Abstracts

Rationale: The benefits of continuous electroencephalogram (cEEG) monitoring in intensive care unit settings are increasingly appreciated. Given the expanding indications for cEEG monitoring, the ability to monitor all patients at high risk for seizures may be limited by local infrastructure. cEEG monitoring during neonatal hypothermia has become the standard of care, with monitoring throughout cooling and rewarming. Recent data suggests that the majority of seizures occur within the first 24 hours of cEEG monitoring. We hypothesized that abnormal brain imaging would help stratify risk for neonatal seizures and help guide risk stratification for neonatal seizures.

Methods: We conducted a retrospective review of EEGs from neonates who received cooling for treatment of neonatal hypoxia. We identified one hundred thirty-four patients. We obtained clinical data through chart review. Archived EEGs were reviewed to determine the presence of electrographic seizures and the time to the first seizure. The EEG background was scored as has been previously described and stratified into two groups: normal/mildly abnormal and moderately/severely abnormal. The presence of imaging abnormalities was determined based on the review of the radiology reports documented in the medical record.

Results: 29% of cooled neonates demonstrate seizures throughout cooling and rewarming. Both a significant increase in the incidence of clinical events concerning for seizures prior to the EEG and an abnormal EEG background was seen in the group with neonatal seizures. The proportion of patients with an abnormal head ultrasound (HUS) did not differ between groups, though few patients with seizures had normal brain MRIs. No patients with both a normal HUS and a normal EEG background demonstrated seizures during cooling.
Of neonates that seized, 83% did so within the first 24 hours of EEG monitoring. In the patients who seized after 24 hours, the frequency of an abnormal EEG background was lower than in the early seizure group. No patients who seized after 24 hours had a normal HUS, though one subsequently had a normal brain MRI.

Conclusions: These data reinforce published findings that a vast majority of cooled neonates seize within the first 24 hours of EEG monitoring. They further support that the risk for new-onset electrographic seizures after 24 hours of monitoring in cooled neonates with both a normal EEG background and a normal head ultrasound is very low. These data may aid in further risk stratification in circumstances where cEEG monitor resources are limited.

Funding: Please list any funding that was received in support of this abstract.: None.