Abstracts

The yield on intensive inpatient monitoring depends on, among other variables, the reasons for admission and the duration of monitoring. The latter variable is linked to the length of patients hospital stay (LOS), which in turn is closely monitored by third party payors and hospital administrators. We undertook this study to assess the yield of intensive inpatient seizure monitoring in the context of short term LOS. We retrospectively reviewed our database of adult patients electively admitted to the Epilepsy Monitoring Unit (EMU) at Westchester Medical Center for further diagnostic evaluation and management, during a 5-year period February 1999-February 2004. All cases were evaluated by one of two adult epileptologists as outpatients prior to EMU admission. The reasons for EMU admission included: 1) to document event, 2) to establish specific diagnosis, 3) to classify seizure type, 4) to localize seizure focus, 5) pre surgical evaluation and 6) antiepilepsy drugs (AED) adjustment. Pre and post EMU variables to be analyzed included seizure diagnosis, EEG interictal and ictal abnormalities, documentation of habitual or new clinical events, and AED treatment. Admitting diagnosis was sub classified as follows: clinical diagnosis of established epilepsy, probable epilepsy, suspected nonepileptic psychogenic seizures (NEPs), or simply to rule out (r/o) seizure disorder. Activation procedures consisted of sleep deprivation, hyperventilation, photic stimulation, and aggressive AED taper. 123 subjects were identified. The age range was 18 from 86 years (mean 38 years) 43 male, 80 female. Mean length of stay was 3.0 days. Initial admitting diagnosis: established epilepsy 56/123, probable epilepsy 30/123, NEPs 22/123, r/o seizure disorder 15/123. Clinical events were recorded in 81/123 (65.8%). Out of 81, 54 (66.6%) had ictal EEG correlate. 25/81 subjects (30.8%) were diagnosed as definite NEPs. In two subjects, the study was inconclusive due to failure to capture the habitual clinical events. In our study, positive diagnostic yield (specific diagnosis established) was obtained in 79/123 (64%) in contrast to a negative yield or inconclusive study in the remaining 44 patients. We consider that the reasons for the high positive yield despite the short restrictive LOS include careful and consistent adherence to admission criteria and a systematic approach with activation procedures implementation. Our experience has demonstrated that carefully planned intensive seizure monitoring is a cost effective procedure in a restrictive managed care environment.