Abstracts

Rationale:
Continuous long-term video EEG monitoring is the ‘gold-standard’ diagnostic test to assess whether epilepsy is the etiology of a seizure-like event. Ambulatory EEG is an alternative but has difficulties in differentiating artifacts from ictal/interictal phenomena, and characterization of event semiology The purpose of this study was to review the yield and usefulness of ambulatory EEG within the pediatric neurology/epilepsy population, quantified based on age, quality of recording, characterization of seizures/events of concern, outcome, and need for subsequent continuous long-term video EEG monitoring.
Method:
This study is a retrospective review of pediatric patients who underwent ambulatory EEGs in  a comprehensive Epilepsy Program. Studies were performed with standard 10-20 electrode placement using Natus Xltek Trex systems. The primary outcome was determination of whether the study was deemed helpful in terms of answering the question specified on the requisition.  Other data collected included patient age, prior  diagnosis of epilepsy, previous routine EEG, quality of  EEG recording, indication for the study, reported event/seizure frequency, whether events were captured, if events were classifiable, and whether subsequent continuous long-term video EEG monitoring was performed.
Results:
A total of 116 studies were analyzed.  67% studies were deemed helpful, and another 4% were deemed partially helpful.  Age range was 1 to 17 years, with an average age of 9.98 years.  Age was a predictor of helpfulness, with 66-73% of studies being helpful in those studies performed older than 4 years and less helpful (50%) in those under the age of 3 (Figure 1). 52% studies maintained good quality throughout the recording. Quality was a predictor of helpfulness, in that 72% of good quality recordings were helpful.  Age was a factor in predicting quality, with only 33% good quality studies in the 4-6 years group, and good quality in 50-57% in the other age group. The indication for the study was divided into 4 groups: characterization of event (seizure vs. non-epileptic), quantifying seizure activity in known epilepsy, determination of electrical status epilepticus in sleep (ESES), and other.  The highest prediction for helpfulness was in those studies for ESES (92%), as well as in those for quantifying seizure activity in known epilepsy (86%), helpfulness for other indications was 81% and least helpfulness occurred in those studies which were performed for characterization of events at 51%. 43% studies captured events, out of those studies 88% were able to classify events.  The likelihood of capturing events was highest (88%) with reported event frequency as multiple times per day, this dropped to 44% in which reported event frequency was 1-6/week. 11% required further in-patient continuous video-EEG monitoring after ambulatory EEG.
Conclusion:
This study enabled us to determine that classification of interictal had the highest yield of usefulness for ambulatory EEG studies, followed by quantifying seizure activity in patients with known epilepsy.  Patient age, a prior abnormal routine EEG and frequency of events were positive predictors of usefulness.  Ambulatory EEGs had a relatively low yield for determining if events were seizure vs. non-epileptic events.  With this information we are in the process of developing a pathway for determining yield of ambulatory EEG which may guide clinicians in the decision-making process regarding ordering an ambulatory EEG vs in-patient continuous video-EEG monitoring (Figure 2).
Funding:
:Dr Jacobs and the Epilepsy Program are supported by the Alberat Childrens Research Insitute and Alberta Childrens Hospital Foundation