# Statistics data dictionary

Prepared November 17, 2026, for Psychedelic Therapy Index.

These files are source-linked editorial extracts of published aggregate facts. They contain no patient-level data and are not new clinical research. Keep each row's source and qualifications with any reused statistic. The linked original publishers and agencies retain their own rights; this document does not grant rights to their articles, figures or full datasets.

## Files

| File | Contents | Rows |
| --- | --- | ---: |
| `ketamine-study-statistics.csv` | Selected ketamine/esketamine study endpoints | 8 |
| `psychedelic-study-statistics.csv` | Selected psilocybin/MDMA study endpoints | 7 |
| `oregon-psilocybin-service-statistics.csv` | Three service-volume measures for each of two official reporting periods | 6 |

The two esketamine-monotherapy rows describe two doses in **one trial**, not two independent samples. The two MAPP2 rows describe two outcomes in **one trial**. Never sum their study sample sizes.

## Clinical study fields

| Field | Definition |
| --- | --- |
| `study_id` | Editorial identifier; suffixes distinguish outcomes or comparisons within a trial. |
| `publication_year` | Year of the cited original report, or first online publication when explicitly described. |
| `population` | Studied clinical population. This is not a complete eligibility list. |
| `study_design` | Short description of how the study was run. |
| `treatment`, `comparator` | Tested intervention and assigned comparison. A blank comparator means none was used. |
| `study_sample_size` | Source sample size described in the row. Read `analysis_denominator_note`; this may be randomized people, completers or an analytic cohort. |
| `analysis_denominator_note` | Clarifies endpoint groups, exclusions, completers and repeated rows. |
| `endpoint` | Exact type of measure extracted. “Response,” “remission” and symptom-score change are different outcomes. |
| `timepoint` | Assessment time or period, rather than a promise of durable effects. |
| `treatment_value`, `comparator_value` | Source-reported point estimate or range. For a between-group difference, `treatment_value` is the difference and `comparator_value=0` denotes its reference, not the comparator group's observed score. |
| `unit` | Percentage, percentage of days, score points or range. Negative score changes mean lower symptoms on the relevant scale; they are not negative percentages. |
| `confidence_interval_or_note` | Source interval or a necessary qualification. A missing interval is not a claim of certainty. |
| `interpretation_limit` | Main reason the result should not be generalized or combined with another row. |
| `source_url` | Primary publication or original report indexed by PubMed. |
| `source_checked` | Date the source was checked for this extraction, in YYYY-MM-DD format. |

## Oregon administrative fields

| Field | Definition |
| --- | --- |
| `geography` | Oregon regulated psilocybin services; not the United States as a whole. |
| `period_start`, `period_end` | Inclusive reporting-period boundaries, YYYY-MM-DD. |
| `metric` | Official measure represented by the row. |
| `value`, `unit` | Published aggregate count and unit. |
| `denominator_note` | Prevents treating service-center reports as deduplicated nationwide people or clinical outcomes. |
| `source_url`, `source_location` | Original agency file and location of the extracted value. |
| `source_checked` | Extraction check date. |

## Missing values and calculations

Blank fields mean “not extracted” or “not applicable,” never zero. No suppressed Oregon fields are included. In the full OHA data, `-99` can mark a value suppressed for confidentiality; it must not be treated as a negative count or recoded as zero.

Published percentages are retained at their published precision. Do not infer a numerator by multiplying a rounded percentage by a total sample size. Model-adjusted means can differ from simple arithmetic means.

No pooled effect size, overall success rate, provider ranking, market estimate or statewide adverse-event rate was calculated. These extracts cannot support those conclusions.

## Files on the website

Serve these four files under `/downloads/` with the exact filenames. The Markdown dictionary can be offered directly as a text download or rendered through a simple accessible text view at the same URL. The full evidence tables and key qualifications must remain visible in the article HTML without downloading a file.
