Social Media-Delivered Sexual Health Intervention A Cluster Randomized Controlled Trial

Background: Youth are using social media regularly and represent a group facing substantial risk for sexually transmitted infection (STI). Although there is evidence that the Internet can be used effectively in supporting healthy sexual behavior, this has not yet extended to social networking sites. Purpose: To determine whether STI prevention messages delivered via Facebook are efficacious in preventing increases in sexual risk behavior at 2 and 6 months. Design: Cluster RCT, October 2010-May 2011. Setting/participants: Individuals (seeds) recruited in multiple settings (online, via newspaper ads and face-to-face) were asked to recruit three friends, who in turn recruited additional friends, extending three waves from the seed. Seeds and waves of friends were considered networks and exposed to either the intervention or control condition. Intervention: Exposure to Just/Us, a Facebook page developed with youth input, or to control content on 18-24 News, a Facebook page with current events for 2 months. Main outcome measures: Condom use at last sex and proportion of sex acts protected by condoms. Repeated measures of nested data were used to model main effects of exposure to Just/Us and time by treatment interaction. Results: A total of 1578 participants enrolled, with 14% Latino and 35% African-American; 75% of participants completed at least one study follow-up. Time by treatment effects were observed at 2 months for condom use (intervention 6896 vs control 56%, p=0.04) and proportion of sex acts protected by condoms (intervention 63% vs control 57%, p=0.03) where intervention participation reduced the tendency for condom use to decrease over time. No effects were seen at 6 months. Conclusions: Social networking sites may be venues for efficacious health education interventions. More work is needed to understand what elements of social media are compelling, how network membership influences effects, and whether linking social media to clinical and social services can be beneficial.