How does epilepsy self‐management reduce seizure frequency? A causal mediational analysis of a program targeting rural and underserved people with epilepsy and prior negative health events

OBJECTIVE: A recent clinical trial demonstrates that an epilepsy self-management program, self-management for people with epilepsy and a history of negative health events (SMART), improves seizure reduction. Understanding which aspects of SMART mediated this improvement allows refinement of epilepsy self-management programs for effective inclusion into epilepsy care. METHODS: A total of 160 adults with epilepsy were randomized to receive the SMART epilepsy self-management program or to be placed on a waitlist. Assessments were conducted at baseline, 13 weeks, and 6 months. The outcome was a binary variable measuring ≥20% seizure reduction at 6 months. Potential mediators were secondary outcomes that significantly improved during the SMART trial. Improvements in these potential mediators were evaluated over the different intervals to distinguish early (baseline to Week 13) and late (Week 13 to Week 26) mechanistic pathways. A regression-based causal mediation framework was performed to estimate the indirect effect of SMART mediated through each candidate mediator. RESULTS: Enrollment in SMART significantly increases the odds of achieving ≥20% seizure reduction at 6 months (odds ratio [OR] = 2.15, p < 0.001). SMART is significantly associated with improvements in depression, self-efficacy, seizure worry, social functioning, and self-management behaviors. Portions of the effect of SMART are significantly mediated by improvement in depression symptoms at 6 months (25%, 95% confidence interval [CI]: 2-48, of total effect) and in early improvements in self-efficacy (30%, 95% CI: 11-49, of total effect) and seizure worry (46%, 95% CI: 24-69, of total effect) at 13 weeks. Gains in social functioning and self-management behaviors are not meaningful mediators of SMART effects. SIGNIFICANCE: In addition to the direct of effect of SMART on seizure reduction, there are significant partial mediators of seizure reduction accounting for substantial indirect effects. Improvements in depressive symptoms and seizure worry along with gains in self-efficacy are likely mechanisms through which SMART works to control seizures, demonstrating the benefits of a holistic self-management approach to epilepsy care.