{
 "generated": "2026-10-03",
 "studies": {
  "excluded": 806,
  "total": 2437,
  "pct": 33.1
 },
 "site_rows": {
  "excluded": 999,
  "total": 5661,
  "pct": 17.6
 },
 "facility_rows_quarantined": 36,
 "residual_anesthesia_served": 167,
 "residual_note": "Studies the classifier kept but whose title or conditions name an anesthesia or procedural term. Retained deliberately: many are psychiatric trials that merely mention a procedural agent (a psychedelic masked with propofol). The exclusion count above is what was removed, not a claim that the category is now empty.",
 "wellness_listings_dropped": 980,
 "query_artifact_studies_dropped": 93,
 "rules": [
  {
   "stage": "Clinics: source ingest",
   "rule": "HealingMaps listings NOT tagged as a psychedelic/psychedelic-therapy category (GLP-1, peptide, generic wellness clinics the source also lists)",
   "removed": 980,
   "enforced_by": "psychedelic-only keep-rule in harvest_healingmaps.py"
  },
  {
   "stage": "Trials: query expansion",
   "rule": "ClinicalTrials.gov records matching a search term but carrying NO psychedelic intervention (e.g. 'DMT' = disease-modifying therapy in multiple-sclerosis trials)",
   "removed": 93,
   "enforced_by": "no-drug-evidence artifact drop in harvest_ctgov.py"
  },
  {
   "stage": "Trials: relevance",
   "rule": "Studies whose indication is anesthesia/analgesia/procedural-sedation, sharing only a molecule (ketamine) with the therapeutic field",
   "removed": 806,
   "enforced_by": "relevance classifier (psychiatric/anesthesia_analgesia/other); anesthesia excluded from index"
  },
  {
   "stage": "Trial sites: relevance",
   "rule": "Trial-site rows belonging to anesthesia/analgesia studies above",
   "removed": 999,
   "enforced_by": "same classifier, propagated to site rows"
  },
  {
   "stage": "Publications: scope",
   "rule": "Psychedelic term appears only in the abstract and the title is not clinical \u2014 a passing mention (how a COVID-19 meta-analysis and a Japan-Hungary diplomatic history entered the corpus before 2026-10-02)",
   "removed": 13931,
   "enforced_by": "is_relevant() in harvest_openalex.py; scope = psychedelic medicine"
  },
  {
   "stage": "Publications: scope",
   "rule": "Returned by the search but no psychedelic term in title or abstract",
   "removed": 6453,
   "enforced_by": "is_relevant() in harvest_openalex.py"
  },
  {
   "stage": "Publications: scope",
   "rule": "Psychedelic is the title subject but the framing is humanities, ethnography, taxonomy or cultivation, not medicine",
   "removed": 1174,
   "enforced_by": "is_relevant() in harvest_openalex.py"
  },
  {
   "stage": "Facilities: merge",
   "rule": "Non-treatment organizations (law firms, funders, regulators, advocacy groups, drug developers) reclassified out of the clinics layer \u2014 they remain in the industry layer",
   "removed": 147,
   "enforced_by": "industry cross-reference in merge_master.py"
  },
  {
   "stage": "Facilities: merge",
   "rule": "Duplicate operator rows collapsed (same name and country, city missing or contained)",
   "removed": 63,
   "enforced_by": "second-pass dedupe in merge_master.py; every merge logged to dedupe_merges.csv"
  },
  {
   "stage": "Facilities: merge",
   "rule": "Contact values stripped as source-template leaks (e.g. an NIMH inquiry address printed on hundreds of unrelated clinic listings)",
   "removed": 1163,
   "enforced_by": "shared-contact guard in merge_master.py; values logged to shared_contacts_stripped.csv"
  },
  {
   "stage": "Facilities: validation",
   "rule": "Rows quarantined at merge: non-facility organizations, wellness residue, unidentifiable rows",
   "removed": 36,
   "enforced_by": "validate.py gate; quarantine.csv with written reasons"
  }
 ]
}