{
  "study_id": "NCT05047120",
  "agent": "framework",
  "run": 4,
  "p_primary_hypothesis_supported": 0.72,
  "effect_direction": "favours intervention",
  "effect_size_90pct_interval": {"low": 0.2, "high": 2.0, "metric": "Between-arm difference in 0-10 NRS pain change (positive = more pain reduction with HYP-CT)"},
  "moderator_prediction": "Higher-baseline-pain patients show larger within-session drops (more room, more salient render); neuropathic vs nociceptive difference small, perhaps slightly larger for nociceptive/musculoskeletal pain less contradicted by constant neural input.",
  "claim_ids_used": ["PM-0023", "PM-0369", "PM-1286", "PM-0086", "PM-0132", "PM-0232", "PM-0128", "PM-0333", "PM-0137", "PM-0276", "PM-0104"],
  "rationale": "Primary outcome is immediate pre-post pain change around live sessions, a render-layer outcome measured right after induction, where the framework predicts hypnosis acts most strongly: critical faculty suspended, suggestion written with reduced resistance, delivered by an authoritative PhD in person. Pain education is propositional content paying the normal tax, so its immediate session effect should be small. Repeated sessions and recordings create conditioned entry cues. Framework uncertainty sits in durability (4-week/6-month average pain, contradicted daily by neuropathic input), not the session-level primary. Risks: n=88 with attrition, active comparator with attention and warmth, possible re-designation of weekly average pain as primary, which would lower support odds."
}