Introduction: Patient-generated health data provide a great opportunity for a more detailed ambulatory monitoring and more personalized treatments in many diseases. In epilepsy, more robust diagnostics applicable to the ambulatory setting are needed. Diagnosis and treatment decisions in current clinical practice primarily rely on patient self-reports, which are often inaccurate. Recent work using wearable devices has focused on methods to detect and forecast epileptic seizures. Whether wearable device signals may also contain information about the effect of antiseizure medications (ASMs), which may ultimately help to better monitor their efficacy, has not been evaluated yet. In our study we systematically investigated the effect of ASMs on different data modalities (electrodermal activity, EDA, heart rate, HR, heart rate variability, HRV) simultaneously recorded by a wearable device over several days in the video EEG monitoring unit. Patients & Methods: We included multi-day data from 50 epilepsy patients, and patients were equipped with one or two biosensor wristbands (Empatica E4) on each side of the body. We systematically analysed the effects of ASMs on the different data modalities during seizure-free periods over multiple days. To obtain a continuous estimate of ASM load we modelled each ASM by the first-order kinetics and then obtained an estimate of total ASM load in the case of multiple drugs using their individual defined daily doses. Artifactual data were removed from further analysis using reported signal quality indices. We focused our analyses on EDA (raw signal and power in the low frequency band (0-0.35 Hz)), HR and HRV (variance of inter-beat intervals, VAR, root mean squared of successive inter-beat intervals, RMSSD, power in the low frequency (LF, 0.04-0.15 Hz) and high frequency bands (HF, 0.15-0.4 Hz)). Results: All modalities (HR, EDA and HRV) exhibited characteristic diurnal patterns. Only HRV diurnal patterns (VAR, RMSSD, LF, HF) were consistently decreased in the setting of high ASM load in comparison to low ASM load. Subgroup analyses indicated that the reduction of HRV by ASMs was not limited to sodium channel blocking ASMs, supporting the notion that changes are not based on cardiac effects but on ANS effects. Conclusion: Our study emphasizes the ability to use easy-worn wearables to monitor various physiological signals related to ANSactivity in parallel. Signals exhibited diurnal variations, and HRV, but not HR, EDA, and EDA power, was reduced in the setting of ASM application. Future work using longer, multidien data may investigate these metrics on longer, multidien cycles, their relationship to ASM action, and, consequently, their utility for detecting seizures, assessing seizure risk, or informing treatment interventions.
Read full abstract
Nov 13, 2022
Alexander Pokutsa + 1
Open Access
