Abstracts

Rationale: Neonatal seizures affect 2-4 in 1000 live births and are most commonly treated with phenobarbital (PHB). Current data suggest that prolonged treatment with PHB does not prevent postnatal epilepsy (PNE) but may have a negative impact on cognitive development and prolong duration of hospitalization. In our study, we sought to determine the extent of PHB use, determine factors that may influence prolonged PHB use, and evaluate whether previous data have influenced a change in practice at our institution. Methods: This study involved a single site retrospective chart review of 84 neonates with symptomatic seizures before 44 weeks postmenstrual age. All infants identified had clinical and/or electrographic seizures at presentation. Etiology of seizures, onset and duration of seizures, antiepileptic therapy, and incidence of PNE were recorded. This data was then analyzed for factors associated with duration of PHB treatment which reflect provider treatment decisions Results: Median duration of PHB treatment was 80 days with only 22% of children weaned from PHB prior to one month of treatment. Duration of PHB treatment was associated with underlying etiology of seizures (p=0.022). Use of a second antiepileptic at the time of PHB wean (p=0.016) was associated with earlier discontinuation. The presence of seizures seen on EEG did not impact duration of treatment (p=0.48). Children had a median follow-up of 17.9 months. Nineteen children (23%) went on to develop postnatal epilepsy. Younger gestational age at birth (p=0.018) and longer duration of seizures at onset were associated with PNE (OR 1.16, 95% CI 1.029-1.308). Conclusions: This study demonstrates that despite clinical evidence spanning the past decades and expert recommendations, provider practice has not changed. Our rate of discontinuation prior to one month, 22%, is similar to that seen by Guillet and Kwon in 2008 of 16%. Providers appear to use factors such as etiology and the ability to provide prophylaxis with other anti-epileptic medications to determine PHB prophylaxis duration. Although significant evidence exists that clinical identification of neonatal seizures is unreliable, this does not appear to impact duration of PHB use. To alter the practice of PHB discontinuation, we propose establishing institutional Guidelines of Care to affect change and address possible barriers to implementation. Funding: None