ClinicalTrials Database Dynamic

A scientific computing skill for searching and retrieving data from ClinicalTrials.gov — the comprehensive registry of clinical studies maintained by the National Library of Medicine. ClinicalTrials Database Dynamic helps you find trials by condition, intervention, or sponsor, and retrieve detailed protocol and results data.

When to Use This Skill

Choose ClinicalTrials Database Dynamic when:

• Searching for clinical trials by disease, drug, or sponsor
• Retrieving trial design details (endpoints, enrollment, phases)
• Monitoring trial status updates and result postings
• Building systematic review search strategies across registries

Consider alternatives when:

• You need published trial results (use PubMed)
• You need drug approval information (use FDA database)
• You need pharmacokinetic data (use dedicated PK databases)
• You need real-world evidence (use claims databases)

Quick Start

Copy
claude "Find active Phase 3 trials for Alzheimer's disease treatments"

Copy
import requests

# ClinicalTrials.gov API v2
base_url = "https://clinicaltrials.gov/api/v2/studies"

params = {
    "query.cond": "Alzheimer Disease",
    "filter.overallStatus": "RECRUITING",
    "filter.phase": "PHASE3",
    "pageSize": 10,
    "sort": "LastUpdatePostDate:desc",
    "fields": "NCTId,BriefTitle,OverallStatus,Phase,EnrollmentCount,LeadSponsorName,StartDate,PrimaryCompletionDate"
}

response = requests.get(base_url, params=params)
data = response.json()

for study in data.get("studies", []):
    info = study["protocolSection"]["identificationModule"]
    status = study["protocolSection"]["statusModule"]
    print(f"\n{info['nctId']}: {info['briefTitle']}")
    print(f"  Phase: {status.get('phases', ['N/A'])}")
    print(f"  Status: {status['overallStatus']}")

Core Concepts

ClinicalTrials.gov API v2

Endpoint	Purpose	Example
/studies	Search studies	?query.cond=cancer
/studies/{nctId}	Get single study	/studies/NCT12345678
/studies?fields=	Select specific fields	fields=NCTId,BriefTitle
/stats/size	Count matching studies	Get result count

Study Phases

Phase	Description	Typical Size
Early Phase 1	First-in-human, dose finding	10-30
Phase 1	Safety, pharmacokinetics	20-80
Phase 2	Efficacy signal, dose ranging	100-300
Phase 3	Confirmatory efficacy trial	300-3000+
Phase 4	Post-marketing surveillance	1000+

Advanced Search

Copy
def search_trials(condition=None, intervention=None, sponsor=None,
                  phase=None, status="RECRUITING", max_results=50):
    """Search ClinicalTrials.gov with multiple filters"""
    params = {"pageSize": min(max_results, 100)}

    if condition:
        params["query.cond"] = condition
    if intervention:
        params["query.intr"] = intervention
    if sponsor:
        params["query.spons"] = sponsor
    if phase:
        params["filter.phase"] = phase
    if status:
        params["filter.overallStatus"] = status

    all_studies = []
    next_token = None

    while len(all_studies) < max_results:
        if next_token:
            params["pageToken"] = next_token

        resp = requests.get("https://clinicaltrials.gov/api/v2/studies",
                          params=params)
        data = resp.json()
        studies = data.get("studies", [])
        all_studies.extend(studies)
        next_token = data.get("nextPageToken")
        if not next_token:
            break

    return all_studies[:max_results]

# Example: CRISPR therapies in recruiting trials
trials = search_trials(intervention="CRISPR", status="RECRUITING")

Configuration

Parameter	Description	Default
api_version	API version (v2 recommended)	v2
page_size	Results per request (max 100)	20
default_status	Default study status filter	RECRUITING
fields	Default fields to retrieve	All fields
sort	Sort order for results	LastUpdatePostDate:desc

Best Practices

1. Use specific condition terms from MeSH vocabulary. "Alzheimer Disease" returns more targeted results than "Alzheimer's" or "dementia." ClinicalTrials.gov uses MeSH terms for indexing — match their vocabulary for precise searches.

2. Filter by status for actionable results. Use RECRUITING for trials currently enrolling, ACTIVE_NOT_RECRUITING for ongoing trials, or COMPLETED for trials with potential results. Unfiltered searches include terminated and withdrawn studies.

3. Request only needed fields. The full study record is large. Use the fields parameter to request only the fields you need (NCTId, BriefTitle, Phase, Status). This reduces response size and parsing time significantly.

4. Paginate through all results for systematic reviews. ClinicalTrials.gov returns max 100 studies per request. Use pageToken for complete result sets. For systematic reviews, document your search strategy including all filters for reproducibility.

5. Check for posted results on completed trials. Completed trials may have results posted on ClinicalTrials.gov before (or instead of) journal publication. Check the hasResults flag and retrieve outcome measures, adverse events, and participant flow data.

Common Issues

Search returns too many irrelevant trials. Combine filters — use query.cond with filter.phase and query.intr together. Adding a specific intervention or sponsor dramatically reduces noise. Avoid generic terms like "cancer" without additional qualifiers.

API returns 429 Too Many Requests. ClinicalTrials.gov enforces rate limits. Add 0.5-second delays between paginated requests. For bulk data needs, consider using the ClinicalTrials.gov bulk download files instead of API queries.

Enrolled numbers don't match published results. ClinicalTrials.gov enrollment counts may differ from publications due to screen failures, withdrawals, and different counting methodologies. Always note the data source when citing enrollment figures.