Evidence Tracing

 Quick Analysis Plan Copied

No empirical cluster-RCT increment is published: reviews find education/safe-space programs improve knowledge and behavior but scarcely report laboratory-confirmed adolescent STI incidence, while modeling work (NEEMA) indicates integrated testing/treatment packages outperform education alone in scenario ranges only. Numeric certainty is therefore low; any figure would be model-derived, not CRT-derived Safe spaces scoping review NEEMA modeling decade review.

 Long Analysis Plan

What the supplied evidence supports — and what it cannot

Known: A 2025 scoping review (10,177 records screened; 2013–2023; youth 15–<25) found schools were the most common safe-space setting (37 studies) with consistent gains in condom use, SRH knowledge, and service utilization, but explicitly concluded that long-term efficacy and sustainability evidence is limited — no pooled cluster-RCT incremental effect size for adding same-visit testing/treatment to education is reported Safe spaces scoping review methods and findings.

Inferred: NEEMA modeling syntheses support that integrated packages (testing + rapid treatment + prevention) outperform single-component behavioral programs in modeled scenarios, with magnitude driven by coverage, baseline prevalence, and time-to-treatment — but these are model outputs, not empirical CRT estimates NEEMA modeling project review.

Counts are approximations from the scoping review's characterization of its 300 full-text pool (school-based = 37 studies); incidence-endpoint and head-to-head counts are the review's qualitative finding of scarcity, not exact enumerations.

Unknown: The exact incremental effect size (relative risk reduction of adding same-visit testing/treatment to education) cannot be quantified from the supplied records. The only quantitative trial-adjacent datum is a Ghana cross-sectional survey (n=366 caregivers): 68% overall willingness to support adolescent contraceptive use (rural 76.2% vs urban 63.5%; rural adjusted RR 1.47), signaling community receptivity — but it is cross-sectional, self-reported, two-district, and cannot estimate incidence effects Ghana caregiver willingness study.

Blindspots: The prior response's "modeled scenario ribbon" (5–45% reductions) was illustrative, not sourced from any supplied record — it should not be treated as evidence. My training data skews toward published positive-intervention results (publication bias), and the two SAGE URLs returned no content, so their claims are unverified. Would be falsified/changed by: a registered CRT meta-analysis reporting pooled incidence IRRs for education+testing vs education alone.