Abstracts

Rationale: Biofeedback therapy offers a non-invasive and apparently side effect-free behavioral intervention. The use of biofeedback in epilepsy has been reported using different physiological parameters (such as EEG and GSR). GSR reflects dynamic changes in psychophysiological arousal and is an indicator of sympathetic activity. Control of psychophysiological arousal can be achieved through biofeedback and this may be harnessed as a therapy for epilepsy, to mitigate pre-ictal states. Evidence is accumulating for the clinical efficacy of GSR biofeedback; Our recent study demonstrated that biofeedback induced functional connectivity changes (front-limbic network) is linked to seizure reduction. A systematic review and meta-analysis was conducted to understand the evidence for a clinical efficacy of GSR biofeedback from currently available literature. Methods: A literature search was conducted through MEDLINE and Cochrane databases (1950–2018). The search was focused on interventional studies of GSR biofeedback for epilepsy. A meta-analysis was undertaken to extract evidence of clinical efficacy using percentage seizure reduction as an indicator of therapeutic efficacy induced by GSR biofeedback. The study procedures of each article were also investigated in line with the PRISMA statement. Results: Twenty one articles were retrieved for GSR/EDA/Skin conductance biofeedback. Of these, four studies qualified as interventional trials (99 patients with drug-resistant epilepsy in total). The difference in percentage seizure frequency change between Biofeedback and Control was between − 54.4 and − 74.0% with an overall weighted mean difference of − 64.3% (95% CI: − 85.4 to − 43.2%). The response rates (proportion of patients manifesting > 50% reduction in seizure frequency) varied from 45 to 66% across studies, similar to the other non-pharmacological interventions such as Vagus Nerve Stimulation (30-70%) and Ketogenic diet (30-55%). Conclusions: Despite small samples currently available, the meta-analysis demonstrated encouraging evidence regarding the clinical efficacy of GSR biofeedback for the management of seizures in patients with drug resistant epilepsy. The systematic review clarified the important elements to look into in the future studies. These include size of patients population, seizure types, double-blinding, allocation of well-designed control group and robust randomization methods. Funding: No funding