JMIR Publications

ABSTRACT

Background:

Digital health interventions present an opportunity for addressing sexual and reproductive health outcomes among young adults in Francophone West Africa. With the growing presence of social media in this region, studies must also consider how social media use may influence online health interventions.

Objective:

The purpose of this study was to identify social media use patterns in a population of young adults in French-speaking West Africa, to examine the demographic predictors of identified social media use patterns, and to assess how social media behavior might shape the impact of a digital health messaging intervention on Facebook.

Methods:

Young adult (ages 15-24) social media users were recruited from Senegal, Côte d’Ivoire and Burkina Faso and took part in an online study on Facebook. Participants were randomly assigned to four study groups: control condition, peer role model, online influencer, and a mixed peer role model and online influencer strategy. Respondents completed pre-and post- intervention surveys between December 2023-February 2024 reporting on general social media use as well as knowledge, provider awareness and service use related to sexual and reproductive health. This study employed a latent class analysis approach to identify self-reported social media use patterns among study participants (N=262) and maximum likelihood estimation to examine how these patterns impact study outcomes.

Results:

Two types of social media users were identified in our sample: Selective users (i.e., daily use of WhatsApp and Facebook) and Multi-platform users (i.e., daily use of WhatsApp, Facebook, Instagram, YouTube, and TikTok). Additionally, being 20-24 years old, relative to being 15-19 years old, was associated with three times the odds of being a multi-platform user relative to a selective user (OR 3.61, P=.008), while residing in Senegal was associated with a greater likelihood of being a multi-platform user, compared to respondents residing in Côte d’Ivoire (OR 0.27, P=.03) and Burkina Faso (OR 0.23, P=.02). Finally, among selective user types only, when compared to the control intervention, the peer role model approach was associated with a 1.84 unit increase in contraceptive knowledge (P=.02), a 1.81 unit increase in service provider awareness (P<.001), and a 1.49 unit increase in service utilization (P=.04). Both online influencer and mixed approaches did not result in significant outcomes for either user type.

Conclusions:

Online health interventions should use different messaging strategies to target various types of social media users. Targeted interventions must consider both demographic contexts and media exposure patterns to maximize the potential impact of social media messaging.