Abstracts

Rationale: Sudden unexplained death in epilepsy (SUDEP) is a common cause of death in persons with epilepsy. To further explore the mechanism of SUDEP, simultaneous recording of respiratory parameters and electroencephalography (EEG) is necessary. However, no studies to date assess the feasibility of respiratory monitoring in this population or signal integrity of respiratory sensors during ictal recordings. The purpose of this study was to determine the integrity of respiratory signals during seizures recorded from patients with epilepsy who underwent laboratory polysomnography (PSG).Methods: We retrospectively reviewed laboratory PSGs recorded from December 2007 to May 2011. The most common reason for referral was for the evaluation of sleep disorder symptoms. Antiepileptic therapy was not altered in advance of testing. Fifty-one seizures were recorded. Critical sensors included nasal pressure, thermal airflow sensor, and oxygen saturation (SpO2). Non-critical sensors included respiratory effort, end-tidal CO2 and electrocardiogram. Seizure duration and type and duration of respiratory events (apnea, hypopnea) were measured. Short seizures were defined as seizure duration of ? 60 seconds. Long seizures were defined as seizure duration of > 60 seconds.Results: Of 51 seizures recorded, 34 (66.7%) were short seizures (mean of 26.6 seconds, range of 11 to 51 seconds) and 17 (33.3%) were long seizures (mean of 119.9 seconds, range 65 to 257 seconds). For all short seizures, signals from both critical and non-critical sensors were interpretable. For long seizures, 2 seizures (11.8%) lost signal from 1 critical sensor and 1 seizure (5.9%) lost signal from 1 non-critical sensor; these signals were deemed to be uninterpretable. Only 1 seizure required immediate post-event sensor maintenance. Ten (29.4%) of the short seizures were associated with ? 1 respiratory events (1 obstructive apnea, 7 central apneas, 2 hypopneas, mean duration 18.0 seconds). Twelve (70.6%) of the 17 long seizures were associated with ? 1 respiratory event (7 obstructive apneas, 6 hypopneas, mean duration 34.4 seconds).Conclusions: Regardless of the duration of seizure, respiratory signal integrity appears to be adequate during ictal recording in the sleep laboratory. Almost none of the seizures required post-event sensor maintenance. Thus, incorporation of respiratory sensors with EEG recording is achievable and may aid in the understanding of the mechanism of SUDEP, perhaps in both attended and unattended/home settings.