Can Technology Save the National Health Service?
2025-08-11
The UK’s National Health Service (NHS) has published a 10-year plan built on three “past-to-future” pillars: shifting from hospital to community care; from sickness to prevention; and from analogue to digital. While I can’t speak to the first two, I have been reviewing the third.
Technologically, this breaks down into three key elements: a new NHS App consolidating public-facing services and information; a single source of patient, health, and treatment data; and an access platform for NHS providers (hospitals, specialists, GPs) that draws on both.
All are desperately needed. I love our NHS—its clinical staff are diamonds, and their achievements often miraculous—but as a system, it is broken. Our ability to meet the healthcare needs of our citizens is hampered by outdated technologies, fragmented systems, and inefficient processes, as the 2024 Darzi review confirmed.
The billion-pound question is whether better use of technology can make a genuine, material difference. The NHS isn’t unique in this—globally, healthcare lags other industries in innovation. Might better systems and processes reduce risk, improve outcomes, and release funds that can be put towards patient care?
I’d love to say technology can “fix” healthcare or provide a solid platform for the future. Unfortunately, history suggests otherwise. The NHS already has layers of technology; adopting something new is more of a problem than letting go of the old.
As both a technologist with hands-on healthcare experience and an industry watcher, I’ve seen many successes—and just as many failures. In the spirit of the NHS plan’s past-to-future theme, here are seven imperatives based on my own experience:
1. Prioritize change management over technological hope
Technology is powerful, but it doesn’t work in a vacuum. A couple of decades ago, data warehousing was hailed as transformative. Then it was the Cloud. Now it’s AI.
This isn’t just vendor marketing hype: systems integrators, consultants, and technology leaders have also believed that stating a goal will somehow make it happen. The NHS plan’s current statement of “We’ll make an app, that’ll sort things out” is another example.
Time and again, we’ve backed new technology waves without addressing the real challenge: change management. Transformation must lead; “digital” follows. After decades of “starting to see results,” it’s time to put change first–which means building stakeholder buy-in, proving by results, and the rest.
2. Build from the bottom of the stack up
Interoperability is key. The big-ticket items—the app, the data lake—mean little without the underlying glue: secure, standards-based data exchange, operational manageability, and upgradability.
Platform strategy needs to consider these needs at every level of the stack. At what level should data-in-motion privacy be guaranteed, for example – with tag-based microsegmentation at network level, or using container-level approaches?
A shared, policy-based architecture (based on common design patterns) across NHS organizations, integrators and indeed service providers, would enable delivery whilst assuring compliance, sovereignty, and risk.
3. Accept legacy data, and leave it where it is (for now)
The vision of a single, centralized NHS data source is compelling, but unrealistic. Instead, we should be building and maintaining an open, accessible library of known data assets through an intermediate “façade” layer.
Modern data fabrics enable us to leave data in place, manage and timestamp access, and gradually migrate what’s current and useful, like a slowly draining reservoir.
This avoids bottlenecks, reduces costs, and supports future-safe, API-driven data management.
4. Anticipate technical blockers before they happen
Even the best solutions can fail in deployment, often for predictable reasons. Operational complexity, such as that found in the NHS, can make even the best-planned delivery slow to a halt.
Equally, innovators can move too fast, missing something essential from the specification along the way. Once it is discovered, it is too late to address. Such issues are common: as software luminary Ed Yourdon (RIP) once said, “Death march projects are the norm, not the exception.”
Rather than assuming “this time will be different,” plan for worst-case scenarios. Put deployment risk on the board’s agenda and treat stakeholder skepticism as a valuable tool for mitigation, not a barrier to progress.
5. Replace proof-of-concept pilots with minimum viable products
As the quote goes, the NHS may have “more pilots than British Airways,” but such studies are designed to measure potential, not to have an impact or deliver change.
In addition, they can fall victim to prototype thinking: what starts as a test model can become a tool in active use, even though it was not built for robustness or scale.
A better approach is product-based thinking. Base units of delivery on stakeholder needs, not project goals, and deploy small, complete iterations that provide immediate value.
6. Lead with open vendor relationships, not lock-in
Suppliers of all descriptions—vendors, integrators and service providers—aim for long-term retention through extended contracts, licensing policies, and technical dependencies. Indeed, the NHS strategy paper talks about some contracts having 10-year terms.
This is expected, but can be ignored up-front by customers and prospects. It is for procurement best practice to protect against lock-in at every level, from cloud data export, to technological incompatibility, to skills development.
This also goes to the value of healthcare data, and the role of third-party providers and insurers. The NHS strategy covers public healthcare only, but its data platforms can be made available to private sector health delivery, at a negotiated cost.
7. Start with the patient, not the clinician
Finally, we need to address the fundamental missing piece of modern healthcare: patient centricity. Healthcare systems are built for clinicians, not patients. Information is delivered in medical jargon, and interfaces can assume high levels of digital literacy.
At a recent healthcare day organised by AWS, one clinical speaker said, “Until we confront these blind spots, our digital transformation will remain skin-deep.” And recent NHS studies are still provider, rather than patient-centric.
The NHS must move from provider-based data management models to ones which support how patients think and act. These include analogue options supported by digital means, for example with AI-driven interfaces.
Getting Ahead of the Curve
Healthcare lags other industries, but that gap offers an opportunity to leap forward. New technologies such as AI can help—but without addressing foundational issues, they risk becoming part of yet another missed opportunity.
This isn’t about needing more money. The NHS already receives around £200 billion annually—8% of UK GDP—with about a quarter set aside for potential litigation from care failures. Patients are suffering unnecessarily, when they could be part of the solution: for example, catalysing demand for better information could drive improvements.
It’s time to tackle the technological elephants in the treatment room, reduce waste, and focus spending on better capabilities, greater visibility, and significantly improved patient outcomes.
As both technologists and people, our future health depends on it.