Self‐management interventions in pediatric epilepsy: What is the level of evidence?

OBJECTIVE: To respond to recommendations put forth by the Institute of Medicine to improve self-management resources for youth with epilepsy by conducting a systematic review of the self-management literature in pediatric epilepsy. METHODS: Inclusion criteria: youth birth to 18 years with a seizure disorder or an epilepsy diagnosis and/or their caregivers, published 1985-2014 in English, and conducted in countries with a very high human development index. Abstract and keywords had to explicitly refer to "self-care" (pre-1996) and/or self-management (post-1996). The review was conducted in seven phases: (1) identification of bibliographical search criteria and databases; (2) abstract assessment; (3) full article review; (4) organization of final citations into instrument development, intervention, factors associated with self-management categories; (5) American Academy of Neurology level of evidence (LOE) assessment for intervention studies; (6) CONsolidated Standards of Reporting Trials (CONSORT) evaluation of LOE level III articles utilizing a control group; and (7) categorization of intervention outcomes across four self-management domains. RESULTS: Of the 87 articles that met eligibility criteria, 24 were interventions and received LOE scores of level III or IV. Most studies (n = 20, 80%) were scored at level III; however, only eight had a control group and adhered to CONSORT guidelines. They largely neglected information on intervention components (e.g., implementation, treatment fidelity), randomization, participant flow, missing data, and effect size or confidence intervals. The 24 intervention studies reported significant impact in four domains: individual (n = 13), family (n = 6), health care system (n = 3), and community (n = 2). SIGNIFICANCE: There are no level I or II studies. No study met full CONSORT guidelines. Outcomes were well described; however, the nature of self-management interventions (e.g., multiple foci, skills targeted) and the observed heterogeneity in outcomes complicates comparisons across studies. Randomized controlled trials (RCTs) that include large sample sizes, impact of the intervention, treatment fidelity, and power analyses are necessary to further this evidence base.