Mental health-related stigma is widespread and has major adverse effects on the lives of people with mental health problems. Its two major components are discrimination (being treated unfairly) and prejudice (stigmatising attitudes). Anti-stigma initiatives often include mass media interventions, and such interventions can be expensive. It is important to know if mass media interventions are effective. To assess the effects of mass media interventions on reducing stigma (discrimination and prejudice) related to mental ill health compared to inactive controls, and to make comparisons of effectiveness based on the nature of the intervention (e.g. number of mass media components), the content of the intervention (e.g. type of primary message), and the type of media (e.g. print, internet). We searched eleven databases: the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 7, 2011); MEDLINE (OvidSP),1966 to 15 August 2011; EMBASE (OvidSP),1947 to 15 August 2011; PsycINFO (OvidSP), 1806 to 15 August 2011; CINAHL (EBSCOhost) 1981 to 16 August 2011; ERIC (CSA), 1966 to 16 August 2011; Social Science Citation Index (ISI), 1956 to 16 August 2011; OpenSIGLE (http://www.opengrey.eu/), 1980 to 18 August 2012; Worldcat Dissertations and Theses (OCLC), 1978 to 18 August 2011; metaRegister of Controlled Trials (http://www.controlled-trials.com/mrct/mrct_about.asp), 1973 to 18 August 2011; and Ichushi (OCLC), 1903 to 11 November 2011. We checked references from articles and reviews, and citations from included studies. We also searched conference abstracts and websites, and contacted researchers. Randomised controlled trials (RCTs), cluster RCTs or interrupted time series studies of mass media interventions compared to inactive controls in members of the general public or any of its constituent groups (excluding studies in which all participants were people with mental health problems), with mental health as a subject of the intervention and discrimination or prejudice outcome measures. Two authors independently extracted data and assessed the risk of bias of included studies. We contacted study authors for missing information. Information about adverse effects was collected from study reports. Primary outcomes were discrimination and prejudice, and secondary outcomes were knowledge, cost, reach, recall, and awareness of interventions, duration/sustainability of media effects, audience reactions to media content, and unforeseen adverse effects. We calculated standardised mean differences and odds ratios. We conducted a primarily narrative synthesis due to the heterogeneity of included studies. Subgroup analyses were undertaken to examine the effects of the nature, content and type of mass media intervention. We included 22 studies involving 4490 participants. All were randomised trials (3 were cluster RCTs), and 19 of the 22 studies had analysable outcome data. Seventeen of the studies had student populations. Most of the studies were at unclear or high risk of bias for all forms of bias except detection bias.Findings from the five trials with discrimination outcomes (n = 1196) were mixed, with effects showing a reduction, increase or consistent with no evidence of effect. The median standardised mean difference (SMD) for the three trials (n = 394) with continuous outcomes was -0.25, with SMDs ranging from -0.85 (95% confidence interval (CI) -1.39 to -0.31) to -0.17 (95% CI -0.53 to 0.20). Odds ratios (OR) for the two studies (n = 802) with dichotomous discrimination outcomes showed no evidence of effect: results were 1.30 (95% CI 0.53 to 3.19) and 1.19 (95% CI 0.85 to 1.65).The 19 trials (n = 3176) with prejudice outcomes had median SMDs favouring the intervention, at the three following time periods: -0.38 (immediate), -0.38 (1 week to 2 months) and -0.49 (6 to 9 months). SMDs for prejudice outcomes across all studies ranged from -2.94 (95% CI -3.52 to -2.37) to 2.40 (95% CI 0.62 to 4.18). The median SMDs indicate that mass media interventions may have a small to medium effect in decreasing prejudice, and are equivalent to reducing the level of prejudice from that associated with schizophrenia to that associated with major depression.The studies were very heterogeneous, statistically, in their populations, interventions and outcomes, and only two meta-analyses within two subgroups were warranted. Data on secondary outcomes were sparse. Cost data were provided on request for three studies (n = 416), were highly variable, and did not address cost-effectiveness. Two studies (n = 455) contained statements about adverse effects and neither reported finding any. Mass media interventions may reduce prejudice, but there is insufficient evidence to determine their effects on discrimination. Very little is known about costs, adverse effects or other outcomes. Our review found few studies in middle- and low-income countries, or with employers or health professionals as the target group, and none targeted at children or adolescents. The findings are limited by the quality of the evidence, which was low for the primary outcomes for discrimination and prejudice, low for adverse effects and very low for costs. More research is required to establish the effects of mass media interventions on discrimination, to better understand which types of mass media intervention work best, to provide evidence about cost-effectiveness, and to fill evidence gaps about types of mass media not covered in this review. Such research should use robust methods, report data more consistently with reporting guidelines and be less reliant on student populations.
