Baseline findings of a cluster-randomized trial of SKILLZ among Grade 9 South African learners

DETAILED SUMMARY

Importance

• HIV prevalence among 15–24 year-olds in South Africa is among the highest in the world, with young females facing nearly twice the risk of young males. Key drivers include age-disparate sex, multiple partnerships, harmful alcohol use, and gender-based violence; while rape and intimate-partner violence perpetration have been documented among adult South African men, little research had assessed prevalence of these behaviors among adolescents.

Objective

• To describe baseline knowledge, attitudes, and behaviors related to HIV risk and gender-based violence among Grade 9 learners in Cape Town and Port Elizabeth, South Africa, prior to a two-year cluster-randomized trial evaluating the effectiveness of the Generation Skillz sport-based HIV prevention intervention.

Study details

• Study design: Randomized controlled trial (baseline data)
• Data collection: Baseline surveys only (146-item self-administered questionnaire via Open Data Kit software on touchscreen mobile phones, available in English or Xhosa with audio assistance)
• Participants: Mixed-gender Grade 9 learners (median age 15 years)
• Implementation setting: 46 public secondary schools in peri-urban townships of Cape Town and Port Elizabeth, South Africa (peri-urban)
• Study period: Baseline surveys conducted March to September 2012, as part of a two-year trial

Intervention

• Other (Generation Skillz): A sport-based HIV prevention program delivered by trained Grassroot Soccer Coaches, using soccer themes, activities, and metaphors to spark conversations about HIV, sex, and violence among Grade 9-10 learners; compared with standard Life-Orientation classes delivered by teachers.

Results

• Sample: 4,485 Grade 9 learners assessed at baseline (2,173 intervention, 2,312 control).
• More control than intervention participants reported HIV stigma (27.4% vs. 23.8%, p=0.007), though this and other baseline differences lost statistical significance after adjusting for age and school-level clustering.
• More control than intervention females reported ever having had sex (45.1% vs. 40.9%, p=0.038), while more control than intervention males reported depressive symptoms (31.5% vs. 25.0%, p=0.001), an age-disparate partner in the last year (15.7% vs. 12.0%, p=0.06), perpetration of intimate-partner violence (37.6% vs. 33.4%, p=0.051), and perpetration of rape (29.1% vs. 23.7%, p=0.029).
• No significant baseline differences between groups in socio-economic status, HIV knowledge, self-efficacy, or number of sexual partners after adjustment.
• The control group was slightly older than the intervention group (mean age 15.8 vs. 15.5 years, p=0.071), and more control than intervention participants reported not living with both parents (67.8% vs. 64.1%, p=0.009); the vast majority of participants (96.3%) were Xhosa speakers.

Conclusions and Implications

• A small baseline age imbalance between study arms appears to explain most observed baseline differences, and outcome analyses for the main trial were adjusted for age.
• Distinct gender differences in knowledge, attitudes, and behavior were observed among Grade 9 learners, with males reporting higher rates of harmful attitudes and behaviors, including high numbers of sexual partners and high prevalence of intimate-partner violence, rape perpetration, and harmful alcohol use, underscoring these as important areas for intervention; Grassroot Soccer incorporated these findings into its Generation Skillz curriculum for Grade 9 and 10 learners.