Abstracts

EEGs are most useful when they reveal abnormalities supporting the clinical suspicion of epilepsy or when they provide evidence of a specific epilepsy syndrome. Recording sleep during an EEG is one way to try and provoke abnormalities that improve the test[rsquo]s diagnostic yield. We reviewed the likelihood of recording sleep and the usefulness of sleep for provoking abnormalites not present in wakefulness and drowsiness. All outpatient EEGs performed over a 13 month period on children less than 19-year-old were reviewed. They were scored for presence of awake, drowsy, and stage II sleep samples. The presence or absence of epileptiform discharges (ED[rsquo]s) in the awake/drowsy state or sleep state was noted along with the specific type of discharge. 259 EEGs met the criteria for review. All included wakefulness. Drowsiness was recorded in all but 3. Forty (15%) EEGs did not include stage II sleep. Therefore 219 EEGs (85%) included both wakefulness and stage II sleep.
79 records (30%) revealed ED[rsquo]s.
Of the 40 EEGs where no sleep was recorded, 12 (30%) contained ED[rsquo]s.
Of the 219 EEGs with both awake and sleep states recorded, 67 (31%) contained ED[rsquo]s. 6 (3%) of these recordings revealed ED[rsquo]s in wakefulness but not during sleep. 11 (5%) revealed ED[rsquo]s in sleep but not in wakefulness. 50 (23%) showed ED[rsquo]s in EEG samples from both wakefulness and sleep. Of the 11 children with spikes only during sleep, 10 had symptomatic epilepsy; the remaining patient had nonepileptic staring spells and two brief irregular generalized spike wave bursts in sleep (suggesting a seizure diathesis but no supporting the clinical suspicion of absence epilepsy). Most EEGs recorded in children in our outpatient EEG laboratory included sleep without resorting to sedation. Preparation with sleep deprivation and performing the study in a quiet, darkened room with appropriate technologist techniques was usually sufficient to get children to fall asleep. In only 5% of cases where a child did fall asleep did the EEG reveal ED[rsquo]s in sleep but not in wakefulness. While this is not a trivial number, it would only have added 2 [quot]positive[quot] cases among the 15% of our patients who did not achieve stage II sleep. In other words, EEGs performed on a pediatric outpatient population were [quot]false negatives[quot] about 5% of the time.
At the same time, 3% of recordings showed ED[rsquo]s only in the awake state. In this case there would be a 3% false negative rate. This raises a concern for EEGs obtained only in the sedated state, since wakefulness may not be present at all. The best diagnostic yield for a single EEG from a pediatric patient will come from recordings obtained with the child in the awake and asleep states, but repeating the EEG with sedation obtain sleep adds little additional diagnostic information. (Supported by Harman Clinic Fund at Lucile Salter Packard Children[apos]s Hospital)