Introduction
The most ambitious digital transformation agenda this country has ever set for its health service is now in place.
Digital by default. Single Patient Records accessible across the entire NHS. AI-assisted diagnostics. The majority of outpatient care delivered outside hospital settings by 2035. Five transformative technologies — data, AI, genomics, wearables, and robotics — being pursued simultaneously across one of the largest and most complex organisations on earth.
I have read the NHS 10-Year Plan carefully. I think the vision is admirable. The direction of travel is the right one. The conditions that made previous large-scale NHS digital programmes difficult — fragmented systems, poor data quality, underfunded infrastructure, a workforce without digital tools adequate to their clinical needs — are at least now being named and addressed in a national ‘plan’ with genuine political backing.
I also think we have been here before. And I say that as someone who spent the best part of twenty years delivering complex digital programmes inside the NHS — not advising on them from outside, but doing the work.
I was part of the NHS National Programme for IT when it was the world’s largest civilian change programme. I led the delivery of digital infrastructure for mental health trusts, primary care, community health, and academic research. I held implementation assurance responsibilities at national level, running the process by which suppliers were assessed before being given authority to deploy. I have watched NHS digital transformation succeed in individual organisations and stall at system level, sometimes in the same programme, often for reasons that were entirely foreseeable.
So when I say what follows, it is not cynicism. It is the honest perspective of someone who wants this one to work — because the stakes are too high for another generation of promising starts and stalled programmes.
Here is what the NHS 10-Year Plan will actually take to deliver.
1. Delivery leadership, not just strategy leadership
The ‘plan’ is well-written. The vision is compelling. The case for change is made clearly and the strategic priorities are coherent. But a strategic plan is not a programme, and the distance between a well-constructed strategy document and a working system deployed at scale in a district general hospital is where most NHS digital transformation programmes have historically lost their way.
The NHS has always been pretty good at strategy. It is significantly less consistent at execution — at the unglamorous, iterative, problem-solving work of moving from a roadmap to a deployed solution in a complex clinical environment with resistant stakeholders, constrained resources, and legacy systems that behave unpredictably.
What the 10-Year Plan needs now — urgently, at every level from individual trust to ICB to national programme — is experienced delivery leadership. People who know the difference between a Gantt chart and a delivery plan. People who can manage suppliers who overpromise and under-deliver. People who understand clinical environments well enough to design change management approaches that work with clinicians rather than around them. People who can hold a programme on course when the original timeline was wrong and the stakeholders are getting impatient.
That talent exists in the NHS ecosystem. It is also in short supply, and it tends to be concentrated in the organisations that already perform well. The national programme infrastructure needs to actively invest in finding, developing, and deploying it — not assume it will emerge organically from the existing leadership base.
The lesson from NPfIT: the deployments that succeeded did so because they had experienced delivery leadership at local level as well as national governance. The deployments that struggled had strong strategy and weak execution. The ratio of investment between those two things should reflect that reality.
2. An honest conversation about legacy infrastructure
You cannot build a digital-first NHS on analogue foundations. That sentence is not a metaphor. It is a description of the literal technical challenge that sits underneath every ambition in the 10-Year Plan.
A significant proportion of NHS trusts are running clinical and operational systems that were old when I started working in this sector. Some of the patient record systems still in use in acute trusts were deployed in the 1990s. Interoperability between those systems is not a configuration problem — it is an architectural one, and in some cases the systems simply cannot share data in the ways that a federated data model requires.
The plan acknowledges this. The commitment to a Federated Data Platform is a serious and substantive response to a problem that has constrained NHS digital capability for decades. But acknowledging the infrastructure problem and funding its remediation at the scale required are two different things. Single Patient Records only work if the underlying patient data is clean, consistently structured, and governed in ways that allow it to flow across organisational boundaries with appropriate consent and security controls. That is not a technology project. It is a data management programme of considerable complexity and duration, and it needs to be resourced and governed as one.
The governance model that sits underneath the Federated Data Platform — who owns what data, who decides what can be shared with whom, what happens when a trust’s data governance approach conflicts with a national standard — needs to be as carefully designed as the technology itself. I know from direct experience on the UK-CRIS programme at Oxford, which built a platform for six million NHS patient records, that the information governance work is not a precondition for the technical work. It is the technical work, in very large part. Getting that right at the scale of the entire NHS is a multi-year programme in its own right.
The lesson from two decades of NHS digital delivery: the trusts that have the cleanest, best-governed data will implement the 10-Year Plan’s ambitions fastest. The trusts that have not invested in data quality and governance infrastructure will find that every new system they implement surfaces the same underlying problems. The plan needs a dedicated data readiness programme — not as a precursor to digital transformation, but as a core component of it.
3. Change management as a first-class investment
The 10-Year Plan will put new technology into the hands of a workforce that is already working under extraordinary pressure, in many cases still carrying the weight of the pandemic, and in some cases still carrying a justified scepticism about digital transformation programmes inherited from the NPfIT era.
Technology does not transform healthcare. The behaviour of the people who use technology transforms healthcare. That distinction — between deploying a system and achieving clinical adoption of it — is where the majority of NHS digital transformation investment has historically been mis-calibrated.
On the NPfIT, the single clearest predictor of whether a deployment would succeed or struggle was not the quality of the technology. It was the quality of the clinical engagement that preceded go-live. The trusts that had invested in genuine clinical leadership of their deployment — where the clinical champion had been involved in configuration decisions, where the training was designed by clinicians rather than by IT, where the go-live timeline had been set by clinical readiness rather than by supplier contract — achieved adoption rates and satisfaction levels that the trusts without that investment could not approach.
The same principle applies at the scale of the 10-Year Plan. Every pound invested in a new diagnostic AI system needs a matching investment in the radiologist or the GP who will decide whether to trust it. Every new patient record system needs a workforce change programme that is designed around the clinical workflow, not around the system’s data model. Every new digital service needs a patient engagement process that goes beyond a consultation document.
Change management has historically been the first item cut when NHS programme budgets come under pressure. The 10-Year Plan needs to treat it as non-negotiable — because a system that no one uses is not a digital NHS. It is an expensive lesson.
The lesson from 29 NHS Trust deployments: the cost of under-investing in change management is not paid at go-live. It is paid over the following eighteen months, in low adoption rates, clinical workarounds, support ticket volumes, and the reputational damage to digital transformation programmes that makes the next deployment harder to achieve. Front-load the investment. It is cheaper.
4. Governance that holds across organisational boundaries
Some of the most intractable challenges in NHS digital transformation are not technical. They are jurisdictional.
Data sharing between trusts. Integration across Integrated Care Boards with different priorities and different levels of digital maturity. Interoperability between systems procured by different organisations under different frameworks at different points in time. These are governance problems as much as technology problems, and they cannot be solved by better technology alone.
The ICS structure created by the Health and Care Act 2022 was designed in part to address the fragmentation that had hampered previous NHS digital programmes. It has improved some things. It has also created new complexity — new organisational boundaries, new accountability structures, new questions about which organisation is responsible for what in a shared digital infrastructure.
The clinical research data platforms I was involved in at Oxford showed clearly that cross-organisational data governance, done properly, is achievable — but it requires every organisation in the network to agree, explicitly and in detail, on data standards, consent frameworks, security models, and the processes for handling disputes and breaches. Arriving at those agreements takes time, senior commitment, and legal resource. It is not work that can be delegated to a programme manager with a spreadsheet. It requires sustained executive attention at every level.
The 10-Year Plan’s ambitions for federated data and cross-organisational care pathways are achievable. But they will only be achieved if the governance architecture is designed with the same rigour and the same resources as the technical architecture — and if that governance work starts before, not after, the technology procurement.
The lesson from federated NHS data delivery: governance that is designed retrospectively to fit a technology that has already been built will always be weaker than governance that shaped the technology from the outset. The sequence matters. Technology last, not technology first.
5. Sustained political and executive sponsorship
This is, in my view, the most significant risk the 10-Year Plan faces. And I want to be direct about it, because it is also the factor that is least amenable to programme management solutions.
Ten years is a very long time in NHS politics. Secretaries of State change. NHS chief executives change. ICB leadership changes. The priorities that dominate a board’s attention in 2025 will not be the same priorities that dominate it in 2030. The programmes that survived the NPfIT era intact were the ones whose sponsors stayed engaged — not just at launch, not just at the annual progress review, but through the difficult middle period of any complex programme when things are not going as planned, when stakeholders are losing confidence, and when the path of least resistance is to restructure, relaunch, or quietly de-prioritise.
The history of large-scale NHS digital initiatives is, in significant part, a history of programmes that had excellent sponsorship at inception and inadequate sponsorship in execution. The NPfIT had Ministerial backing when it launched. What it needed — and what it did not consistently have — was sustained, active, visible sponsorship at regional and local level through the long deployment period. Where that sponsorship existed and held, the programme delivered. Where it evaporated, the programme stalled.
The structural answer to this problem is to build sponsor accountability explicitly into the governance of the 10-Year Plan — not as a statement of commitment at launch, but as an ongoing governance requirement with measurable expectations at every level. Executive sponsors who disengage should be replaced. Programmes that lose board-level support should be escalated, not quietly shelved. The political environment should be managed actively, with regular reaffirmation of commitment from the key sponsors, rather than being treated as a stable backdrop against which delivery proceeds.
The lesson from thirty years of complex NHS programmes: the single clearest predictor of whether a multi-year NHS digital programme will sustain momentum through its difficult middle period is not the quality of its technology or its programme management. It is the quality and consistency of its executive sponsorship. Protecting that sponsorship — and having a plan for what happens when it comes under pressure — is a delivery task, not a political one.
The ambition is right. The delivery has to match it.
None of what I have set out here is a reason not to pursue the 10-Year Plan’s ambitions. On the contrary. The case for a genuinely digital NHS — for a health service in which every clinician has access to the information they need to make good decisions, every patient can see their own records and to some extent manage their own care, and the extraordinary power of health data, which can be used to improve population health and accelerate research — is overwhelming. The cost of not doing this, measured in clinical errors avoided and lives extended and a health service that can be sustainable in the long term, is incalculable.
But the NHS has been here before. It has launched programmes with ambitions comparable in scale and importance to those in the 10-Year Plan, and it has seen them stall — not because the technology failed, but because the delivery disciplines failed. Because strategy was treated as a substitute for programme management. Because legacy infrastructure was acknowledged rather than addressed. Because change management budgets were cut. Because governance was designed after the fact. Because sponsorship faded.
The 10-Year Plan deserves better than that outcome. So does every patient who will wait longer than they should, and every clinician who will need to make a worse decision than they could, if this programme does not deliver.
We are currently working with NHS clients on digital transformation strategy and programme leadership aligned to this agenda. If you are working on the delivery side of the 10-Year Plan — at trust, ICB, or national level — we would welcome a conversation about what that looks like in practice.
