The utility of routine EEG in the diagnosis of sleep disordered breathing

Sleep disordered breathing (SDB) is a common medical condition. Its manifestations of snoring, nocturnal choking, arousals, and sleep fragmentation can lead to excessive daytime sleepiness, neuropsychological slowing, lapses of consciousness, and accidents that can be misinterpreted as epileptic phenomena. Moreover, patients with documented epilepsy commonly exhibit similar symptomatology because of the undiagnosed coexistence of sleep apnea. Therefore, a large proportion of patients referred to the electroencephalogram (EEG) laboratory primarily to confirm or refute the diagnosis of epilepsy could suffer from latent sleep apnea and the routine EEG has the potential to divulge it. We retrospectively evaluated the reporting of sleep apnea symptomatology (snoring, choking, gasping/deep breath, apnea, desaturation, excessive drowsiness) in routine inpatient and outpatient adult EEG studies performed in our institution over the past 12 years (39,130 studies, approximately half of which recorded at least early stages of sleep). Comparisons were performed with the medical records to ascertain the coexistence of objectively diagnosed SDB with polysomnography before or after the EEG study and the importance of reporting variations in assisting with the diagnosis. Two illustrative examples are provided. Sixty-nine EEG studies were identified, performed primarily to confirm, or refute the diagnosis of epilepsy. The mean age of the subjects at EEG was 64 years (range 30-89), and 55 (80%) were male. 36% of them suffered from known epilepsy. Snoring was the most commonly reported sign in 48 (70%) of the studies, followed by arousals in 29 (42%), apnea in 16 (23%), excessive drowsiness in 13 (19%), gasping/deep breath in 9 (13%), and desaturation in 7 (10%). A sleep disorder was suggested in 25 (36%) of the interpretations and a direct recommendation for a sleep study was made in 22 of them (32%). This interpretation was included in the impression of the report in 21 (30%) of the cases, in the detail in 20 (30%) of the cases and in both in 28 (40%). Only 14 (20%) patients underwent polysomnography, and all of them were formally diagnosed with SDB. Seven (50%) of them were diagnosed with obstructive sleep apnea, 2 (14%) with central sleep apnea, 3 (22%) with both, 1 (7%) with upper airways resistance syndrome, and 1 (7%) with primary snoring. From these 14 patients, 9 (64%) were diagnosed with a sleep study performed after the EEG, 4 (29%) before the EEG interpretation, and 1 (7%) had a repeat study after the EEG. In the logistic regression model applied, with the exception of the presence of arousals (odds ratio = 4.63, P = 0.033), none of the aforementioned symptomatology or the reporting of suspicion for SDB or the location (impression vs. detail) of the reporting showed a statistically significant association with the completion of a sleep study. Routine EEG offers a unique opportunity of direct clinical observation along with electrophysiologic and cardiorespiratory monitoring. When sleep is recorded, it can help identify clinical and electrographic features of sleep apnea and prompt confirmation with a polysomnogram in the appropriate clinical context. It can therefore serve as a valuable, adjunctive tool for the diagnosis of SDB. Our data highlight that potential but unveil its decreased use in the neurology community. Increased awareness is required by the EEG technologists, interpreting neurologists, and referring physicians, regarding reporting and using sleep apnea features on the EEG.