Authors :
Presenting Author: Caralynn Li, D.O. – Vanderbilt University
Ushtar Amin, M.D. – University of South Florida; Alfred Frontera, M.D. – University of South Florida; Selim Benbadis, M.D. – University of South Florida
Rationale:
With recent technological advancements, ambulatory video-EEG monitoring (AVEM) has become more readily available. Several studies have assessed the diagnostic yield of AVEM in capturing a habitual event in the adult population, and results generally range from 44% to 85%. Given this wide range, this calls for a better method to stratify the pretest probability of capturing an event so that resources can be properly allocated, and a clinician can order the study fittingly. The purpose of this study was to assess the likelihood of capturing a typical event in question based on certain baseline patient or event characteristics.Methods:
We retrospectively reviewed the results of 300 consecutive ambulatory video-EEG studies between June of 2021 and August of 2022. Patients in this study were seen at our academic epilepsy center at the University of South Florida and the James A. Haley Veterans Administration in Tampa Florida. Patients were included in event analysis if the study was ordered for the purpose of capturing an event (and excluded for all other purposes).
Results:
One hundred and ninety-five studies were ordered for the purpose of capturing an event. After exclusion of infrequent events or events with unclear frequency, 149 studies were included in event analysis.
Sixty-eight patients (46%) had their typical event captured, of which 94% were on camera. The diagnosis was an epileptic seizure in 17 patients (25%), psychogenic non-epileptic seizure in 7 (10%) and other non-epileptic events in 44 (65%). Of the 81 patients who did not have their typical event captured, 63 patients had no events, and 18 had atypical events (Table 1).
Regarding baseline event frequency, for patients who on average had daily events, 84% had events captured, which corresponded to significantly increased odds (OR 17.90, 95% CI 7.55-42.44, p< 0.001). For those that had events < 1 per week to