RubyIntelligence's picture
Update README.md
ac94cb4 verified
|
Raw
History Blame Contribute Delete
5.86 kB
metadata
license: cc-by-4.0
language:
  - en
tags:
  - clinical-trials
  - biomedical
  - pharma
  - healthcare
  - parquet
  - phase3
pretty_name: Clinical Trials Phase 3 Intelligence
size_categories:
  - 10K<n<100K

clinicaltrials-completed-phase3-full — Data Card

Version: 1.1.0 | Generated: 2026-06-14T06:18:47.568731+00:00 | Records: 29,633

Dataset Summary

clinicaltrials-completed-phase3-full contains structured clinical trial records from ClinicalTrials.gov, the U.S. federal registry of clinical studies operated by the National Library of Medicine (NIH). Each record includes trial status, design parameters, eligibility criteria, sponsor information, enrollment counts, and primary outcome measures — extracted, normalized, and validated by the jade pipeline. Intended for clinical researchers, pharmaceutical data teams, market intelligence analysts, and quantitative healthcare investors.

Source

  • Primary source: ClinicalTrials.gov, the U.S. National Library of Medicine's registry of clinical studies, operated by NIH. Accessed via the ClinicalTrials.gov REST API v2. Enrichment fields added post-hoc from the same source.
  • Collection method: Automated pipeline using jade v0.1.0
  • License: U.S. Government Open Data (Public Domain) — no restrictions on use or redistribution
  • PII status: No personally identifiable information. Investigator contacts and patient-level data excluded at collection time.

Quality Metrics

  • Dataset quality score: 0.9867 (threshold: 0.92)
  • Validation layers: Pydantic schema · Business rules · Statistical checks · LLM cross-check
  • Records passing validation: 29,633 of 29,633 in this version

Field Definitions

Field Type Description
nct_id str ClinicalTrials.gov unique identifier (NCT######)
title str Brief title of the clinical study
official_title str None
status str Overall recruitment status (e.g. COMPLETED, RECRUITING)
start_date str None
completion_date str None
sponsor_name str Lead sponsor organization name
sponsor_class str None
conditions list[str] Medical conditions or diseases studied
study_type str None
phases list[str] Clinical trial phases (e.g. PHASE3)
enrollment int None
allocation str None
masking str None
interventions list[dict] Interventions with type, name, and description
primary_outcomes list[dict] Primary outcome measures with measure and time_frame
secondary_outcomes list[dict] Secondary outcome measures with measure and time_frame
minimum_age str dict
maximum_age str dict
sex str None
results_available bool Whether results have been posted to ClinicalTrials.gov
locations_count int None
countries list[str] Countries where the study is conducted
trial_duration_days int None
therapeutic_area str Disease area classification: Oncology, Cardiovascular, Neurology / CNS, Infectious Disease, Metabolic / Endocrine, Immunology / Rheumatology, Respiratory, Gastroenterology, Dermatology, Ophthalmology, Rare Disease / Orphan, Other.
primary_endpoint_met str Whether the primary endpoint was achieved: 'true', 'false', or 'unknown'. Coverage: ~45% of records (those with posted results on ClinicalTrials.gov). Extracted by LLM from the results section.

Known Limitations

  • Location data (countries, locations_count) is absent for approximately 15% of studies where ClinicalTrials.gov omits the contactsLocationsModule.
  • sponsor_class may include values outside the standard set (e.g. 'OTHER') for a small fraction of records; these receive a statistical soft warning but are not excluded.
  • Completion dates reflect the study's reported anticipated or actual completion, not verified outcomes.
  • Age fields (minimum_age, maximum_age) are returned as parsed {value, unit} objects rather than plain strings.
  • primary_endpoint_met is 'unknown' for ~55% of records where ClinicalTrials.gov has not yet posted results. This is expected — results are posted voluntarily and often lag trial completion.
  • therapeutic_area is LLM-assigned and may differ from formal MeSH classification for trials spanning multiple disease areas.

Update Cadence

Weekly (every Friday night)

Provenance

All records include a provenance object documenting:

  • Source URL and retrieval timestamp
  • Extraction model and pipeline version
  • Validation model and quality score
  • Whether LLM cross-check was performed

License

This dataset is derived entirely from U.S. federal government public records (ClinicalTrials.gov, operated by the National Library of Medicine, NIH). The source data is in the public domain. This structured, cleaned version is provided under CC BY 4.0 — attribution required.

How to Access the Full Dataset

The 50-row sample.csv in this repository is free to use under CC BY 4.0.

Full dataset (29,633 records, 26 fields, Parquet format): Available for purchase at Ruby Intelligence on Gumroad

  • One-time download: $1,499

Generated by jade pipeline v0.1.0 · 2026-06-14T06:18:47.568731+00:00