A mixed-methods, exploratory, quasi-experimental evaluation of a radio drama intervention to prevent age-disparate transactional sex in Tanzania

INTRODUCTION: Age-disparate transactional sex (ADTS) is associated with HIV, unintended pregnancy, school dropout and violence, yet few interventions have successfully prevented it, and none have set ADTS prevention as their primary outcome. This exploratory evaluation aimed to assess indications of change after exposure to the Learning Initiative on Norms, Exploitation and Abuse (LINEA) intervention, a mass-media, gender-transformative social norms intervention aimed at preventing ADTS in Tanzania. METHODS: In a condensed implementation 331 participants were instructed to listen to the LINEA radio drama over seven weeks, and 60 were randomly allocated to household discussion sessions about content. In-depth interviews (n = 81) from girls aged 12–16 years, and women and men caregivers were collected at baseline (September 2021), midline (November) and endline (December 2021). Surveys were conducted (n = 120) at baseline and endline using the Norms and Attitudes on ADTS Scale (NAATSS) and the Gender Roles and Male Provision Expectations (GRMPE) scale. Interviews were thematically analyzed using a framework approach. Age-stratified linear regression models adjusted for baseline scores were used to measure association between the intervention and endline scale scores. RESULTS: Longitudinal data were available from 59 qualitative (73%) and 95 quantitative participants (79%). Qualitative evidence revealed the drama facilitated family conversations about adolescent challenges, allowing caregivers to advise daughters. Some girls gained confidence to refuse men's gifts, learning that accepting them could necessitate sexual reciprocation. Some caregivers felt increased responsibility for supporting girls in the community to avoid ADTS. Blame for ADTS shifted for some from girls to men, suggesting increased understanding of inequitable power dynamics and reductions in victim blaming. Marginal quantitative evidence revealed that highly exposed girls had improved gender equitable beliefs on the GRMPE (β = −6.26; 95% CI: −12.94, 0.42). Moderately exposed men had increased gender inequitable norms on the NAATSS subscale (β = 0.42 95% CI: 0.05, 0.79), but there was no effect in highly exposed men. CONCLUSIONS: Given the small sample results should be interpreted cautiously. Our initial findings indicate high engagement with the LINEA intervention shows promise in shifting knowledge, behaviors, and attitudes, beliefs and social norms driving ADTS in Shinyanga, Tanzania, supporting a robust impact evaluation.