School-based interventions that aim to modify sexual health knowledge, attitudes and behaviors have mixed and often unsustained effects on adolescent sexual health outcomes. However, observational evidence suggests that broader school-related factors, such as school climate and academic attainment, can influence outcomes. METHODS: Nine databases were searched in July 2017 for randomized and quasi-experimental evaluations of interventions addressing school-level environment or student-level educational assets, to examine whether such interventions can promote young people's sexual health. Searches were limited to studies published since 1990 but were not restricted by language. Studies were assessed for risk of bias and synthesized narratively and meta-analytically. RESULTS: Searches yielded 11 evaluations, published from 1999 to 2016, of interventions related to school-level environment or student-level educational assets. Because of inconsistent reporting, the risk of bias was not clear for most studies, and meta-analysis was possible for only one outcome. The meta-analysis of three randomized trials provided some evidence that school-environment interventions may delay sexual debut (pooled odds ratio, 0.5). Narrative synthesis of the remaining outcomes found mixed results, but suggests that interventions addressing school-level environment may delay sexual debut and that those addressing student-level educational assets may reduce risk of pregnancy and STDs.

CONTEXT School‐based interventions that aim to modify sexual health knowledge, attitudes and behaviors have mixed and often unsustained effects on adolescent sexual health outcomes. However, observational evidence suggests that broader school‐related factors, such as school climate and academic attainment, can influence outcomes. METHODS Nine databases were searched in July 2017 for randomized and quasi‐experimental evaluations of interventions addressing school‐level environment or student‐level educational assets, to examine whether such interventions can promote young people's sexual health. Searches were limited to studies published since 1990 but were not restricted by language. Studies were assessed for risk of bias and synthesized narratively and meta‐analytically. RESULTS Searches yielded 11 evaluations, published from 1999 to 2016, of interventions related to school‐level environment or student‐level educational assets. Because of inconsistent reporting, the risk of bias was not clear for most studies, and meta‐analysis was possible for only one outcome. The meta‐analysis of three randomized trials provided some evidence that school‐environment interventions may delay sexual debut (pooled odds ratio, 0.5). Narrative synthesis of the remaining outcomes found mixed results, but suggests that interventions addressing school‐level environment may delay sexual debut and that those addressing student‐level educational assets may reduce risk of pregnancy and STDs. CONCLUSIONS Additional and more rigorous evidence is needed to assess the probability that interventions addressing school‐related factors are effective and to provide better understanding of the mechanisms by which they may work to improve adolescent sexual health.