Seizure Duration and Timepoint Cutoffs for Statistically Defining a Prolonged Seizure: A Post Hoc Analysis of the SCORE Video-EEG Database
Bersabeh Sile1, Pirgit Meritam Larsen2, Anna Kuba3, Sami Elmoufti4, Jenna Roberts1, Eugen Trinka5, Cedric Laloyaux6, Sandor Beniczky2
1UCB, Slough, UK, 2Filadelfia, Danish Epilepsy Center, Dianalund, Denmark, 3UCB, Warsaw, Poland, 4UCB, Morrisville, NC, USA, 5Department of Neurology, Neurocritical Care, and Neurorehabilitation, Member of European Reference Network EpiCARE, Center for Cognitive Neuroscience, Christian Doppler University Hospital, Paracelsus Medical University, Salzburg, Austria; Neuroscience Institute, Center for Cognitive Neuroscience, Christian Doppler University Hospital, Paracelsus Medical University, Salzburg, Austria; Institute of Public Health, Medical Decision Making and Health Technology Assessment, University for Health Sciences, Medical Informatics, and Technology, Hall in Tirol, Austria, 6UCB, Brussels, Belgium
Objective:
We explored statistically appropriate timepoint(s) for defining possible/probable prolonged seizures by seizure type.
Background:
Most seizures are brief and self-limiting, but some will cluster, grow prolonged and more severe. Defining prolonged seizures accurately informs when to ideally give acute medication to prevent status epilepticus (SE). Prolonged seizures have been defined as >2 minutes for generalized absence seizures and the convulsive phase of bilateral tonic-clonic seizures (focal/generalized onset), and >5 minutes for focal seizures.
Design/Methods:
Post hoc analysis of 2742 seizures from 887 video-electroencephalography (V-EEG) recordings of 725 patients (including outliers). Seizure duration was measured clinically/via EEG and analyzed using Tukey’s box plot. For each seizure type, outliers were categorized as possibly (>1.5×IQR above Q3) or probably (>3×IQR above Q3) prolonged. Patients with SE were excluded (n=20), which is a limitation.
Results:
Using Tukey’s method, 4% (97/2742) of seizures observed clinically and 6% (159/2742) of those observed via EEG were classified as either possibly/probably prolonged. Cutoffs (clinical/EEG duration, seconds) for possibly prolonged seizures: focal preserved consciousness 48.75/60.5; focal consciousness unknown 48.5/70.25; focal tonic 45.75/51.75; focal to bilateral tonic-clonic 167.75/189.5; focal clonic 48.5/56.5; focal impaired consciousness 111.5/146.75; atonic 1.9/1.6; generalized tonic 24.5/21.25; eyelid myoclonia 6.3/6; myoclonic tonic 9/7; absence with eyelid myoclonia 10/9; generalized clonic 14/15.25; atypical absence 44.25/50; generalized tonic-clonic 127/120.25; tonic spasm 15.5/15.5; typical absence 20/23.25; myoclonic absence 20.75/21.
Conclusions:
These data suggest an appropriate focus on absence, tonic-clonic and focal seizures as being more prolonged and support the 2 minutes cutoff for tonic-clonic seizures (focal/generalized onset). However, the statistical cutoff of ~1 minute for absence seizures and 1-2 minutes for focal seizures suggests the consensus definition may be too conservative at 2 and 5 minutes, respectively. Focal preserved consciousness seizures may become prolonged from 1 minute, and focal impaired consciousness seizures from 2 minutes. Other generalized seizure types may become abnormally prolonged earlier.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.