The Perils of Thinking Big in Digital Government

Donald Trump quotation, accessed via Donald Trump Quotes — Inspiration Boost

The world’s most powerful figures love to ‘think big’. Whether it’s business moguls reshaping global markets or governments trying to digitise entire nations, scale is often seen as strength. But in digital transformation, being too big and too top-down can be a fatal flaw.

Recent events, specifically in the US with the formation of Elon Musk’s DOGE by executive order, evidence how influential tech figureheads shape government digitisation programmes through their big ideas and big influence. But this is not a new phenomena, even back in 2002, Bill Gates “hooked” then UK Prime Minister Tony Blair with a nationwide mission to digitally transform the entire NHS.

10 years and £10bn later, that mission, named the UK’s National Programme for IT, or NPfIT was unceremoniously dismantled. It was the largest healthcare IT project in the world at the time, inspired in part by Gates’ vision of digital economies and backed by Tony Blair’s government with a ‘big bang’ approach. But thinking big was its downfall.

Why does this matter today?

The NHS is still repeating similar mistakes and as the UK and other nations continue to digitise at scale, the lessons from NPfIT are more relevant than ever. This blog explores how the NPfIT’s massive scale and top-down approach led to its failure, why the NHS is still trapped in the mindset of thinking big and what can be learned from Estonia’s X-Road. By examining this different approach, insights can be gained on how the UK could approach digitising the NHS moving forward.

The National Programme for IT — a brief history

Launched in 2002 with a £6.2 billion budget, the NPfIT aimed to modernise healthcare in the NHS. This would be done by implementing integrated electronic patient records, online ‘choose and book’ services, computerised referral and prescription systems and supporting network infrastructure.

Taken from: 2014 — The National Program for IT in the NHS — A Case History.pdf — Google Drive

From 1992 up until the initiation of the NPfIT in 2001, successive UK Governments had attempted to put in place a mix of strategies and targets with the aim of digitising NHS systems. These included several, forward thinking principles that still underscore digitalisation of government 30 years on: that information should be secure and confidential, and the creation of a minimum viable dataset that should be maintained and be accessed on a strictly evidenced, need-to-know basis. However, after delays, increasing costs, opposition and product failure, the programme was discontinued by the coalition government. The total cost, including contract exit fees, reached £12.7 billion by 2013/14.

The National Programme for IT — size and approach

The NPfIT took an ‘everything, everywhere, all at once’ approach. The programme was centrally managed from the outset with a new department, NHS Connecting for Health, overseeing a top-down, fait accompli waterfall delivery.

One of its many aims was to rollout national electronic patient record (EPR), electronic prescribing and appointment booking systems that were being tendered for and built by external suppliers simultaneously.

To do this, it split the NHS geographically into five huge and unwieldy clusters. IT enterprise solutions were procured simultaneously across these. Policies were developed, and developers procured to ‘take the whole thing away…based on assumptions’ and build.

Ironically, in many ways its huge scale proved restrictive. Due to its size, smaller companies were locked out of procurement tenders and local trusts lost autonomy and flexibility in system design and implementation. Efforts to find out what users needed were effectively obstructed as solutions were not tested by doctors and nurses. By 2010, this approach had alienated local trusts, with no buy-in from those expected to implement the system.

In 2011, Sir David Nicholson admitted the programme’s scale was its downfall, recognising the need for “restoring local control over decision-making and enabling greater choice for NHS organisations”. While parts of the database remained, record-keeping and booking systems were handed back to local trusts.

Continued impact

For better or worse, the failure of the NPfIT as a big, central system triggered a ‘diversification revolt’. Today, different NHS trusts use different Electronic Patient Records (EPRs). Some can share data with partner trusts, others cannot. A small number still rely on paper records, despite a £2bn Frontline Digitisation Programme to ensure all trusts have EPRs by 2025, now pushed back to 2026. Ultimately, this fragmentation leads to patient and admin errors that are inefficient but also risk patient safety. The 2018 WannaCry cyberattack exposed these vulnerabilities, locking NHS trusts out of records and delaying care.

Local NHS trusts also make the mistake of thinking big

While the NPfIT’s failure led to greater localisation, the same mistakes still persist in individual NHS trusts. Local doesn’t necessarily mean small. The largest ‘local’ trusts each serve millions of UK residents. Often, their independent EPR launches are treated as all-or-nothing events, assuming bigger means better. Temporary digital teams (usually staffed by the enterprise system supplier) are parachuted in for go-live days, but it’s uncertain what happens next. Who ensures systems evolve, feedback is integrated and improvements made?

As Jen Pahlka says:

“Treating software as just another commodity overlooks the fact that mission-critical software cannot simply be bought the way you buy a truck or even a building. It’s an integral part of the service you provide, and that service and the environment in which it operates are dynamic.”

Yet, local NHS digital environments remain stubbornly static. IT teams are siloed, rarely embedded or perhaps even interacting with clinical staff. They are often physically inaccessible, housed in buildings not on hospital sites. User research and design roles are largely absent. Digital tools such as EPRs that are integral to smooth service delivery remain outsourced and detached from local user needs.

The result? Hyper-local procurement of non-local digital systems, the worst of both worlds. Trusts duplicate procurement efforts while failing to adapt systems to local workflows. As Eaves, Pope and McGuire recognise, “the siloed nature of government means that the purchasing and building of technology occur[s] in parallel, even when the requirements or business operations [are] similar or even identical”. Across the NHS, 32 different EPR systems are used. One of the largest EPRs, EPIC, is used independently and in a generally disconnected manner by 13 different trusts, with a catchment population totaling nearly 8 million.

EPIC’s patient-facing portal, My Chart, allows patients to download referral and treatment letters, view appointments and access test results, all critical functions for managing health. Yet, three of the four most common complaints about MyChart relate to poor design. In healthcare, clarity isn’t just a convenience, it’s a necessity. When information is confusing or inaccessible, the risks are serious and could have real implications on access to and delivery of care.

Fig 1: A screenshot of negative highlight summary of My Chart reviews (349 total) on Apps Hunter taken from https://appshunter.io/ios/app/382952264/reviews?sort=0

Hyper-localisation and segregation of systems also means NHS trusts can’t share patient data easily, requiring separate data-sharing agreements before information can flow. This is out of sync with modern life, patients move and expect healthcare to follow them. Instead, they’re left navigating a patchwork of disconnected systems. Or, as one My Chart reviewer puts it:

Fig 2: screenshot of latest My Chart review on Apps Hunter https://appshunter.io/ios/app/382952264/reviews?sort=0

How can the NHS strike the right balance and create a seamless, efficient system while empowering local teams to adapt and improve it? Is there a way to ensure national consistency without sacrificing the flexibility needed for local innovation, all while providing a better experience for patients?

A global example: Estonia’s X-Road

One of the best examples of a system that balances central infrastructure with local flexibility is Estonia’s X-Road. This communication framework enables secure data exchange between government agencies and private companies, built on a national identification layer. Instead of each organisation creating its own isolated databases and applications, services must comply with shared data standards and integrate with X-Road. This means developers can focus on user experience while benefiting from a unified, interoperable infrastructure that fosters co-production and cross-sector collaboration.

A similar approach in the NHS could provide both a scalable platform and hyperlocal design, avoiding a one-size-fits-all system while ensuring a secure foundation for data sharing. This could transform coordination between health and social care services. For example, older patients who have suffered a fall often cannot be discharged until local councils and support services have arranged at-home care. The delays cost time, money, and hospital capacity when many patients would rather be at home. A shared platform could allow real-time data exchange, speeding up the process and ensuring better patient outcomes.

To sum up

Digital transformation fails when it’s too big, too fast and too top-down. The NHS’s NPfIT approach is a prime example. It led to wasted resources, resistance and eventual failure. In response, the NHS decentralised decision-making, but this shift hasn’t solved the core issue. Many trusts procure large enterprise systems in a waterfall approach, often with poor patient interfaces. Even when trusts have the flexibility to refine these systems, they may lack the internal capacity to do so, as digital design and user research, despite being widely recognised as essential, are not always embedded in practice.

While decentralisation offers flexibility, reinventing the wheel each time leads to duplication and waste. Estonia’s X-Road provides a possible model; a scaled digital infrastructure that enables interoperability, while providing local adaptation and continuous improvement. This platform approach could help the NHS avoid the pitfalls of both centralisation and fragmentation, ensuring it doesn’t fall into the trap of being too big and too top-down while still fostering collaboration and innovation at a local level.