{
  "study_id": "NCT06632691",
  "agent": "framework",
  "run": 2,
  "p_primary_hypothesis_supported": 0.5,
  "effect_direction": "favours intervention",
  "effect_size_90pct_interval": {"low": -0.4, "high": 2.0, "metric": "Mean difference in VAS pain, 0-10 (positive = less pain with the headset)"},
  "moderator_prediction": "Anxious patients benefit more: threat-primed receivers price the procedure's pain heavily, and the hypnotic state change offers the most to rebook; no clear sex difference expected beyond what anxiety explains.",
  "claim_ids_used": ["PM-0023", "PM-0369", "PM-1340", "PM-0132", "PM-1286", "PM-0086", "PM-0372", "PM-0216", "PM-0845", "PM-0601", "PM-0232", "PM-0360", "PM-0092"],
  "rationale": "The framework expects hypnosis to act at the render layer, so self-rated pain is its most reachable outcome (PM-1286, PM-0372); VR induction is a state change past the critical faculty (PM-0023, PM-1340). But delivery is a recorded headset without live hypnotist authority or rapport (PM-0232, PM-0360), and patients carry an installed prior from earlier rectosigmoidoscopies that files against the new entry (PM-0216). Rectosigmoidoscopy pain is modest, compressing the room for a difference; n=100 with SD around 2.5 powers roughly a 1.4-point effect. Unblinded self-report inflates the arm difference. Expect roughly 0.8 points benefit, near the significance threshold."
}