Abstracts

Rationale: To identify the probability of a diagnosis being established depending on the duration of video EEG monitoring. Additional aims are to determine the relationship between video EEG results and different clinical characteristics. Methods: We studied EMU (epilepsy monitoring unit) length of stay and assessed the utility of prolonging studies in patients who had not yet received a diagnosis. Clinical characteristics in 212 consecutive patients admitted for scalp video-EEG monitoring were recorded. We collected data including reason for admission, frequency of seizures/spells, gender, age, age of seizure onset, handedness, family history, history of neurologic disease, current and past Antiepileptic Drugs (AEDs) and prior work up. Subjects were categorized into five diagnostic groups: epileptic seizures (Epi), nonepileptic events (NEE), mixed epileptic and non-epileptic events (Mixed), nonepileptic events from a physiologic cause (NEEP), and nondiagnostic study without results recorded (ND). Variables for these diagnostic categories were analyzed and the significance of these findings was determined by the non parametric Krushkal Wallis test for discrete/continuous data and Fisher’s exact test for categorical data. In determining the relationship between seizure frequency and the day of first event, we report Spearman’s rank-based correlation and use the bootstrap to derive 95% confidence intervals. Results: The most diagnoses were made during the first day of admission (45%) and by day three, 82 patients remained without a diagnosis. On day three, 25 of these patients (33%) received a diagnosis, on day four seven (22%) additional patients received a diagnosis, on day five 5 patients (35%) received a diagnosis, and by day six only one additional patient (11%) was given a diagnosis. Significant differences were found between diagnostic groups for admission reason, duration of EMU stay, total seizures, age of seizure onset, duration of epilepsy, and seizure frequency. The duration of epilepsy was significantly longer in patients diagnosed with epilepsy, 14 years as compared to three to eight years in the other groups (p<0.001). Age of seizure onset was significantly older in the non diagnostic group and in the NEE group, 28 and 29 years respectively as compared to 15 to 19 years in the other diagnostic groups (p<0.001). There were also significant differences in seizure frequency between groups. There was an average event frequency of 5.0 overall and the greatest frequency was an average of 19.5 events in the mixed group and 13.0 events in NEEP group (p<0.001). Conclusions: Our findings show that the majority of patients are diagnosed within the first two or three days of monitoring and we found a limited benefit of prolonging inpatient video-EEG studies beyond five days. Additionally, patient demographics were significantly different depending on VEEG diagnosis, which can help predict the utility of completing video EEG studies in individual patients. Funding: No funding