Cluster randomized controlled trial of Generation SKILLZ among adolescents in South Africa

DETAILED SUMMARY

Importance

• Despite progress in increasing uptake of HIV testing and treatment, preventing new HIV infections remains a challenge for South Africa, which has more than 6.4 million people living with HIV and approximately 469,000 new infections per year, roughly 30% of which occur among 15- to 24-year-olds.
• Interest in using sport to promote health has grown internationally, and SMS messaging is increasingly used in health promotion, but rigorous randomized evidence supporting the effectiveness of sport-based HIV prevention, and of complementary SMS campaigns, had remained limited.

Objective

• To assess the effectiveness of Generation SKILLZ, a sport-based HIV prevention intervention, with and without an accompanying SMS campaign, in reducing reported age-disparate sex, multiple sexual partnerships, and rape perpetration, and in improving HIV-related knowledge and attitudes, among adolescents in South African informal settlement schools.

Study details

• Study design: Cluster randomized controlled trial (2×2 factorial design)
• Data collection tools: Self-administered questionnaires completed on touchscreen mobile phones at baseline and midline (8–11 months post-intervention)
• Participants: 4,485 Grade 9 learners at baseline (mean age 15.66 years, SD 1.39; median age 15 years); 3,442 participants retained at midline (median age 16 years)
• Implementation setting: 46 secondary schools in informal settlements in Cape Town and Port Elizabeth, South Africa (urban)
• Study period: 2012–2014 (baseline in 2012; midline follow-up in 2013; endline planned for 2014)

Intervention

• Generation SKILLZ: Grassroot Soccer’s eleven-session sport-based HIV prevention curriculum. Schools were randomized to receive the intervention or standard life-orientation classes only, and intervention schools were further randomized to receive, or not receive, a 12-month biweekly SMS campaign reinforcing program messaging.

Results

• 4,485 learners completed the baseline questionnaire across 46 schools; 3,442 (70.5%) were retained at midline, with 1,309 (29.5%) lost to follow-up. Riskier, older, and Cape Town participants had higher odds of attrition.
• There was very strong evidence of a positive effect of the intervention on HIV-related knowledge among both males and females (β=0.39, 95% CI: 0.25–0.53), on reported ever testing for HIV among males (OR=1.38, 95% CI: 1.11–1.72), and on reported HIV testing in the last 12 months among males (OR=1.47, 95% CI: 1.13–1.90).
• There was strong evidence of a negative effect on reported multiple partnerships in the last six months among males (OR=1.34, 95% CI: 1.08–1.66) and on reported ever perpetrating physical intimate-partner violence among males (OR=1.27, 95% CI: 1.00–1.60).
• Within the intervention arm, there was strong evidence that the SMS campaign reduced reported multiple partnerships in the last six months among males and females combined (OR=0.75, 95% CI: 0.58–0.96); however, comparing SMS schools directly to control schools, no evidence of a difference was found on any of the three primary behavioral outcomes (age-disparate sex, multiple partnerships, or rape perpetration).

Conclusions and implications

• Midline results suggested the Generation SKILLZ intervention was not effective in achieving its primary behavioral objectives in the first year, despite increases in HIV-related knowledge and self-reported HIV testing. Findings helped inform Grassroot Soccer’s decision to de-implement Generation SKILLZ.
• Findings provided further evidence that targeted SMS messaging may be an effective tool for sexual health promotion; further qualitative research was planned to investigate why the intervention may have led to an increase in certain risk behaviors, including hypothesized links to increased social media use and coach-participant dynamics.