AUTHOR=Timpte Katharina , Rosenkötter Ulrike , Honrath Philipp , Weber Yvonne , Wolking Stefan , Heckelmann Jan 

TITLE=Assessing 72 h vs. 24 h of long-term video-EEG monitoring to confirm the diagnosis of epilepsy: a retrospective observational study

JOURNAL=Frontiers in Neurology

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2023.1281652

DOI=10.3389/fneur.2023.1281652

ISSN=1664-2295

ABSTRACT=Introduction: Paroxysmal seizure-like events can be a diagnostic challenge. Inpatient video-EEG monitoring (VEM) can be a valuable diagnostic tool, but recommendations for the minimal duration of VEM to confirm or rule out epilepsy are inconsistent. Here, we aim to determine whether VEM of 48 or 72 hours was superior to 24 hours. 
Methods: In this monocentric, retrospective study, we included 111 patients with paroxysmal, seizure-like events who underwent at least 72 hours of VEM. Inclusion criteria were (1) preliminary workup was inconclusive, (2) VEM admission occurred to confirm a diagnosis, and (3) at discharge, the diagnosis of epilepsy was conclusively established. We analyzed the VEM recordings to determine the exact time point of the first occurrence of epileptic abnormalities (EA; defined as interictal epileptiform discharges or electrographic seizures). Subgroup analyzes were performed for epilepsy types and treatment status.
Results: In our study population, 69.4 % (77/111) of patients displayed EAs during VEM. In this group, first occurrence of EAs was observed within 24 hours in 92.2% (71/77) of patients, and within 24 to 72 hours in 7.8 % (6/77). There was no statistically significant difference in the incidence of EA between medicated and non-medicated patients or between focal, generalized epilepsies and epilepsies of unknown type. Of 19 recorded spontaneous electroclinical seizures, six (31.6 %) occurred after 24 hours. 
Discussion: VEM of 24 hours may be sufficient in the diagnostic workup of paroxysmal seizure-like events under most circumstances. Considering the few cases of first EA in the timeframe between 24-72 hours, a prolonged VEM may be useful in cases with a high probability of epilepsy or where other strategies like sleep-EEG or ambulatory EEG show inconclusive results. Prolonged VEM increases the chance to record spontaneous seizures. Our study also highlights a high share of subjects with epilepsy that do not exhibit EAs during 72 hours of VEM.