Abstracts

Despite the fact that refractory status epilepticus (RSE) represents an important cause of neurological morbidity and mortality, there have been few studies of its treatment, particularly comparisons among drugs used for coma-induction. Most studies are case series in which a specific antiepileptic drug (AED) was prescribed, and do not accurately reflect the variety of approaches used in clinical practice. Moreover, the impact on outcome of EEG suppression remains unclear. Among 127 episodes (107 patients) of status epilepticus (SE) collected from a retrospective database of two tertiary-referral hospitals in Boston, we identified cases that were refractory to first-line (benzodiazepines) and second-line (phenytoin valproate, phenobarbital, or other) AED, and needed intubation and induced coma for clinical management. Coma induction was assessed with regard to the AED used and the extent of EEG suppression achieved. Short-term mortality was analyzed in relationship to demographic and clinical variables, and the treatment strategy used. Statistical analysis was performed with Fisher exact tests for categorical variables and t-tests for continuous variables. We identified 49 episodes of RSE occurring in 47 patients. RSE tended to occur more frequently in incident SE than in recurrent SE episodes (p=0.06). Mortality was 23% for RSE and 8% for non-RSE patients (p=0.05); patients with RSE episodes were more likely to lack a history of epilepsy (p=0.01) or of SE (p=0.05) prior to the study period, and had a higher prevalence of nonconvulsive SE with coma (p=0.07). One coma-inducing AED was prescribed in 45%, whereas two ore more were used in 55% of episodes; the most used agents were barbiturates (31 episodes), propofol (27) and midazolam (19); in 20/33 (61%) monitored episodes, the medication was titrated to EEG burst-suppression. Demographics, clinical variables and outcome did not differ significantly between the different coma-inducing strategies, or between episodes with and without EEG burst-suppression. Barbiturates showed a tendency to be used preferentially in episodes related to ominous etiologies. RSE has a higher prevalence in incident than in recurrent SE, and shows higher mortality than non-RSE. In our series, RSE episodes were managed with a combination of coma-inducing AED markedly more often than in series published previously. In this setting, which may more accurately reflect usual clinical practice, mortality appears to be independent of the specific AED used for coma induction and the extent of EEG suppression, suggesting, in agreement with previous studies, that the underlying etiology is the main determinant of outcome. (Dr. Rossetti is supported by the Swiss National Science Foundation and the SICPA Foundation, Prilly, Switzerland.)