Commentary

In this review of adolescent HIV risk reduction interventions, Johnson et al used a comprehensive search strategy, explicit selection criteria, and 2 independent raters to extract study data. They obtained data directly from the researchers for about half of the 22 studies lacking sufficient information to calculate effect size estimates.

The authors used a fixed effects model for meta-analysing the data, despite significant heterogeneity in the results of studies examining condom use. They did further analyses to investigate the effect of methodological and participant characteristics on the variability of study results. Study variability was explained by several factors: condom use increased most among non-institutionalised adolescents who were given condoms and received active training in the skills needed to use and negotiate condom use with a partner. This analysis provides information of great relevance to health policy.

Public health nurses should share the findings of this systematic review to dispel the myth that HIV sexual risk reduction interventions for adolescents increase the frequency of sexual behaviour, the number of sexual partners, or the onset of sexual debut. Based on these findings, HIV prevention programmes for youth in the community should incorporate both active instruction in condom use and provision of condoms. Evaluations should focus on behavioural outcomes. Concurrently, investigators need to explore social determinants of health, including environmental structures and social processes within families, schools, and neighbourhoods[1] to identify additional factors that might promote or inhibit risk reduction behaviours in youth. Findings from such research may increase our understanding of the complex social environment that influences the adoption of risk reduction behaviours.