Healthcare & Medical Data MCP Server — FDA, PubMed, CMS, NIH & CDC for AI Agents

Connect Claude, ChatGPT, Cursor, and custom LLM agents directly to authoritative U.S. healthcare, regulatory, and public-health data through the Model Context Protocol (MCP). One server, nine agent-callable tools spanning the FDA, PubMed/NLM, CMS, NIH RePORTER, and the CDC — no enterprise data contracts, no per-source API onboarding.

This is a clinical / regulatory / public-health data layer for AI agents. It orchestrates a fleet of dedicated, production NexGenData data actors behind a single MCP interface so your agent can go from "what adverse events are reported for ozempic?" to structured JSON in one tool call.

What You Get (tools the agent can call)

• search_drug_labels(query, data_type, max_results) — openFDA drug labels, adverse events, recalls, and device records
• search_pubmed(query, max_results) — PubMed biomedical literature: titles, authors, abstracts, journals, PMIDs
• fda_recalls(search_term, category, severity, max_results) — FDA product recalls and enforcement actions (drug/food/device)
• fda_drug_approvals(date_range, approval_type, manufacturer, ...) — recent FDA approvals (NDA/BLA/ANDA/NME) with special designations
• fda_adverse_events(search_term, event_type, serious_only, ...) — FAERS (drug) and MAUDE (device) adverse event reports
• fda_warning_letters(manufacturer_name, product_type, ...) — FDA warning letters and Form 483 inspection observations
• cms_open_payments(year, payment_type, physician_last_name, ...) — CMS Open Payments: industry payments to physicians/hospitals
• nih_grants(fiscal_years, agencies, organization_name, pi_last_name, ...) — NIH RePORTER federally funded research projects
• cdc_surveillance(disease, state_filter, age_group, ...) — CDC FluView / WONDER respiratory-disease surveillance

Use Cases

• AI clinical-research assistants — pull PubMed evidence, FDA labels, and adverse events to draft an evidence summary
• Pharma & medtech regulatory intelligence — monitor approvals, recalls, warning letters, and 483s for a competitor or product class
• Pharmacovigilance co-pilots — surface FAERS/MAUDE adverse-event signals for a drug or device on demand
• Payment-transparency / compliance tooling — query CMS Open Payments by physician or manufacturer
• Grants & research-landscape analysis — map NIH funding by organization, PI, or therapeutic area
• Public-health dashboards — feed agents live CDC influenza / COVID-19 / RSV surveillance data
• Health-data chatbots — give consumer and provider LLM apps grounded medical data

How It Works

Each tool maps to a dedicated, individually maintained NexGenData actor that pulls from the relevant public source (openFDA, NCBI/PubMed E-utilities, openPaymentsData.cms.gov, NIH RePORTER API, CDC FluView/WONDER). The MCP server runs each underlying actor on demand and returns its dataset items as structured MCP tool output. You get one connection, one credential, and a consistent JSON shape across all nine data domains.

Quick Start

Wire it into an MCP-compatible client (Claude Desktop, Cursor, Windsurf, n8n) using the streamable-HTTP endpoint:

https://nexgendata--healthcare-medical-mcp-server.apify.actor/mcp

Example Claude Desktop config:

{
  "mcpServers": {
    "healthcare-medical": {
      "url": "https://nexgendata--healthcare-medical-mcp-server.apify.actor/mcp?token=YOUR_APIFY_TOKEN"
    }
  }
}

Then ask your agent things like:

• "Search PubMed for recent GLP-1 cardiovascular outcome trials."
• "List FDA drug approvals with breakthrough designation in the last 90 days."
• "What serious adverse events have been reported for ozempic in the past 90 days?"
• "Show CDC influenza surveillance for the past 12 weeks."

Data Sources

FDA openFDA, FDA recall/enforcement databases, FDA FAERS & MAUDE, FDA warning letters & Form 483 inspections, NLM/NCBI PubMed, CMS Open Payments, NIH RePORTER, and CDC FluView/WONDER. All sources are public U.S. government data.

This server provides data for informational and research purposes only. It is not medical advice and is not a substitute for professional clinical judgment.

Related NexGenData actors

• FDA Orange Book Drug Patents
• Openfda Drug Search
• FDA Recall Monitor
• Clinical Trials Tracker
• NPI Healthcare Provider Leads