Overcoming health misinformation in marginalized groups: a systematic review.

BACKGROUND: Marginalized populations experience persistent health disparities that may be exacerbated by exposure to health misinformation. This systematic review aimed to identify interventions designed to address health misinformation in marginalized populations and characterize their effectiveness for changing health beliefs, intentions or behaviors. METHODS: Medline, Embase, PsycINFO and Web of Science Core Collection were searched from inception to January 2025, with additional hand-searching of journals and citations. All studies were assessed for eligibility by two reviewers. Quantitative, qualitative, or mixed-method studies were eligible if they aimed to address health misinformation in a marginalized population and reported an outcome relating to health beliefs, intentions or behaviors. Studies that focused only on program design or engagement, and those lacking outcome data, were excluded. The Effective Public Health Practice Project Quality Assessment Tool and the National Institute for Health and Care Excellence Quality Appraisal Checklist were used to assess study quality and risk of bias. Findings were described through narrative synthesis. RESULTS: Thirty-two studies met eligibility criteria; seven were assessed as moderate quality, 25 as weak. Studies were classified based on intervention components: message-focused (n = 17); capacity- or trust-building (n = 6); or both (n = 9). Studies of message-focused interventions provided moderate-to-low quality evidence for use of pre/debunking to change misinformation beliefs and to increase intentions to comply with recommended health behaviors. Sixteen studies included community engagement in design, and 14 were delivered by community members; early community participation in ideation, design or delivery was reported to improve engagement and trust in health information, contributing to intervention efficacy. However, most studies had limitations in randomization, blinding, and outcome measurement. CONCLUSION: Evidence on interventions to overcome health misinformation in marginalized populations remains limited and generally weak. Debunking and prebunking approaches show promise in changing health beliefs and intentions in some marginalized groups. Representatives from marginalized populations should be routinely involved in design, delivery and evaluation of such interventions. REGISTRATION: The protocol and search strategy were registered and published in the PROSPERO International Prospective Register of Systematic Reviews (CRD42024550534).