FAIR CRF & Metadata Stewardship Prototype

This Streamlit dashboard demonstrates a prototype system for managing Case Report Forms (CRFs) and clinical metadata aligned with FAIR principles, CDISC standards, and data governance responsibilities

🔍 Features

• CRF Design Templates: View standard CRFs for Adverse Events (AE), Demographics (DM), and Lab Tests (LB).
• Metadata Repository: Curated metadata per domain, including field names, definitions, and controlled terms.
• Terminology Compliance Checker: Automatically highlights non-compliant entries using CDISC-controlled terminologies.
• Indication-Level CRF Library: Switch between oncology and cardiology-specific CRFs.
• LLM-Powered Copilot: An OpenAI-based assistant for answering CRF metadata questions.
• Governance Request Tracker: Submit and log requests for metadata changes, with justification and field-level detail.

📁 Project Structure

crf_metadata_dashboard/
├── app.py                    # Streamlit app
├── crf_metadata_csvs/        # Sample CRF data and metadata
│   ├── crf_ae_sample.csv
│   ├── metadata_repository_sample.csv
│   ├── filled_crf_noncompliant_sample.csv
│   └── cdisc_terminology.csv
└── README.md

🚀 How to Run

Install dependencies and run the app:

pip install streamlit pandas openai
streamlit run app.py

Set your OpenAI API key in .streamlit/secrets.toml:

[general]
OPENAI_API_KEY = "your_openai_api_key"

🌐 Live Demo

Hosted on Streamlit Cloud: 👉 Try the dashboard

⚠️ Disclaimer

This prototype is intended for demonstration purposes only. It is not a validated clinical system and may not reflect the complexity of real-world CRF standards. Clinical data standards require expert domain knowledge and regulatory oversight.

👩‍💼 Author

Shima Dastgheib LinkedIn | GitHub