A Federated Architecture and Operating Model for National Digital Health at Small-Nation Scale: A Design Blueprint for NHS Wales

Small nations repeatedly attempt national digital health through a single delivery body that builds, buys and runs everything — and repeatedly fail. This paper presents the alternative as a specified design rather than a policy aspiration: a federated architecture and operating model for a 3.1-million-person health system, developed for NHS Wales and generalisable to jurisdictions of comparable scale. The design applies one constraint throughout: no component is proposed unless it already operates at scale in a comparator jurisdiction — Denmark's standards-body pattern (MedCom, Sundhedsdatastyrelsen), Estonia's federated exchange (X-Road, TEHIK), or NHS England's externally-recruited delivery leadership. The architecture comprises six independently procurable layers under five design principles, the central two being the separation of standards from delivery and federation of the patient record as a query across the systems that hold the data, rather than a monolithic national platform. The operating model specifies a standards-and-interoperability body of roughly 400 staff at £45–60M per year — against the current 1,263-staff, ~£200M monopoly — with a named destination for every current post, a 36-month strangler-fig transition costed at £5–15M, and a first-90-days sequence. A striking property of the adoption path is that almost none of it requires primary legislation: the monopoly was made by executive instruments and can be unmade by them. The design states its own falsification criteria: six published tests for the ten-year digital programme the Welsh Government has announced, and a monitoring baseline whose 12- and 36-month targets the reform must meet. The contribution is a worked, costed, testable existence proof that small-nation digital health does not require a monopoly — and a template for the class of jurisdictions between one and six million people facing the same choice.