Abstracts

Rationale: Video electroencephalographic (VEEG) monitoring is widely used as a diagnostic tool for patients suspected of a seizure disorder. A common goal of an inpatient admission for VEEG is to capture the patient’s typical events with continuous video and EEG monitoring so that a definitive diagnosis can be made. A definitive diagnosis dictates treatment, decreases long-term costs and, in the case of a non-epileptic seizure diagnosis, reduces the risk of unnecessary anti-epileptic drug (AED) therapy. VEEG is a highly specific test, but is not very sensitive. Repeat admissions for VEEG may increase diagnostic yield, but the test is resource intensive, time-consuming, expensive, and poses some potential risk to patients. The purpose of this study is to determine the diagnostic value of repeat admissions for VEEG and to determine if commonly available clinical information can predict the diagnostic yield of a repeat study. Methods: All patients completing scalp VEEG monitoring at the UAB Epilepsy Center from 2002-present were screened (n=3,727). Minors, mentally retarded patients, and patients readmitted for surgical procedures or pre-surgical localization were excluded. Of the remaining 3,183 patients, any patient with a non-diagnostic VEEG and at least one repeat admission was included in this study. The following data were extracted from a chart review: number of admissions, final diagnosis, suspected seizure type, pre-admission seizure frequency, length of stay, age, sex, race, and the presence of epileptiform discharges during the non-diagnostic study. Multiple logistic regression was used to determine if any of the parameters could predict the diagnostic yield of repeat VEEG studies. Results: Of the 3,183 patients, 87 (2.7%) were readmitted due to a non-diagnostic first test. Just over half of these were diagnosed in a subsequent study (48, 55.2%) with the rest remaining undiagnosed. Of the 48 patients, 46 (96%) were diagnosed with the first readmission and two by the second. The total duration of monitoring was a mean 190.2 hours for those diagnosed and 169.1 hours for those never diagnosed. 33 (68.8%) patients were ultimately diagnosed with epilepsy, 15 (31.2%) with non-epileptic seizures, and 2 (4.2%) with both. Only age was independently predictive of a diagnostic readmission (p<0.05) with younger patients more likely to be diagnosed, while total duration of monitoring was suggestive (p=0.07). Number of admissions, sex, race, seizure type, seizure frequency, diagnosis, and interictal findings were not predictive of a diagnostic readmission. Conclusions: Only 2.7% of patients with non-diagnostic studies were readmitted and a repeat VEEG study was only useful in 55.2%. However, most were diagnosed with only one additional admission. Seizure frequency was not predictive because patients’ assessments of seizure frequency were extremely inaccurate. While patient age and length of stay may predict diagnostic yield, clinical expertise remains the best tool for planning potential readmissions for VEEG monitoring.