The NHS Has World-Class Values. Is Its Digital Architecture Undermining Them?

When I first moved to Australia, I carried a quiet confidence that the NHS was, despite its flaws, the best health system in the world. Not from nostalgia, but principle: universal access matters, equity matters, and the NHS’s early leadership in outcomes transparency was genuinely world-class. At the time, it was also among the most cost-effective systems globally.

What I did not appreciate then was how the NHS’s closed nature - digitally, culturally, and institutionally - would later reveal itself as a barrier to integration, learning, and innovation. That became clear only after working across both UK and Australian systems, and more starkly after returning to the UK to build a service designed to support NHS objectives while sitting outside its organisational boundaries.

Full disclosure: I lead Clean Slate Clinic, which faces these integration barriers directly and would likely benefit commercially from greater interoperability. But that vantage point also makes structural patterns visible - patterns that shape how the UK health system recognises, funds, and learns from effective care delivered beyond the NHS itself.

When Interoperability Determines Legitimacy

The NHS does not struggle to recognise good care. It struggles to recognise good care that happens outside its institutional perimeter. In the UK, innovation faces a structural binary: operate within the NHS, navigating constrained capacity, complex procurement, and finite funding, or operate outside it, accepting that impact may go unmeasured, unrecognised, and treated with institutional scepticism. This is not a moral failing; it is an architectural one.

Nowhere is this clearer than in digital interoperability.

When we assessed clinical systems for UK operations, interoperability was a core requirement. What followed was a rude awakening. In practice, meaningful interoperability is extremely difficult without an NHS contract. Systems described as “interoperable” are interoperable within the NHS, not across the wider ecosystem of not-for-profits, private providers, and social enterprises delivering care to the same patients.

Interoperability in the UK is effectively a commissioning privilege, not a patient right.

This contrasts sharply with Australia. National infrastructure such as My Health Record is provider-agnostic: public, private, NGO, Aboriginal community-controlled, and virtual providers can all participate within clear accreditation, consent, and privacy frameworks. The system is imperfect, but the philosophical choice is different. Australia prioritised openness over uniformity; the UK prioritised control and standardisation.

That choice is understandable given the NHS’s scale and political accountability, but it has consequences.

Why This Matters in Practice

These architectural decisions are not abstract.

 At Clean Slate Clinic we support vulnerable people who are unable or unwilling to access NHS services. When someone needs something as fundamental as a prescription for a safe home detox, they face additional costs because NHS e-prescribing systems are unavailable to non-commissioned providers. We face additional friction delivering care that aligns with NHS objectives but sits outside its digital infrastructure.

Pathology illustrates the same issue. In Australia, if a registered clinician refers a patient for blood tests, those tests are government-subsidised based on clinical value and accessible through any pathology provider. In the UK, access is tightly coupled to the treating NHS service or local GP, creating delay, duplication, and real cost for those outside NHS pathways.

Over time, these frictions compound. The further care develops outside NHS integration, the harder future integration becomes. This is not just an interoperability problem; it is a policy choice about how legitimacy is conferred, and to whom.

What the NHS Gets Right

None of this reflects a lack of digital capability. The NHS demonstrates genuine leadership in specific domains: OpenSAFELY’s privacy-preserving pandemic analytics, NICE’s internationally respected evidence frameworks, and NHS England’s work on FHIR standards all show what is possible when leadership and resources align.

The issue is not competence. It is whether the current digital architecture supports value wherever it is created, or only where it is institutionally sanctioned.

Provider Exceptionalism and Invisible Care

The NHS rightly prides itself on universal access. But that principle has, at times, drifted into a quieter assumption: that legitimacy flows primarily from being an NHS provider. This form of provider exceptionalism is rarely explicit, yet it shapes funding, measurement, and trust. The result is that large volumes of effective care, particularly in mental health, addiction, and long-term conditions, occur outside the system’s digital field of vision.

What cannot be seen cannot be benchmarked. What cannot be benchmarked cannot improve. And what is not digitally visible increasingly may as well not exist.

The Case for Controlled Integration

The NHS could reasonably argue that opening digital infrastructure to non-commissioned providers creates risks: fragmentation of records, data governance failures, and diversion of scarce resources from frontline care. These are serious concerns.

But they argue for designed openness, not architectural closure.

Accreditation requirements such as CQC registration, information governance certification and clinical governance standards can act as entry conditions for interoperability. Robust consent, privacy, and security frameworks can extend beyond NHS boundaries, as demonstrated in Australia, Canada, and parts of Europe.

Rather than wholesale reform, bounded interoperability pilots with explicit safeguards and evaluation criteria could generate evidence about what works, and what risks actually materialise.

AI Will Make These Choices Durable

This is where digital and innovation leaders should pause.

Artificial intelligence does not democratise health systems by default. It amplifies existing structures. AI systems learn from the data they can access. They optimise what is visible and reinforce what is institutionally legible.

In a closed digital architecture, AI will become exceptionally good at optimising hospital flows, acute episodes, and institutional performance - because that is where the data lives. Care delivered outside those boundaries will not merely be overlooked; it will be optimised around.

AI-driven referral optimisation cannot route patients to the highest-quality provider if that provider is digitally invisible. Federated learning approaches that could improve addiction treatment across NHS and non-NHS settings cannot function without interoperability, regardless of algorithmic sophistication.

Invisibility, once a manageable oversight, becomes a compounding disadvantage. AI makes architectural choices sticky - and resistant to later correction.

The critical window is now, before AI deployment at scale. Once systems are optimised on partial data, incentives to broaden visibility diminish. The system already “works,” at least by its own metrics.

Constraints, and Leadership

Any discussion of NHS digital architecture must acknowledge reality: chronic underfunding, intense political scrutiny, legacy IT estates, workforce shortages, post-pandemic demand, and the long shadow of past data-sharing failures. Australia does not face these pressures at the same intensity.

But constraint does not negate choice. It sharpens it.

A tangible step digital leaders could take now is establishing structured pathways for CQC-regulated non-NHS providers to participate in tightly governed interoperability pilots - not full openness, but experimental openness where learning is the objective.

The deeper question is whether we are willing to design digital systems that recognise value wherever it is safely created, rather than only where it has always lived.

A Final Thought

In a decade, we may look back and realise we conflated centralisation with safety - and assumed interoperability could only be engineered top-down.

Digital health will not save the NHS by digitising existing structures alone. It requires selective, governed openness: architecture that extends beyond institutional boundaries while preserving the accountability that makes the NHS trustworthy.

AI will not rescue us from this choice. It will simply reveal, with brutal clarity, whether we were prepared to make it.

Join the conversation at the HETT Leaders’ Summit

These questions about openness, legitimacy, and digital architecture are not theoretical. They sit at the heart of how health systems evolve under pressure. Pia Clinton-Tarestad will explore these themes further at the HETT Leaders’ Summit in the session Global Lessons, National Leadership: What the World Gets Right and What the NHS Must Do Next.

Drawing on international experience and real-world system design challenges, the session will examine how different models of digital transformation shape outcomes, where the NHS’s strengths truly lie, and what leadership decisions will matter most as health systems become increasingly data and AI driven. If you are grappling with how the UK can learn from global approaches while protecting its core values, this is a conversation worth joining.