Predictors of length of stay in children admitted for presurgical evaluation for epilepsy surgery

Rationale One challenge for families whose children are undergoing presurgical evaluation for epilepsy surgery is the unpredictable length of hospitalization for video-electroencephalograph monitoring. The goal of this study was to retrospectively evaluate length of stay in children admitted for presurgical evaluation at a tertiary referral center. Methods Duration of stay for children with medically intractable epilepsy admitted for presurgical evaluation to the Pediatric Epilepsy Monitoring Unit at Mayo Clinic Rochester between 2010 and 2013 was evaluated retrospectively. Results Of 140 children, surgical candidacy was determined in 122 (87.1%) (72 candidates, 50 noncandidates). The mean length of stay was 4.0 ± 3.7 days and was not predicted by candidacy for surgery, age at monitoring, duration of epilepsy, number of antiepileptic drugs at admission, or focal/hemispheric magnetic resonance imaging abnormality. Shorter length of stay was predicted by younger age at epilepsy onset (P < 0.05) and shorter interval since most recent seizure (P = 0.001). Subtraction ictal single-photon emission computed tomography coregistered to magnetic resonance imaging was performed in 43 (35.2%) children, and correlated with longer length of stay (mean 5.1 ± 4.1 days for subtraction ictal single-photon emission computed tomography coregistered to magnetic resonance imaging users versus 3.5 ± 3.3 days for nonusers, P = 0.022). Antiepileptic drugs were reduced either upon or after admission in 67 (54.9%) children, and the length of stay was significantly longer in these patients (mean 5.5 ± 4.1 days if antiepileptic drugs were reduced versus 2.2 ± 2.1 days if not reduced, P < 0.001). Conclusions Significant predictors of shorter length of stay include younger age at epilepsy onset, shorter interval from most recent seizure, lack of subtraction ictal single-photon emission computed tomography coregistered to magnetic resonance imaging, and lack of need for AED reduction on or after admission.