Abstracts

Because of its favorable side effects and safety profile, levetiracetam (LEV) has frequently been used in place of other AEDs in the treatment of patients admitted to the neurosurgical ICU for seizure prophylaxis or treatment. Hemorrhage, tumors, and trauma are the frequently reported reasons why this AED is used, but population based studies of efficacy and safety of LEV in the NSICU setting are lacking. In this study, we compare the clinical status and outcomes of patients treated with LEV vs. other AEDs. We retrospectively reviewed 552 charts of patients admitted to NSICU at the University Hospital in Cincinnati between 9/1/03 and 9/30/04. Our measures of outcome were discharge GCS and difference in GCS between admission and discharge, clinical impression of the treating physician (no improvement vs. improvement), and recurrence or new occurrence of seizures in patients receiving LEV vs. other AEDs. We conducted bivariate tests (t-tests and Pearson [chi]2-tests) for differences between groups of patients who received or did not receive levetiracetam. Analyses were performed for subsamples of patients with the most frequent diagnoses (tumor, trauma, ICH, SAH, SDH) and for the entire sample. 358 patients received AEDs during the stay in the NSICU. Data on 16 patients was excluded due to death during admission. Final sample included 342 patients. 124/358 (34.6%) received LEV with dose ranging between 250 mg bid and 2000 mg bid with most frequent titration of 500 mg bid for 1-3 days and then an increase to 1000 mg bid. LEV did not have an effect on the discharge GCS for all diagnoses or the entire sample (all p[gt]0.156) except slightly lower discharge GCS in patients with the diagnosis of trauma (n=96) treated with LEV vs. other AEDs (11.2 vs. 12.9, respectively; p=0.009). The clinician[apos]s impression of improvement was significantly different only for patients with brain tumors (n=72): 8/32 (25%) of patients who received LEV were deemed to be improved while only 3/40 (7%) of patients who did not receive LEV improved (p=0.04). Finally, of 38/342 patients who had seizures after the admission to the NSICU, 27/119 (p[lt]0.001) were started on LEV vs. 11/224 (4.9%) were started on other AEDs (p[lt]0.001). While there was no correlation between age and treatment with LEV at the time of admission to ICU (old defined as age [gt]64 years; p=0.8); therapy with LEV during NSICU stay was preferentially started in older patients (r=0.245; p[gt]0.001). In general, no major differences in outcomes were found in NSICU patients treated for seizures or seizure prophylaxis with standard AEDs vs. LEV. These preliminary findings should be treated with caution until more data become available, especially data based on prospective studies. (Supported by UCB Pharma.)