Abstracts

Rationale: Scant data exist on inpatient seizure epidemiology and management. This study examined in-hospital patterns of seizure occurrence with a particular focus on phenytoin use. Methods: Medical records were screened for inpatient seizures for all patients over age 17 admitted to Tisch Hospital (NYU) in 2007 who were discharged with ICD9 codes 345.xx or 780.3x or who were treated at any time with phenytoin, fosphenytoin, levetiracetam, phenobarbital or valproic acid. Subjects were excluded if they were admitted to manage seizure(s) that occurred prior to hospitalization. Statistical comparisons were performed with chi-square, Fisher’s exact or Wilcoxon rank sum tests. Results: We identified 112 cases, of whom 66 (61%) had new-onset seizures (NOS). A neurologist consulted in 106 cases (95%) during hospitalization. Multiple seizures occurred in 73 cases (65%) and seizures occurred on multiple days in 44 (39%). Age, choice of anticonvulsant, location at time of seizure and service at time of seizure were not significant predictors of recurrent seizure (data not shown). Status epilepticus (SE) was the index seizure in 10 (9%). SE cases were predominantly treated with phenytoin (50%) and had a high risk of death (50%). Clinical features of new-onset seizure cases versus cases with a history of seizure or epilepsy are presented in Table 1. After excluding cases with a history of epilepsy or with status epilepticus as the index seizure, initial drug selection in the remaining 60 cases was: phenytoin 26 cases (43%), levetiracetam 24 (40%), valproic acid 1 (2%), magnesium 1 (2%), midazolam 1 (2%), phenobarbitol 1 (2%), no treatment 6 (10%). Phenytoin was prescribed at discharge in 8 cases (16%, excluding deaths). Likelihood of phenytoin use did not vary with age, seizure etiology, service at time of index seizure or occurrence in an ICU (data not shown). Conclusions: Cases with new-onset seizure were older, more likely to be in an ICU, more likely to be treated with phenytoin, and had higher mortality than cases with a history of seizure or epilepsy. Stroke was the most common etiology. Most cases had recurrent seizures. Phenytoin was most often the initial choice to treat new-onset seizures (excluding status epilepticus) but was less often prescribed at discharge.