HEOR Agent
AI-powered Health Economics and Outcomes Research (HEOR) MCP server for pharmaceutical, biotech, CRO, and medical affairs teams.
What it does
Automates the full HEOR workflow: literature review, evidence synthesis, economic modeling, indirect comparisons, and HTA dossier preparation — all callable as MCP tools from Claude.ai, Claude Code, or any
MCP-compatible host.
14 Tools
| Tool | Purpose |
|---|---|
literature_search | Search 41 data sources with PRISMA audit trail |
screen_abstracts | PICO-based relevance scoring and study design classification |
evidence_network | Build treatment comparison network, assess NMA feasibility |
indirect_comparison | Bucher method + Frequentist NMA |
population_adjusted_comparison | MAIC/STC per NICE DSU TSD 18 |
survival_fitting | Fit 5 parametric distributions to KM data (NICE DSU TSD 14) |
cost_effectiveness_model | Markov / PartSA with PSA, OWSA, CEAC, EVPI, EVPPI, scenarios |
budget_impact_model | ISPOR-compliant year-by-year BIA |
hta_dossier_prep | NICE, EMA, FDA, IQWiG, HAS, EU JCA — with auto-GRADE |
validate_links | HTTP validation of citation URLs |
project_create | Persistent project workspace |
knowledge_search / read / write | Project knowledge base |
41 Data Sources
Biomedical: PubMed, ClinicalTrials.gov, bioRxiv/medRxiv, ChEMBL
HTA Appraisals: NICE TAs, CADTH/CDA-AMC, ICER, PBAC, G-BA AMNOG, HAS, IQWiG, AIFA, TLV, INESSS, ISPOR
Cost References: CMS NADAC, PSSRU, NHS Costs, BNF, PBS Schedule
Epidemiology: WHO GHO, World Bank, OECD Health, IHME GBD, All of Us
FDA Regulatory: Orange Book, Purple Book
Enterprise (require API key): Embase, ScienceDirect, Cochrane, Citeline, Pharmapendium, Cortellis, Google Scholar
LATAM: DATASUS, CONITEC, ANVISA, PAHO, IETS, FONASA
APAC: HITAP
Installation
Claude Code
claude mcp add heor-agent -- npx heor-agent-mcp
Claude Desktop
{
"mcpServers": {
"heor-agent": {
"command": "npx",
"args": ["heor-agent-mcp"]
}
}
}
Try it
Web demo (BYOK): https://web-michael-ns-projects.vercel.app
Example prompt:
▎ Create a project for semaglutide in obesity. Search literature for evidence, screen results for adults with obesity comparing semaglutide to placebo for weight loss, then draft a NICE STA dossier.
This single prompt exercises: project creation → literature search (3-run stability) → PICO screening → HTA dossier with auto-GRADE.
Links
- GitHub: https://github.com/neptun2000/heor-agent-mcp
- npm: https://www.npmjs.com/package/heor-agent-mcp
- Features: https://github.com/neptun2000/heor-agent-mcp/blob/master/docs/FEATURES.md
- Changelog: https://github.com/neptun2000/heor-agent-mcp/blob/master/CHANGELOG.md
License
MIT
Server Config
{
"mcpServers": {
"heor-agent": {
"command": "npx",
"args": [
"heor-agent-mcp"
]
}
}
}