Abstracts

Rationale: Increasingly, hospitals are striving to shorten length of stay (LOS) in an effort to improve care. Beyond the impact on expense, longer stays in the hospital increase the risk of iatrogenic or other complications. The epilepsy monitoring unit (EMU) is a unique hospital setting where the admission goal focuses on capturing one or more events/seizures. Seizure provocation is performed in a controlled manner to minimize the risk of secondarily generalized convulsions and status epilepticus, and to maintain patient safety. To better establish benchmarks for care, we evaluated both the impact of admission objectives and select patient characteristics on EMU LOS. Methods: We reviewed the medical records of 905 consecutive patients who were admitted for diagnostic scalp video-EEG monitoring at NYU s Comprehensive Epilepsy Center from January 1, 2011 to December 31, 2011. Reasons for admission, number of anti-epileptic drugs on admission, and time to target events were documented.Results: Of the 905 patients included, the mean length of stay was 4.8 3.2 days. The median length of stay was 4 days. 95 subjects were admitted for pre-surgical evaluation with a mean LOS of 7.2 days, vs. 4.5 days (p<0.001) for those admitted for diagnostic evaluations. 12.8% (116/905) of patients admitted to the EMU had a previous history of epilepsy surgery and the majority of these patients were admitted for medication adjustment; patients who had undergone prior epilepsy surgery experienced longer length of stays (5.9 days) compared to those without a prior epilepsy surgical history (4.7 days) (p<0.0002). Patients admitted on three or more antiepileptic drugs (AEDs) on admission experienced longer lengths of stay (6.3 days) vs. patients on less than three concomitant AEDs (4.3 days). Patients were further stratified by reason for admission. 494/905 patients were primarily admitted for seizure characterization and these patients experienced a mean LOS of 3.9 days whereas patients who were admitted for medication adjustment (189/205) had a mean LOS of 6.0 days. The mean LOS of patients admitted to rule out subclinical seizures (105/905) was 4.8 days. Conclusions: LOS on an inpatient video-EEG monitoring unit is largely influenced by the indication for admission, which in turn dictates the number of events one attempts to record. Likewise the number of AEDs present on admission, which reflects the complexity of a patient s epilepsy, directly impacts LOS; the presence of 3 or more concomitant AEDs increases LOS by 2 days. The reason is probably multi-factorial; a conscientious effort to avoid seizure exacerbation and enhance patient safety during AED reduction in refractory patients is a likely contributing factor. These variables should be considered when analyzing LOS metrics for EMUs, and argues against a one size fits all approach.