Forty-one level 1 studies were included.
School-based interventions were assessed in 10 randomised controlled trials (RCTs; n=9,393), 1 post-test only with a comparison group (n=838), 7 panel studies with comparison groups (n=12,594) and 4 repeat cross-sectional studies with comparison groups (n=7,640).
Mass media interventions were assessed in 5 repeat cross-sectional studies with comparison groups (n=4,850 at baseline and n=6,238 at follow-up) and 1 repeat cross-sectional study without a comparison group (n=947 at baseline and n=1,575 at follow-up).
Community-based interventions were assessed in 1 RCT (n=720), 3 repeat cross-sectional studies with comparison groups (n=9,478) and 1 post-test only with a comparison group (n=1,693).
Workplace interventions were assessed in 3 panel studies with comparison groups (n=1,838) and 1 post-test only without a comparison group (n=2,417).
Health facility-based interventions were assessed in 3 repeat cross-sectional studies with comparison groups (2,083 people in the only study reporting the sample size) and 1 panel study with a comparison group (sample size not reported).
Unless specifically stated, all studies mentioned below are level 1 studies.
School-based programmes (22 level 1 studies).
All programmes: programmes varied widely in terms of the syllabus, content and format of delivery. The proportion of all studies reporting a significant positive impact was: 17 of 21 studies assessing knowledge and attitudes; 4 of 11 studies assessing delayed sex; 3 of 6 studies assessing the number of partners; 6 of 10 studies assessing contraceptive use; and 1 of 3 studies assessing service use.
Mass media programmes (6 quasi-experimental studies): 5 of 6 studies assessing knowledge and attitudes found a positive effect on knowledge and attitudes. Three of 4 studies that included social marketing found a positive effect on knowledge and attitudes. The studies found mixed results for behaviour outcomes.
Community-based programmes (5 studies including 1 RCT, 1 post-test study with comparison group and 3 repeated cross-sectional studies): the studies found that community-based programmes improved knowledge about sexually transmitted infections, knowledge and attitudes, educational level, employment, service use and delayed sex.
Workplace programmes (4 studies): all 4 studies found a positive effect on knowledge and attitudes. The 2 studies assessing contraceptive use found increased use with the programme.
Health-facility based programmes. Youth-friendly services (3 studies): the studies found that the programmes improved knowledge and increased service use and contraceptive use.