Abstracts

Rationale: Predictors of outcome have been described in newborns undergoing continuous electroencephalogram monitoring (cEEG) in the intensive care unit (ICU). However, the timeliness of seizure detection and treatment has not been evaluated in previous literature. The aim of this study is to describe factors, including timeliness, associated with short-term mortality in newborns undergoing cEEG in the ICU.Methods: Retrospective observational study. We considered patients from birth to 28 days of life after gestational age 40 weeks who underwent cEEG (defined as more than 3 hours of uninterrupted EEG) in the ICU at Boston Children’s Hospital during the period 2011 to 2013. We excluded patients in whom cEEG was performed in the setting of an elective admission. In patients with multiple episodes on cEEG in the ICU, only the first recording was considered. The main outcome measure in this study was in-hospital mortality. Newborns were defined as patients with age less than 28 days of life since 40 weeks of gestational age. We considered prematurity as birth before 37 weeks of gestational age. For the purposes of this study, we considered electrographic seizures as any seizure detected on cEEG, including electro-clinical and electrographic-only seizures.Results: We reviewed 211 newborns (55.0% males, 24.8% prematures) with a median (p25-p75) postnatal age of 3.95 (0.9-13.1) days. Etiology was structural symptomatic in 123 (58.3%) newborns, non-structural symptomatic in 24 (11.4%), unknown in 4 (1.9%), and a non-central nervous system cause was identified in 60 (28.4%). The median (p25-p75) time from ICU admission to cEEG initiation was 15.1 (4.9-104.8) hours. Electrographic seizures occurred in 74 (35.1%) newborns. Electrogrographic status epilepticus (SE) occurred in 8 (10.8%) newborns. The time from admission to cEEG had a median of 101.8 hrs (7.48(p25)-171.7(p75)) in patients who died. Of the patients who survived their median time to EEG from admission was 14.4 hrs (4.7(p25) – 65.2(p75)). On univariate analysis, prematurity (p=0.02), the presence of electrographic SE (p=0.02), and time from ICU admission to cEEG (p=0.003) were associated with an increased mortality while gender, etiology, and time from ICU admission to first detected seizure were not (Table 1). After controlling for gender and prematurity, factors associated with increased mortality were prematurity OR=2.63 (95% CI: 1.06-6.50), p=0.037, presence of SE OR=8.82 (95% CI: 1.74-44.57), p=0.008, and time from ICU admission to initiation of cEEG OR=1.002 (95% CI: 1.001-1.004), p=0.008, but gender, and presence of seizures were not (Table 2).Conclusions: Both seizure burden (presence of SE) and time from ICU admission to initiation of cEEG were predictors of mortality in newborns undergoing cEEG in the ICU.