Abstracts

Rationale: Subcutaneous EEG (sqEEG) is a new modality with high clinical potential for remote, continuous monitoring in epilepsy, for use cases from seizure detection to seizure forecasting. Good acceptability and adherence are essential requirements for its effectiveness. We aimed to assess patterns of adherence to subcutaneous EEG monitoring, as a proxy for device acceptability.

Methods: Eight patients with refractory epilepsy (50% female, age range 33-63) were enrolled in an ongoing prospective observational study and were implanted with a two-channel unilateral subcutaneous EEG system (UNEEG SubQ^TM), under local anesthesia, over an area of presumed ictal EEG changes (six left implantations). One subject was reimplanted after a device malfunction. Recording usually started one to three weeks after implantation. An external recorder connects to the implant housing via an inductive link, powering the implant and storing data. Subjects were asked to use the system as much as possible during their everyday life, and regular monthly follow-up visits were undertaken to review compliance. Device adherence (percentage of recording time) was investigated at different timescales. A time-of-day effect was assessed via examining individual circadian plots. A weekday effect was investigated through individual-level Kruskal-wallis tests. Long-term trends in adherence (i.e. attrition rate) were analyzed via group-level mixed effects models, with random intercept and slope, also assessed separately for presumed diurnal (10am-4pm) and nocturnal (11pm-5am) periods.

Results: Pooling over subjects, a total of 901 days of sqEEG have been recorded, out of a possible 1100. Recording length ranged from 35 to 231 days. Median adherence was 20.3 hours/day, ranging from 13.2 to 23.6 hours/day. Circadian adherence plots showed individual patterns of adherence, from strictly fixed hours off recording in some subjects, to a preference to record either during the day or night in others. A weekday effect was statistically significant in two subjects, showing worse adherence during the weekend in one. At a larger timescale, group-level linear mixed-effects models showed no significant adherence change over days of study, including during the day or night-time periods.

Conclusions: Contrary to concerns over a high attrition rate in wearable devices, our findings suggest that sqEEG devices are highly acceptable for at least a proportion of refractory epilepsy patients. On the other hand, individual patterns of adherence may influence the monitoring yield of these systems. Overall, these findings reinforce the suitability and acceptability of subcutaneous EEG for chronic implantation and monitoring.

Funding: Please list any funding that was received in support of this abstract.: Epilepsy Foundation of America "My Seizure Gauge" Initiative.