Abstracts

Rationale: Neonatal status epilepticus (NSE) is frequently encountered in the neonatal intensive care unit, most often as the result of acute brain injury or infection. It is believed that electrographic cessation of seizures, rather than clinical resolution, is vital to improving outcome in these infants. The first-line treatment of phenobarbital has a 50% efficacy, which only slightly improves with the addition of the traditional second-line phenytoin. There has been growing interest in the use of midazolam in neonates, with recent literature illustrating variable results. Our objective is to characterize the electrographic response of NSE to continuous midazolam infusion, with specific attention to efficacy, rate of response and adverse effects. Methods: All high risk infants with clinical seizure or moderate-to-severe hypoxic-ischemic injury were monitored on continuous video EEG. All infants received an initial loading dose of phenobarbital, repeated up to total of 30-60 mg/kg. Patients identified as being in NSE despite phenobarbital received IV midazolam bolus (0.1 mg/kg), while continuing on EEG monitoring. If seizures persisted, midazolam was administered by continuous infusion (0.1 mg/kg/hr), and titrated as necessary. Results: Six patients were identified in a 3 month time period as being in NSE despite loading doses of phenobarbital. Seizures started between 0 hours of life and 10 days of life, and were due to various etiologies (see table 1). The midazolam was started within 24 hours of seizure identification in 5 of the 6 patients. Status epilepticus (SE) stopped with initiation of versed in all patients. Electrographic seizures stopped after versed bolus in 1 patient, within several minutes after initiation of midazolam drip in 2 patients, within hours in 2 patients, and decreased dramatically in 1 patient who was subsequently withdrawn for septic shock due to complete bowel infarction. All 5 surviving newborns remained seizure-free on phenobarbital alone after removal of versed drip. There were no notable clinical side effects from the midazolam infusion. Conclusions: Midazolam infusion should be considered an option in the management of NSE. It works rapidly to control status epilepticus, often resulting in complete cessation of seizures, even in those patients who had been refractory to high levels of phenobarbital. More attention and significant work needs to be done to provide better age appropriate options of SE in the NICU population.