Do School-Based Programs Prevent HIV and Other Sexually Transmitted Infections in Adolescents? A Systematic Review and Meta-analysis

Ali Mirzazadeh, M. Antonia Biggs, Amanda Viitanen, Hacsi Horvath, Li Yan Wang, Richard Dunville, Lisa C. Barrios, James G. Kahn, Elliot Marseille

We systematically reviewed the literature to assess the effectiveness of school-based programs to prevent HIV and other sexually transmitted infections (STI) among adolescents in the USA. We searched six databases including PubMed for studies published through May 2017. Eligible studies included youth ages 10-19 years and assessed any school-based programs in the USA that reported changes in HIV/STI incidence or testing. We used Cochrane tool to assess the risk of bias and GRADE to determine the evidence quality for each outcome. Three RCTs and six non-RCTs, describing seven interventions, met study inclusion criteria. No study reported changes in HIV incidence or prevalence. One comprehensive intervention, assessed in a non-RCT and delivered to pre-teens, reduced STI incidence into adulthood (RR 0.36, 95% CI 0.23-0.56). A non-RCT examining chlamydia and gonorrhea incidence before and after a condom availability program found a significant effect at the city level among young men 3 years later (RR 0.43, 95% CI 0.23-0.80). The remaining four interventions found no effect. The effect on STI prevalence was also not significant (pooled RR 0.83 from two non-RCTs, RR 0.70 from one RCT). Only one non-RCT showed an increase in HIV testing (RR 3.19, 95% CI 1.24-8.24). The quality of evidence for all outcomes was very low. Studies, including the RCTs, were of low methodological quality and had mixed findings, thus offering no persuasive evidence for the effectiveness of school-based programs. The most effective intervention spanned 6 years, was a social development-based intervention with multiple components, rather than a sex education program, and started in first grade.