Statistical Analysis Plan — a 2×2 factorial randomised experiment of advice on opioid use in acute low back pain

Résumé fourni par la source

This registration records the statistical analysis plan for a 2×2 factorial online randomised controlled trial of guideline-concordant advice about opioids for a hypothetical acute low back pain (ANZCTR: ACTRN12626000892314; University of Sydney HREC 2026/HE000484). Two factors are randomised simultaneously: advice framing (opioids are "not recommended" vs "could be considered as a last resort") and delivery mode (a clinician delivered advice versus a matched AI-generated avatar delivered advice). The primary outcome is intention to take opioids, measured immediately after a short advice video. The advice framing comparison is a superiority comparison; the delivery mode comparison is a non-inferiority comparison with a 5-percentage-point margin. Recruitment opened on 7 August 2026 and closed on 11 August 2026. This registration is made after recruitment closed but before any analysis of the primary outcome. Under the two-stage blinding procedure set out in the plan, the statistician will conduct all pre-specified analyses under a dummy allocation generated by re-randomisation; the true allocation will be then restored and the analyses repeated unchanged, with no analysis code written or modified after unblinding. The files registered here are the statistical analysis plan.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.