Nason/clinicaltrials-database

---
license: cc0-1.0
task_categories:
- tabular-classification
- tabular-regression
tags:
- clinical-trials
- clinicaltrials-gov
- aact
- duckdb
- medical
- health
- government-data
pretty_name: ClinicalTrials.gov Database
size_categories:
- 10M<n<100M
---

# ClinicalTrials.gov Database (AACT)

A clean, queryable DuckDB database built from the [AACT flat files](https://aact.ctti-clinicaltrials.org/) -- the most comprehensive structured export of ClinicalTrials.gov data.

**56,397,018 rows** across **48 tables** covering every registered clinical trial.

Built with [clinicaltrials-database](https://github.com/ian-nason/clinicaltrials-database).

## Quick Start

### DuckDB CLI

```sql
INSTALL httpfs;
LOAD httpfs;
ATTACH 'https://huggingface.co/datasets/Nason/clinicaltrials-database/resolve/main/clinicaltrials.duckdb' AS ct (READ_ONLY);

-- Trials by phase
SELECT overall_status, COUNT(*) as n
FROM ct.studies
GROUP BY overall_status
ORDER BY n DESC;
```

### Python

```python
import duckdb
con = duckdb.connect()
con.sql("INSTALL httpfs; LOAD httpfs;")
con.sql("""
ATTACH 'https://huggingface.co/datasets/Nason/clinicaltrials-database/resolve/main/clinicaltrials.duckdb'
AS ct (READ_ONLY)
""")
con.sql("SELECT * FROM ct.studies LIMIT 5").show()
```

DuckDB uses HTTP range requests, so only the pages needed for your query are downloaded.

## Tables

| Table | Description | Rows |
|-------|-------------|------|
| `reported_events` | Summary of reported adverse events | 11,446,277 |
| `outcome_measurements` | Summary data for outcome measures by study group | 4,777,352 |
| `browse_conditions` | NLM-generated MeSH terms for study conditions | 4,194,046 |
| `design_outcomes` | Planned outcome measures and observations | 3,566,612 |
| `facilities` | Facility names, addresses, and recruiting status | 3,410,973 |
| `baseline_measurements` | Demographic and baseline measures by group | 2,808,533 |
| `browse_interventions` | NLM-generated MeSH terms for study interventions | 2,480,003 |
| `result_groups` | Aggregate list of group titles/descriptions for results reporting | 2,120,694 |
| `outcome_counts` | Sample sizes for each outcome by study group | 1,537,817 |
| `keywords` | Investigator-provided keywords describing the study | 1,523,516 |
| `design_group_interventions` | Cross-reference mapping design groups to interventions | 1,283,125 |
| `study_references` | Citations to publications related to the study | 1,078,399 |
| `design_groups` | Protocol-specified groups or cohorts assigned to interventions | 1,056,031 |
| `conditions` | Disease or condition names studied in each trial | 1,025,650 |
| `interventions` | Interventions or exposures: drugs, devices, procedures, vaccines | 973,952 |
| `sponsors` | Sponsor and collaborator names and types | 920,618 |
| `milestones` | Participant progress through each stage of the study | 855,865 |
| `countries` | Countries where the study has sites | 783,722 |
| `id_information` | Identifiers other than NCT ID (org study IDs, secondary IDs) | 748,312 |
| `outcomes` | Outcome measure descriptions and time frames | 638,016 |
| `outcome_analysis_groups` | Groups involved in each outcome analysis | 609,509 |
| `drop_withdrawals` | Summary of participant withdrawals: counts and reasons | 577,923 |
| `reported_event_totals` | Totals of reported adverse events by category | 577,536 |
| `calculated_values` | AACT-computed fields: months to report results, facilities count, etc. | 576,029 |
| `studies` | Core study record: title, status, phase, type, dates, enrollment, and regulatory info | 576,029 |
| `brief_summaries` | Brief text summary of the study protocol | 575,064 |
| `detailed_descriptions` | Detailed text description of the study protocol | 575,064 |
| `eligibilities` | Participant eligibility criteria (inclusion/exclusion, age, sex) | 575,064 |
| `designs` | Study design details: allocation, masking, assignment, purpose | 571,286 |
| `responsible_parties` | Parties responsible for submitting study information | 557,553 |
| `overall_officials` | People responsible for overall scientific leadership | 519,601 |
| `search_term_results` | | 494,618 |
| `intervention_other_names` | Synonymous names for interventions | 470,456 |
| `facility_contacts` | Contact information at each study facility | 409,886 |
| `outcome_analyses` | Statistical analyses performed on outcomes | 315,026 |
| `baseline_counts` | Sample sizes at baseline for each study group | 230,199 |
| `facility_investigators` | Investigator names at each study facility | 223,509 |
| `central_contacts` | Primary and backup contacts for enrollment questions | 217,960 |
| `ipd_information_types` | Individual participant data sharing information types | 87,849 |
| `provided_documents` | Protocol, SAP, and informed consent form documents | 78,018 |
| `result_contacts` | Points of contact for scientific information about results | 76,963 |
| `result_agreements` | Agreements between sponsor and principal investigators about results | 76,963 |
| `participant_flows` | Recruitment and pre-assignment details | 76,963 |
| `links` | Web links relevant to the study | 74,921 |
| `pending_results` | Events related to submission of results for QC review | 32,304 |
| `documents` | Full study protocol and statistical analysis plan | 10,692 |
| `retractions` | Retraction notices for study results or publications | 334 |
| `search_terms` | | 186 |

## Data Source

[AACT (Aggregate Analysis of ClinicalTrials.gov)](https://aact.ctti-clinicaltrials.org/) -- maintained by the Clinical Trials Transformation Initiative (CTTI). Updated daily from ClinicalTrials.gov. This is public domain U.S. government data.

## License

Database build code: MIT. Underlying data: public domain (U.S. government work).

## GitHub

Full source code, build instructions, and data dictionary: [github.com/ian-nason/clinicaltrials-database](https://github.com/ian-nason/clinicaltrials-database)