Awesome Health Architecture

A curated, open knowledge base for digital health architecture, interoperability, standards, platforms, infrastructure, security, data, AI and health information exchange.

This is not a list of software. It is an attempt to answer one question properly:

If you have to architect a digital health system, where do you start?

The material is arranged the way the work actually proceeds — from a health problem, through capability and workflow, down to data, standards, applications, infrastructure, and the governance that keeps all of it alive.

---

The architecture map

                     Health system problem
                              │
                              ▼
                    Business architecture
             (who is served, which capabilities)
                              │
                              ▼
                  Digital health architecture
                              │
        ┌─────────────────────┼─────────────────────┐
        ▼                     ▼                     ▼
      Data              Applications        Interoperability
        │                     │                     │
        ▼                     ▼                     ▼
   FHIR / openEHR       EMR / HMIS / LMIS      HIE / OpenHIE
        │                     │                     │
        └──────────┬──────────┴──────────┬──────────┘
                   ▼                     ▼
              Registries            Terminology
       (client, facility, HWR)   (SNOMED, LOINC, ICD)
                   │                     │
                   └──────────┬──────────┘
                              ▼
                    Identity, consent, trust
                              │
                              ▼
                        Infrastructure
                              │
                              ▼
                     Analytics and reporting
                              │
                              ▼
                              AI

Read it downward when you are designing, and upward when you are debugging: most failures that look like an application problem are really a registry, identifier or governance problem one layer down.

---

Start here

If you are…	Start with
New to digital health	Learning paths → Glossary
Designing a national architecture	Digital health architecture → OpenHIE → Maturity model
Building an integration	Interoperability → HL7 FHIR → Integration engines
Writing a tender or policy	Governance → Standards directory → Checklists
Adding AI to a health system	AI architecture → MCP in healthcare → AI ethics

---

Sections

Architecture

• Architecture overview — the layers and how they relate
• Digital health architecture
• Architecture frameworks — health-specific and general, kept apart
• Enterprise architecture
• C4 model — describing systems at four zoom levels
• Architecture decision records
• Architecture patterns — the pattern catalogue
• Maturity model

Exchange

• OpenHIE — the reference architecture
  • Point-of-service systems
  • Interoperability layer
  • Business domain services
• Interoperability — the four levels
• Health information exchange
• Integration engines
• APIs

Standards

• Standards directory — every standard, with owner and layer
• Exchange — HL7 FHIR, HL7 v2, CDA, DICOM
• FHIR ecosystem — Implementation guides, SMART on FHIR, Bulk Data
• Terminology — SNOMED CT, LOINC, ICD, Terminology services
• Clinical models — openEHR

Registries

• Registries overview
• Client registry / MPI
• Facility registry
• Health worker registry

Identity, security and trust

• Overview
• OAuth 2.0 and OpenID Connect
• Consent and trust
• Security architecture
• ISO 27001, ISO 27799, GDPR

Data and analytics

• Data architecture
• Health data, Open data, Data governance
• Analytics tools, GIS

Platforms

• Platform directory
• OpenMRS, DHIS2, OpenHIM
• FHIR servers, HAPI FHIR
• CLI tools

Infrastructure

• Infrastructure overview
• Docker, Observability, DevSecOps
• AWS, Azure, GCP

AI

• AI architecture
• Machine learning, Clinical AI, Imaging AI
• MCP in healthcare — emerging, not a standard
• AI ethics

Global guidance and domains

• WHO and global guidance, SMART Guidelines, Digital public infrastructure
• Health domains — maternal and child health, community health, offline-first, surveillance, supply chain, financing
• Country architectures — Nepal

Working materials

• Glossary
• Checklists
• Templates — reference architecture, ADR
• Governance
• Learning paths

---

How resources are classified

Every external resource on this site is one of four tiers. The tier is stated wherever it is not obvious from context.

Tier	Meaning	Examples
1 — Official	Normative or authoritative publication by the body that owns the thing	WHO, HL7, ISO, IETF, DICOM, SNOMED International, Regenstrief, a health ministry
2 — Community	Established open-source project or professional community	OpenHIE, OpenMRS, DHIS2, HAPI FHIR, OpenHIM
3 — Educational	Explanatory material, including the pages on this site	Books, courses, tutorials, blog posts
4 — Experimental	Research, prototypes, emerging technology	Draft specifications, MCP in healthcare, most agentic AI patterns

Tier 3 and Tier 4 material is never presented as normative. Where a specification is a draft or under consultation, it is labelled as such rather than described as a standard.

---

Conventions used here

• Architecture first, technology second. Every technology page says which architectural problem the technology solves, where it sits, what the alternatives are, and what it costs to operate.
• No universal winners. Comparisons explain which context suits which option. FHIR does not "beat" openEHR; centralised does not "beat" federated.
• Reference architecture and implementation technology are kept distinct. OpenHIE is an architecture; OpenHIM is one implementation of one of its components.
• Needs verification is preferred to a confident guess. Where a claim could not be traced to a primary source, it is marked rather than asserted.
• Source priority runs: WHO and ITU → HL7, ISO, IETF, DICOM, SNOMED International, Regenstrief, OpenID Foundation, NIST → national health agencies → established open-source projects → academic work → community writing.

---

Contributing and verification

Links on this site are checked mechanically. From the repository root:

npm run check-links   # verify every external URL in the knowledge base
npm run validate      # structural checks: frontmatter, duplicates, orphans

The editorial rules are short: primary sources, no invented resources, tiers stated, and a Last verified date on anything that moves.