Abstracts

Rationale: Ambulatory EEG (AEEG) or home EEG monitoring is a widely used investigation in children with epilepsy or paroxysmal events suspicious for epileptic seizures. The duration of monitoring varies from one to several days. We report the utility of 50 consecutive AEEG monitoring studies from a single institution. Methods: Retrospective chart review was conducted and information gathered included age, gender, reason for AEEG monitoring, duration of monitoring, antiepileptic drugs (AEDs) at the time of the study, frequency of events, interval between the decision for AEEG monitoring and the study itself , result of prior EEG, and result of AEEG.. The study was determined to be helpful if it influenced clinical decision making by providing additional data when compared to the prior EEG. Results: The median age of the patients was 10.25 years (IQR 6.15-14.86), 58% males and 42 % females. The duration of AEEG monitoring was 1 day in forty-one patients, 2 days in four patients, and 3 days in five patients. Epilepsy was already diagnosed in 32 patients. Twenty-nine patients were on 1 AED, four on 2, two on 3, and fifteen were not on AED, at the time of AEEG. Median time to AEEG acquisition was 3-4 weeks. Twenty-seven patients were monitored for event characterization, 12 with daily events and 8 with several events per week. Twelve patients were monitored to classify seizures . In another 8, the AEEG was obtained in preparation of weaning AED. Five patients were monitored for identification of subclinical seizures. Two patients had more than one indication for AEEG monitoring. Events were captured in 15/27 patients (55.5%) monitored for event characterization. Four of the 12 patients monitored to classify epilepsy had epileptiform abnormalities on AEEG. Of these, 3 patients had epileptiform abnormality on the prior EEG. One of the five (20%) patients monitored for subclinical seizures had an electrographic seizure during AEEG monitoring. Overall, 19/27 (70.3 %) of AEEGs done for event characterization were helpful as compared to 3/12 (25 %) done for seizure classification (Logistic regression: Odds Ratio: 7.1; 95% CI: 1.52 - 33.4; p = 0.01). Impact on treatment was not relevant for AEEGs that were done before AED wean, or for subclinical seizures, as the the ordering physician had predetermined that the result, normal or abnormal, would influence clinical decision making. Conclusions: AEEG can be a valuable diagnostic tool when done for event characterization in appropriately selected patients. It is less likely to be helpful for seizure classification.