Abstracts

Rationale: While early surgical treatment can be efficacious for certain types of medically-intractable epilepsy, one challenge for families considering pre-surgical evaluation for their child is the unpredictable length of hospitalization required to determine surgical candidacy. Our goal is to retrospectively evaluate length of stay for evaluation of surgical candidacy in pediatric epilepsy patients at Mayo Clinic, Rochester, a tertiary referral center. Methods: Duration of stay for pediatric patients with medically-intractable epilepsy admitted for pre-surgical evaluation to the Pediatric Epilepsy Monitoring Unit at Mayo Clinic Rochester between 2010 and 2012 was evaluated retrospectively. Surgical candidacy was classified as candidate, noncandidate or indeterminate on the basis of scalp EEG monitoring (all patients) in addition to SISCOM (subset of patients). Length of stay and its predictors were evaluated for candidates and noncandidates. Results: Of 67 children admitted for pre-surgical evaluation, surgical candidacy was determined in 60 (89.6%), of whom 44 were candidates and 16 were noncandidates. The median duration of stay for these children was 3.0 days (interquartile range 1.9, 6.3). Length of stay was not predicted by candidacy for surgery, age, duration of epilepsy diagnosis, or abnormal MRI. Shorter length of stay was predicted by younger age at epilepsy onset (p=0.03), higher number of antiepileptic drugs (AEDs) on at admission (p=0.01), and shorter interval since most recent seizure (p=0.03). SISCOM was performed in 20 children, and its usage was correlated with a nonsignificant trend toward longer length of stay (mean length of stay 5.7 ± 4.6 days for SISCOM users versus 4.0 ± 3.7 days for nonusers, p=0.11). AEDs were reduced either upon or after admission in 29 (48.3%) children and length of stay was significantly longer in these cases (mean length of stay 6.5 ± 4.6 days if AEDs were reduced versus 2.8 ± 2.5 days if not reduced, p<0.001). Conclusions: Median length of stay was only 3 days; however, the range was broad. Factors predictive of shorter length of stay include younger age at epilepsy onset, higher number of AEDs at time of admission, and shorter interval since most recent seizure. SISCOM use was associated with a nonsignificant trend of longer length of stay. Understanding the predictive value of these factors on length of stay can assist families in planning pre-surgical evaluation for their child.