Abstracts

Rationale: Epilepsy remains a clinical diagnosis supported but not always defined by quantitative tests such as encephalography (EEG). Nevertheless, accurate diagnosis is warranted given lifelong antiepileptic costs, side effects, and the need to refine diagnosis in patients who continue to experience spells despite adequate medical therapy. The VA system has a particular need for accurate neurodiagnostics due to the increased incidence of blast trauma and head injuries sustained by returning veterans who present with stereotyped events but with comorbid risk factors for non-epileptic spells. Unfortunately, routine EEG testing has low sensitivity, often requiring multiple studies to establish a diagnosis of epilepsy. Long term inpatient video EEG monitoring remains the gold standard for evaluation of stereotyped spells, but is costly, not universally available, and may not be suitable for infrequent spells or those with particular triggers. Ambulatory EEG has recently been shown to be a viable alternative to inpatient monitoring, but has been poorly studied in the outpatient setting. Here, we investigate the sensitivity and cost efficacy of ambulatory EEG in routine epilepsy diagnosis.Methods: Between 5/1/2012 and 5/1/2013 48 patients underwent ambulatory EEG (24-72 hours, mean 47 hours) testing at the VA Boston Healthcare System as either part of their diagnostic workup after inconclusive routine EEG testing or to evaluate for ongoing seizures despite AED therapy. 43 patients had undergone an average of 2 routine EEGs prior testing. 16 patients had also undergone inpatient video EEG monitoring. The diagnostic sensitivity and costs of EEG testing modalities were calculated and compared.Results: Ambulatory EEG results altered the diagnosis in 20 (47%) patients, demonstrating previously undiagnosed epileptic discharges or seizures in 12 (28%) patients and capturing NES spells in 2 (5%). . Ambulatory EEG results affected a management changein 19 patients, including 9 patients in whom additional AEDs were added. The ambulatory EEG was normal in 6 (14%) patients in whom a subsequent inpatient LTM captured epileptic abnormalities. The sensitivity of ambulatory EEG for epileptic discharges was 73%, of routine EEG was 38%, and LTM was 75%. The sensitivity of ambulatory EEG for any EEG abnormality other than generalized background slowing was 82%, EEG was 70%, and LTM was 93%. Based on these values we estimate that the ambulatory EEG has roughly the diagnostic yield of 3 routine EEG tests. Overall costs would have been minimized by performing ambulatory EEG as the second line test after a non-diagnostic routine EEG.Conclusions: Ambulatory EEG recording in the outpatient setting is a cost effective option providing a higher likelihood of definitive diagnosis compared to routine EEG testing and offers significant cost savings over inpatient video EEG monitoring. This preliminary data argues for expanded use of ambulatory EEG in routine epilepsy workup. While further study is necessary, we believe that ambulatory EEG should be considered earlier in routine epilepsy workup.