Abstracts

Rationale: Extracorporeal membrane oxygenation (ECMO) is a life-saving heart and lung bypass procedure that can often induce substantial EEG artifact. We examined whether using conductive plastic EEG electrodes in place of conventional gold electrodes reduced electrical artifact in standard and extended clinical EEG studies of pediatric ECMO patients.Methods: Initially, 21 consecutive routine and continuous EEG studies (14 using gold electrodes and 7 using plastic electrodes) recorded during ECMO therapy were evaluated by four blinded electroencephalographers. Three single-page EEG samples from each study were assessed for 60 Hz artifact, other possible ECMO-related artifact, and whether total artifact burden compromised confidence in the overall EEG interpretation. Chi-square or Fisher exact tests compared differences based on electrode composition. Electrical artifact was then quantified using 50-70 Hz spectral power. Statistical analyses were performed on the log transformation of these values because the data was highly skewed. To validate this artifact assessment method, t-test comparisons were then performed based on reviewers’ confidence in EEG interpretation. Finally, 50-70 Hz power was measured in an extended consecutive cohort comprised of 14 gold and 43 plastic electrode studies, as well as 4 gold-electrode routine EEGs conducted on members of this cohort after completion of ECMO therapy, then evaluated using analysis of variance and post-hoc Tukey analyses.Results: With plastic electrodes, the EEG reviewers noted less 60 Hz artifact (p<0.001) and ECMO-related artifact (p<0.001), and judged better overall confidence in EEG interpretation (p<0.001). Significantly less 50-70 Hz power (p<0.001) was seen when the majority of reviewers retained confidence in EEG interpretation (12 studies, 7 gold + 5 plastic) than when the majority deemed confidence in EEG interpretation was compromised (7 studies, all gold). In the larger final cohort, there was less 50-70 Hz power in plastic electrode studies compared to gold during ECMO therapy, and less 50-70 Hz power in gold electrode studies outside of the ECMO environment than during ECMO. All comparisons had p<0.001 using both bipolar and referential montages. Plastic electrodes during ECMO had equivocal results compared to gold electrode studies performed outside of ECMO therapy, with less 50-70 Hz power seen in non-ECMO/gold studies than ECMO/plastic using a referential montage (p=0.048), but no significant difference using a bipolar montage (p=0.077).Conclusions: Using conductive plastic electrodes in lieu of conventional gold electrodes diminishes EEG artifacts associated with ECMO therapy in pediatric patients, and improves confidence in EEG interpretation. Prominent neurologic complications are known to affect ECMO patients. Factors such as sedation, neuromuscular blockade, and hemodynamic instability limit clinical and radiographic evaluation to detect these issues. Diminished EEG artifact and improved confidence in EEG interpretation therefore significantly augment the neurologic assessment of ECMO patients.