A Systematic Review of Peer-Supported Interventions for Health Promotion and Disease Prevention
Published in: Preventive Medicine Volume 101 (August 2017), pages 156-170. doi: 10.1016/j.ypmed.2017.06.008
Posted on RAND.org on August 03, 2017
Prior research has examined peer programs with respect to specific peer roles (e.g.; peer support) or specific health/wellness domains (e.g.; exercise/diet), or have aggregated effects across roles and domains. We sought to conduct a systematic review that categorizes and assesses the effects of peer interventions to promote health and wellness by peer role, intervention type, and outcomes. We use evidence mapping to visually catalog and synthesize the existing research. We searched PubMed and WorldCat databases (2005 to 2015) and New York Academy of Medicine Grey Literature Report (1999 to 2016) for English-language randomized control trials. We extracted study design, study participants, type of intervention(s), peer role(s), outcomes assessed and measures used, and effects from 116 randomized controlled trials. Maps were created to provide a visual display of the evidence by intervention type, peer role, outcome type, and significant vs null or negative effects. There are more null than positive effects across peer interventions, with notable exceptions: group-based interventions that use peers as educators or group facilitators commonly improve knowledge, attitudes, beliefs, and perceptions; peer educators also commonly improved social health/connectedness and engagement. Dyadic peer support influenced behavior change and peer counseling shows promising effects on physical health outcomes. Programs seeking to use peers in public health campaigns can use evidence maps to identify interventions that have previously demonstrated beneficial effects. Those seeking to produce health outcomes may benefit from identifying the mechanisms by which they expect their program to produce these effects and associated proximal outcomes for future evaluations. Although we attempted to register our protocol with PROSPERO, we did not meet eligibility criteria because we were past the data collection phase. The full PROSPERO-aligned protocol is available from the authors.