Open data access
Download and inspect the index.
The public files expose the same condition-level values shown on the site, together with source and methodology versions. Current snapshot:2026-09-11 · methodology v2.1.0 · SHA-256 9a17314144e5c9c9…
- Download JSON — the exact snapshot document: totals, per-condition aggregates, standardized treatment terms with mapped patient results, and record-level membership manifests (NCT IDs in and out of each set).
- Download CSV — one condition-level row per condition set, using the columns documented below.
Both endpoints return an ETag containing the snapshot’s SHA-256 hash and aLast-Modified date, so you can verify that two downloads — or a citation and a re-check — refer to byte-identical data. What each refresh changed is logged on the trial record changelog (RSS).
CSV field dictionary
Columns appear in this order in the file. Counts describe registry records, not verified facts about treatments.
| Column | Meaning |
|---|---|
condition | Condition display name, e.g. “Acne vulgaris”. |
condition_slug | URL slug of the condition page (/evidence-index/<slug>). |
source_data_timestamp | ClinicalTrials.gov data cut this snapshot was built from (ISO 8601). |
methodology_version | Version of the published Skinlogie methodology that produced the row. |
source_matches | All study types returned by the broad official condition search. |
interventional_search_matches | Broad search matches whose registry study type is interventional. |
excluded_after_condition_check | Interventional matches removed by the versioned condition-label screen (e.g. non-atopic “eczema”, arthritis-only psoriasis labels). |
included_interventional_studies | Records passing the condition screen — the denominator for every other count in the row. |
completed_studies | Included records with registry status Completed. A registry status, not a quality judgment. |
active_studies | Included records with recruiting or otherwise active registry statuses. |
studies_with_posted_results | Included records of any status with posted registry summary results. |
completed_with_posted_results | Completed included records with posted registry summary results. |
completed_results_percentage | completed_with_posted_results ÷ completed_studies × 100. |
actual_enrollment_sum | Sum of enrollment counts submitted as “actual”. Participant entries, not guaranteed unique people. |
anticipated_enrollment_sum | Sum of enrollment counts submitted as “anticipated” (planned, not achieved). |
race_or_ethnicity_reports | Completed-with-results records whose baseline tables report race or ethnicity. |
race_or_ethnicity_denominator | Completed-with-results records assessed for that reporting. |
race_or_ethnicity_percentage | reports ÷ denominator × 100. |
skin_phototype_reports | Completed-with-results records whose baseline tables report a skin-phototype scale (e.g. Fitzpatrick). |
skin_phototype_denominator | Completed-with-results records assessed for phototype reporting. |
skin_phototype_percentage | reports ÷ denominator × 100. |
listed_country_count | Distinct countries in listed study locations. Where studies ran — not participant nationality or ancestry. |
source_query | The official ClinicalTrials.gov search URL the condition set is built from. |
The JSON download additionally carries the treatment-term detail shown on condition pages: standardized NLM term groups (at most the 60 highest-volume terms per condition, withtreatmentTermCount counting all qualifying terms), registered intervention labels, outcome domains, participant categories, research activity, and conservatively mapped study-level patient results. An empty patient-result array means “unable to determine” — not “no effect”.
The enrichment layer
A second, separately versioned file widens the snapshot’s patient-result mapping from the recent-study candidates to every included study with posted results, applying the same published mapping rules plus three documented gate relaxations (dose-arm fan-out to at most 3 unambiguous treatment groups, matching-placebo co-intervention arms, and per-intervention rather than per-study combination exclusion). Per treatment term it adds: all kept comparative results with acomparatorClass (placebo/vehicle vs active comparator vs dose comparison), submitted serious-adverse-event group counts for mapped studies, registry-linked RESULT publications (PubMed ids), the earliest favourable comparative time point, recruiting study lists, and top-term pair matrices (co-registration and head-to-head counts). Nothing is pooled across studies.
- Download the enrichment JSON — versioned, with the snapshot hash it was built against.
Reuse and citation
Source records are from ClinicalTrials.gov(U.S. National Library of Medicine) and remain subject to itsterms. Skinlogie’s derived aggregates — the screens, groupings, and computed fields in these files — are licensed CC BY 4.0: reuse them freely, including commercially, with attribution to Skinlogie and ClinicalTrials.gov.
When you reuse the data, keep the methodology version and snapshot hash with your copy — they are what make a number checkable later. No field in the dataset is an efficacy score, a comparative benefit estimate, or an appraisal of study quality; please do not present the counts as one.
Cite this index
Skinlogie. Dermatology Clinical Trial Evidence Index (methodology v2.1.0; ClinicalTrials.gov data through 2026-09-11; snapshot 9a17314144e5). https://skinlogie.com/evidence-index. Accessed 2026-09-13.