Abstracts

Rationale: Differentiating convulsive epileptic seizures (CES) from convulsive nonepileptic seizures (CNES) is difficult without long-term monitoring (LTM) with video-EEG[1]. However, LTM may not be readily available and is often costly [2]. Thus, clinicians rely heavily on description of previous events and medical history to direct the initial treatment and diagnostic interventions [3-6]. The aim of our study is to evaluate which patient demographic and historical information is most associated with CES or CNEE in a year-long cohort of new patients referred to a tertiary-care epilepsy clinic who received early LTM referral. Methods: We retrospectively reviewed all Vanderbilt LTM reports for a period of 12 months (October 2010 to October 2011). We identified studies performed on adult new patients who underwent LTM for characterization/classification of undiagnosed events. We only included in our analysis LTM studies that recorded typical convulsive events. The corresponding initial epilepsy clinic notes were reviewed to retrieve patient demographics, past medical histories, epilepsy risk factors, and detailed event description (Table 1). A total of twenty variables with binary, rank order, or continuous data were then analyzed utilizing Kruskal-Wallis tests to compare three groups: CES, CNES and the combined group (CES + CNES). Since there were a large number of variables analyzed relative to subject number, a statistical value of p ≤ 0.01 was considered significant. Results: We identified a total of 119 subjects (with a median age of 42.5 years) meeting inclusion criteria (Table 1). CES was diagnosed in 32 patients, CNES was diagnosed in 80 patients, and both CES and CNES were diagnosed in the remaining 7 patients. The average age of presentation to the epilepsy clinic was similar in all groups (p = 0.712). CNES had a later reported age of onset at 26.1 years (p= 0.007) compared to the CES group (27.2 years of age) and the combined group (24.1 years of age). CNES had a nearly significant female predominance at approximately 80% (p = 0.017). Regarding event history and historical data, the subject having an event during a previous clinical encounter (p = 0.01) or having a history of depression (≤ 0.001) significantly favored a diagnosis of CNES. In contrast, a history of febrile seizures had a statistically significant association with a diagnosis of CES (p = 0.001). There were no differences in the remaining parameters among the three studied groups. Conclusions: Our preliminary findings demonstrate the utility of some clinical features in favoring a diagnosis of CES or CNES. Extension of this study to include a larger number of LTM studies may help to identify additional features of value.