Abstracts

Rationale: Neonatal seizures are frequently subclinical with up to 90% of EEG seizures having no clinical correlate. Continuous EEG monitoring has been accepted as an adequate detection system of neonatal seizures and an objective means to judge response to anti-epileptic treatment. Because of varying resources between hospitals, EEG application is not standardized. A comparison was performed of neonatal continuous EEG quality between three neonatal ICUs in an attempt to determine if different application methods may reveal differences in EEG quality. Methods: The NEOLEV2 study is a prospective blinded randomized clinical trial of levetiracetam vs phenobarbital treatment in neonatal seizures. Inclusion criteria for NEOLEV2 are: infants at risk for having seizures or having seizures, term infants (corrected gestational age between 37-44 weeks), and > 2200g in weight. Three sites used varying EEG application methods including: 1) Needle electrodes applied by EEG technicians in Auckland, New Zealand, 2) Scalp EEGs applied by neonatal nurses in San Diego, CA and 3) Scalp EEGs applied by EEG technicians in Oakland, CA.  Five EEGs from each site were sampled to include EEGs of infants who were cooled/sedated and not cooled/not sedated. Two ABPN (American Board of Psychiatry and Neurology) board certified neurophysiologists were blinded to EEG type and analyzed the quality of EEGs using a scoring form created to evaluate for common EEG artifacts associated with application methods.  The second hour of each EEG was reviewed to avoid reviewing any excessive artifacts associated EEG onset. Standard neonatal EEG montages and filters were utilized during review. Needle electrodes applied by EEG technicians were compared to scalp electrodes applied by EEG technicians. Scalp electrodes applied by neonatal nurses were compared to scalp electrodes applied by EEG technicians.  Fisher’s exact test was used for statistical analysis in the two types of comparisons. Results: Needle and scalp electrodes both applied by EEG technicians did not show any statistical difference in all EEG assessments using P< 0.05 as the significance level. Differences in EEG artifact in EKG (p=0.167), pulse (p=1), electrode pop (p=0.5), loss of electrode contact (p=1), electrode movement (p=1), sweat artifact (p=1), salt bridge (p=1), 50/60 Hz electrical noise (p=1), external device artifact (p=1), glossokinetic muscle artifact (p=1), and excessive surface muscle artifact (p=1) did not reach statistical significance.  Scalp electrodes applied by neonatal nurses and scalp electrodes applied by EEG technicians did not show any statistical difference between the application methods in all EEG assessments using P< 0.05 as the significance level.  Differences in artifact in EKG (p=0.524), pulse (p=1), electrode pop (p=1), loss of electrode contact (p=0.5), electrode movement (p=1), sweat artifact (p=1), salt bridge (p=1), 50/60 Hz electrical noise (p=1), external device artifact (p=1), glossokinetic muscle artifact (p=0.524), and excessive surface muscle artifact (p=1) did not reach statistical significance. Conclusions: This sampling study comparing varying EEG application methods showed no statistical difference between 1) needle electrodes vs scalp electrodes when applied by EEG technicians or 2) scalp electrodes applied by neonatal nurses vs scalp electrodes applied by EEG technicians. Very little information is available regarding the use of needle electrodes in the neonatal ICU. Even less is known about utilizing nurses for EEG application.  Due to high hospital costs in the neonatal ICU, these alternative methods of EEG application may be feasible methods to consider when implementing continuous EEG monitoring in the neonatal ICU. Funding: None