Do health education initiatives assist socioeconomically disadvantaged populations? A systematic review and meta-analyses

Background: Health education interventions are considered critical for the prevention and management of conditions of public health concern. Although the burden of these conditions is often greatest in socio-economically disadvantaged populations, the effectiveness of interventions that target these groups is unknown. We aimed to identify and synthesize evidence of the effectiveness of health-related educational interventions in adult disadvantaged populations. Methods: We pre-registered the study on Open Science Framework https://osf.io/ek5yg/. We searched Medline, Embase, Emcare, and the Cochrane Register from inception to 5/04/2022 to identify studies evaluating the effectiveness of health-related educational interventions delivered to adults in socio-economically disadvantaged populations. Our primary outcome was health related behaviour and our secondary outcome was a relevant biomarker. Two reviewers screened studies, extracted data and evaluated risk of bias. Our synthesis strategy involved random-effects meta-analyses and vote-counting. Results: We identified 8618 unique records, 96 met our criteria for inclusion – involving more than 57,000 participants from 22 countries. All studies had high or unclear risk of bias. For our primary outcome of behaviour, meta-analyses found a standardised mean effect of education on physical activity of 0.05 (95% confidence interval (CI) = -0.09–0.19), (5 studies, n = 1330) and on cancer screening of 0.29 (95% CI = 0.05–0.52), (5 studies, n = 2388). Considerable statistical heterogeneity was present. Sixty-seven of 81 studies with behavioural outcomes had point estimates favouring the intervention (83% (95% CI = 73%-90%), p < 0.001); 21 of 28 studies with biomarker outcomes showed benefit (75% (95%CI = 56%-88%), p = 0.002). When effectiveness was determined based on conclusions in the included studies, 47% of interventions were effective on behavioural outcomes, and 27% on biomarkers. Conclusions: Evidence does not demonstrate consistent, positive impacts of educational interventions on health behaviours or biomarkers in socio-economically disadvantaged populations. Continued investment in targeted approaches, coinciding with development of greater understanding of factors determining successful implementation and evaluation, are important to reduce inequalities in health.