A G7 federated EHR blueprint

The programme

The client, a G7-country federal digital-health corporation, was set up to accelerate a national electronic health record across a federation of sub-national jurisdictions with independent health systems. The design problem was neither a single monolithic system nor ten fully-independent ones; it was a federated architecture in which each jurisdiction retained sovereignty over its own data while sharing enough common structure for a clinician, patient, or public-health authority to work across borders.

The role

For fifteen years Ron led the architecture practice inside the sponsoring corporation, most of that time as Group Director of Architecture. He was responsible for:

• the national reference architecture itself: the master blueprint against which every co-funded jurisdictional project was evaluated;
• the standards portfolio: HL7 messaging, CDA documents, terminology (SNOMED CT, LOINC), and the governance around them;
• the architecture council: a multi-stakeholder body of jurisdictional chief architects that reviewed proposed changes and enforced coherence;
• the capability-building track: turning junior architects across the federation into people who could lead their own jurisdiction’s work.

The craft

Three architectural choices, in retrospect, mattered most:

1. Registries and repositories, not warehouses. The architecture deliberately did not create a national data warehouse. Instead it defined a small number of index and locator patterns (patient, provider, encounter, document) and let clinical data stay in the jurisdiction that produced it. This turned out to be the only politically survivable design.
2. A viewer as the first population-scale artefact. Rather than solve every workflow, the architecture prioritised a clinical viewer (a read-only, cross-jurisdictional view of a patient’s records) as the first thing to work end-to-end. It gave the programme a shipping win early, and it gave every subsequent design decision a concrete acceptance criterion.
3. Standards profiled, not merely adopted. HL7 v2 and CDA were adopted, but the useful artefact was the jurisdictional implementation profile that constrained them to the ninety percent of cases actually used in the field. That profile turned out to be the durable deliverable.

The outcome

Over the decade-plus the reference architecture was in force, roughly CAD 2.1B of federal co-investment was allocated against it. Every co-funded project was reviewed for architectural conformance before funds released. Multiple governments and executives came and went; the architecture and its governance survived them all. The successor architecture team, now running the programme, was largely developed and promoted under Ron’s tenure.

What it teaches

The engagement is a case study in architecture as durable policy instrument. The reference architecture was less a technical artefact than a governance mechanism: a shared object that made federal co-investment tractable when there was no realistic path to central control. That pattern is the model Parker DHC now brings to national programmes in other geographies.