finianoneill/drug-approval-mcp

FDA Drug Approvals MCP Server

A Model Context Protocol (MCP) server that provides access to FDA drug safety data through the openFDA API. This server enables AI assistants to search adverse event reports, drug labeling information, and recall data to support drug safety analysis and research.

Features

🔧 Tools

• search_drug_events: Query FDA adverse event reports (FAERS database) for specific drugs
• get_drug_label_info: Retrieve official drug labeling information including indications, warnings, and dosage
• search_drug_recalls: Search FDA enforcement reports for drug recalls and safety alerts

📊 Resources

• Recent Drug Adverse Events: Latest adverse event reports from the FDA
• Popular Drug Labels: Labeling information for commonly searched medications
• Recent Drug Recalls: Current drug recalls and enforcement actions

💬 Prompts

• analyze_drug_safety: Generate comprehensive safety analysis for a specific drug
• drug_comparison: Compare safety profiles across multiple medications

Installation

Prerequisites

• Python 3.8+
• Required packages: httpx, mcp

Setup

1. Clone the repository:

git clone https://github.com/yourusername/drug-approval-mcp.git
cd drug-approval-mcp

2. Install dependencies:

pip install httpx mcp

3. Run the server:

python fda_mcp_server.py

Usage

As an MCP Server

The server implements the Model Context Protocol and can be integrated with MCP-compatible AI assistants. Configure your MCP client to connect to this server for FDA data access.

Direct Testing

Run the test script to verify functionality:

python test_fda_mcp_server.py

Tool Examples

Search Drug Events

# Search for adverse events related to aspirin
{
  "tool": "search_drug_events",
  "arguments": {
    "drug_name": "aspirin",
    "limit": 10,
    "date_range": "20240101_to_20241231"
  }
}

Get Drug Label Information

# Retrieve labeling information for metformin
{
  "tool": "get_drug_label_info", 
  "arguments": {
    "drug_name": "metformin",
    "limit": 5
  }
}

Search Drug Recalls

# Find Class I recalls for a specific drug
{
  "tool": "search_drug_recalls",
  "arguments": {
    "drug_name": "lisinopril",
    "classification": "Class I",
    "limit": 10
  }
}

API Endpoints Used

This server interfaces with the following FDA openFDA API endpoints:

• Drug Adverse Events: https://api.fda.gov/drug/event.json
• Drug Labeling: https://api.fda.gov/drug/label.json
• Drug Enforcement Reports: https://api.fda.gov/drug/enforcement.json

Data Sources

All data is sourced from the FDA's openFDA API, which provides access to:

• FAERS (FDA Adverse Event Reporting System): Post-market drug safety surveillance
• SPL (Structured Product Labeling): Official drug labeling information
• Recall Enterprise System (RES): FDA enforcement and recall actions

Response Format

Drug Events Response

{
  "total_results": 150,
  "events": [
    {
      "report_id": "12345678",
      "receive_date": "20241015",
      "serious": "1",
      "patient_age": "65",
      "patient_sex": "F",
      "reactions": ["Nausea", "Dizziness"],
      "drugs": [
        {
          "name": "ASPIRIN",
          "indication": "Pain relief"
        }
      ]
    }
  ]
}

Drug Label Response

{
  "total_results": 5,
  "labels": [
    {
      "brand_names": ["Tylenol"],
      "generic_names": ["acetaminophen"],
      "manufacturer": ["Johnson & Johnson"],
      "indications_and_usage": ["Pain relief and fever reduction"],
      "warnings": ["Liver damage warning"],
      "adverse_reactions": ["Nausea", "Rash"]
    }
  ]
}

Recall Response

{
  "total_results": 3,
  "recalls": [
    {
      "recall_number": "D-1234-2024",
      "product_description": "Drug name and strength",
      "reason_for_recall": "Contamination",
      "classification": "Class II",
      "status": "Ongoing",
      "recall_initiation_date": "20241001",
      "firm_name": "Manufacturer Name"
    }
  ]
}

Configuration

Logging

Set the log level using command line arguments:

python fda_mcp_server.py --log-level DEBUG

Rate Limiting

The FDA API has usage limits. The server respects these by:

• Limiting result sets to reasonable sizes
• Implementing proper error handling for rate limit responses
• Using async/await for efficient request handling

Error Handling

The server includes comprehensive error handling for:

• Invalid drug names or search parameters
• FDA API rate limiting and downtime
• Network connectivity issues
• Malformed API responses

Development

Project Structure

drug-approval-mcp/
├── fda_mcp_server.py      # Main MCP server implementation
├── test_fda_mcp_server.py # Test script for validation
├── README.md              # This documentation
├── LICENSE                # MIT License
└── .gitignore            # Git ignore file

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes and add tests
4. Submit a pull request

Testing

Run the test suite:

python test_fda_mcp_server.py

Legal and Compliance

Data Usage

• All data is publicly available through the FDA's openFDA API
• Data should be used for informational and research purposes
• Not intended as a substitute for professional medical advice

Disclaimers

• FDA adverse event reports are not verified for medical or scientific accuracy
• The existence of a report does not establish causation between a drug and adverse event
• Always consult healthcare professionals for medical decisions

License

This project is licensed under the MIT License - see the file for details.

Support

For issues and questions:

1. Check the FDA openFDA documentation
2. Review the error logs for debugging information
3. Submit issues through the repository's issue tracker

Changelog

v1.0.0

• Initial release with core MCP server functionality
• Support for adverse events, drug labels, and recalls
• Comprehensive error handling and logging
• Pre-built prompts for common analysis tasks

TO-DO NOTE: Additional data source will come from: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm