, the post-LTME diagnosis was ES in 8, NES in 40, other physiological in 8 and uncertain in 31. For the 55 with post-LTME diagnosis of NES, 34 (61.8%) required 1 or 2 d of LTME and 48 (87%) were diagnosed within 3 d. LTME length for the 41 with post-LTME diagnoses of ES was ≤ 3 d in 25 (61%) and ≤ 4 d in 35 (85%). Baseline EEG was normal in 34 (61.8%) of those with post-LTME diagnosis of NES, but only 4 (9.8%) of patients with ES. Patients diagnosed with NES were the same age but more likely to be female (66:34) compared to patients with ES (51:49), and had shorter median history of seizures (2 vs. 18 y). Conclusions: LTME duration was longer for patients with ES than NES. Factors suggestive of a NES diagnosis included shorter seizure history, normal baseline EEG, and uncertain pre-LTME diagnosis. Female gender was associated with NES. Surprisingly, 22% of patients with pre-LTME diagnosis of ES had a post-LTME diagnosis of NES.