CareNHS — Campaign for Responsible Leadership in NHS Wales
Patient Safety

The Billion-Pound Question Wales Cannot Afford to Get Wrong

Wales is about to decide how 3.2 million patients' health records will work for a generation. The organisation tasked with delivering it has failed at everything it has ever attempted. England has 39 million people on its NHS App. Northern Ireland deployed Epic across all 5 trusts. Wales produced two reports in nine months. The clock is ticking.

31 March 2026 · 18 min read

Somewhere in the next few years, Wales will make the most consequential technology decision in the history of its NHS. The Senedd election is on 7 May 2026. Every candidate should be asked: who will make this decision, and are those people capable of getting it right?

The question is whether Wales will have a modern Electronic Health Record: a single, integrated system where every clinician treating every patient can see the full picture — medications, test results, referrals, diagnoses, allergies, clinical notes — in real time, from any setting. The answer will determine how every hospital, every GP surgery, and every patient record in Wales works — for a generation.

Every comparable health system in the developed world is answering this question right now. England has committed £10 billion to NHS digital transformation and reached over 90% EPR coverage across its trusts. Northern Ireland deployed Epic across all five Health and Social Care Trusts under a contract originally valued at £275 million, serving 1.9 million people. Denmark built sundhed.dk, giving citizens access to their full medical history going back to 1977. Estonia built its initial national EHR for approximately €10 million — roughly €7.50 per citizen — that now covers 99% of health data.

Wales has the National Target Architecture programme. After nine months and an undisclosed sum paid to Channel 3 Consulting and Aire Logic, it has produced two documents: a Current State Report and an Initial Target State Report. Phase 2 runs to March 2026 and is expected to deliver a Strategic Investment Plan. No EHR vendor has been selected. No build-or-buy decision has been made. No timeline for a national EHR exists.

The person leading this programme is Ifan Evans, Executive Director of Strategy — the man who wrote the Welsh Government's digital health strategy, moved to DHCW to implement it, and now reports on whether his own plan is succeeding. His published qualifications include three degrees, none of which are detailed in publicly available biographies. He has no publicly documented experience in technology architecture or large-scale system delivery.

This is who is designing Wales's digital future.

What an EHR Actually Means for Patients

This is not an abstract technology debate. It is the difference between a clinician knowing what medications you are on and guessing. Between your GP referral reaching the hospital and disappearing into an unsupported system. Between a cancer diagnosis being made in days and being delayed by months because the screening system hasn't been updated in a decade.

Right now in Wales:

  • WPAS — the patient administration system used in almost every Welsh hospital — has, according to Employment Tribunal proceedings, been identified by a Health Board as a contributing factor in at least one patient death. Cardiff and Vale UHB refused to adopt it, running its own system instead. Health boards across Wales regard it with deep frustration.

  • WCCG — the system that carries every GP referral to secondary care — runs on technology unsupported for eight years. Sam Hall, Director at DHCW, acknowledged: "there are just a few people that know how that thing is put together." If it fails, every referral in Wales stops.

  • The Welsh Clinical Portal is primarily a data viewing and aggregation tool — it brings together data from other systems but is not a full EHR supporting comprehensive clinical workflows, order entry, or structured documentation. Dr Simon Barry, a Consultant Respiratory Physician, told the Senedd about "difficulties in linking some national databases" and "a lack of ownership and management of workstreams." A bowel screening system change was requested ten years ago and still hasn't happened.

  • The Royal Colleges warned in July 2025 that patients "regularly experience delays that lead to worsening health" due to digital fragmentation.

A modern EHR replaces all of this. When a patient walks into any hospital in a health system running Epic, the clinician sees everything — the GP records, the medications, the test results, the referral history, the allergies, the clinical notes from every previous encounter — on one screen, in real time. The prescriptions are checked against allergies automatically. The referral reaches the specialist without passing through an unsupported gateway. The screening invitation arrives on time.

This is not futuristic. It is what over 90% of English trusts already have. It is what 1.9 million people in Northern Ireland got in 2023-2025. It is what 3.2 million people in Wales do not have and — under the current trajectory — will not get.

The Right Idea, Destroyed by the Wrong People

The "once for Wales" philosophy — that digital health systems should be built nationally rather than procured separately by each health board — is not inherently wrong. It is, in principle, exactly what Estonia and Denmark did. A small nation builds shared infrastructure that serves everyone. You avoid fragmentation, avoid vendor lock-in, maintain data sovereignty, and build something you own and control.

Estonia proved this works. With 1.3 million people — smaller than Wales — Estonia built a national EHR on its X-Road interoperability platform for €10 million. It covers 99% of health data. It has been running since 2008. Citizens can see their full medical history, control who accesses it, and interact with every part of the health system through a single digital identity. Estonia's broader digital governance infrastructure is estimated to save the equivalent of 2% of GDP annually.

Wales has 3.1 million people. It has had a national digital health body since NWIS was established. It has had DHCW since 2021. It has spent approaching £200 million in identifiable programme costs — and cannot demonstrate a return on investment.

The "once for Wales" idea could have made Wales a European leader in digital health. Instead, it was entrusted to leaders who had never delivered technology at scale, who grew the workforce 78% without demonstrating what it achieved, who accumulated honorary professorships and industry awards while every programme fell further behind schedule, and who responded to criticism by dismissing whistleblowers and blocking the website of the campaign that documented their failures.

The philosophy was right. The execution destroyed it. And now the consequences are arriving.

The Health Boards Have Lost Faith

The consequence of DHCW's delivery failure is that the health boards — the organisations that treat patients every day — no longer trust the national body to build anything for them.

The Senedd's July 2023 joint committee found "serious concerns regarding the effectiveness of collaboration" with "few tangible examples of where this is occurring in practice." The RCGP described "frustration" among its members. Aneurin Bevan UHB called for "urgent action to address the governance around DHCW." Hywel Dda reported "significant integration and connectivity barriers." Cardiff and Vale refused to use WPAS entirely.

When DHCW was asked in oral evidence whose responsibility it was to ensure adoption of WCCIS — the flagship social care system that had consumed £42 million — the answer was: "I think that's a collective responsibility. I don't think it is a wholly DHCW responsibility."

In this environment, individual health boards are naturally looking at what England's trusts have done: buy a proven commercial EHR. The logic is understandable. Guy's and St Thomas' and King's College Hospital deployed Epic for £450 million. Manchester deployed Epic for £181 million. Northern Ireland deployed Epic across all five trusts for £275 million. Somerset and Dorset just signed a £222 million contract for a unified EHR across four trusts. These are working systems, in production, serving real patients.

If Welsh health boards follow this path individually, Wales will end up with multiple different commercial EHR systems — some Epic, some Oracle, some other vendors — exactly mirroring the fragmentation that plagues English hospitals. An Imperial College London study found that on 11 million occasions in a single year, English patients attended a hospital that could not access their records from a previous hospital visit. Digital fragmentation is not a theoretical risk. It is England's lived reality.

Wales would lose the one advantage its smaller size provides: the ability to do it once, do it right, and have every patient record in the country work together.

The False Binary

The current debate is framed as two options:

Option A: DHCW builds on WCP and WPAS. The national body extends its existing platforms into a full EHR. Health boards get a Welsh-built, Welsh-owned system that covers everyone.

Option B: Health boards buy individually. Each board procures Epic, Oracle, or another commercial system. Wales fragments but at least each board gets a working system from a proven vendor.

Both options are catastrophic under current leadership.

Option A fails because DHCW cannot deliver. Nine programmes under Level 3 intervention. Not one delivered on time. The CEO cannot demonstrate return on investment. The NTA programme — which is supposed to define the architecture — produced two documents in nine months. WPAS is hated by health boards and linked to patient death. WCP is a read-only portal, not a clinical workflow system. Asking DHCW to build a national EHR on this foundation, with this leadership, is asking for NPfIT at Welsh scale.

The UK already tried a centrally-imposed national EHR. It was called NPfIT. It cost £12.7 billion. It was dismantled after nine years without delivering. The Public Accounts Committee called it "one of the worst and most expensive contracting fiascos in the history of the public sector." The parallels with DHCW — top-down imposition, no clinician engagement, suppression of criticism, leaders who prioritised their own positions over delivery — are uncomfortable.

Option B fails because fragmentation destroys the national advantage. If each health board buys its own system, Wales gets the worst of both worlds: the cost of commercial EHRs (typically £80-200+ million per organisation in England, potentially exceeding £500 million across Wales) without the national integration that a smaller country should be able to achieve. Patient records won't flow between boards. The referral pathway from GP to hospital — already running on 8-year-old unsupported technology — would need to bridge between different commercial systems from different vendors. And Wales would be locked into those vendors for a decade or more.

Option C: Fix the Leadership, Then Decide

Neither Option A nor Option B works under current leadership. The real answer is to change the people who will make the decision before the decision is made:

  1. Recruit a new executive team through open international competition — CEO, Chief Digital Officer, Executive Director of Operations — with proven EHR delivery experience at national scale. Not from within NHS Wales. Not from the existing leadership pipeline. From the pool of people who have actually built and deployed comparable systems.

  2. Commission an independent EHR architecture review — led by external experts with no ties to DHCW, its suppliers, or the Welsh Government — to evaluate build, buy, and hybrid options transparently, with published findings.

  3. Impose a 90-day moratorium on NTA programme spending until the new leadership is in place and the architecture review is complete. No more undisclosed consultancy contracts producing documents that lead nowhere.

  4. Mandate HL7 FHIR UK Core as the interoperability standard for all Welsh health systems immediately, so that whatever decision follows, the data will flow between systems.

These are not aspirational proposals. They are the minimum preconditions for a decision of this magnitude to be made competently.

The choice is not build or buy. The choice is: fix the people who will make the decision, or accept that the decision will be wrong.

DHCW's Complacency

Inside DHCW, according to multiple sources who have contacted this campaign, there is a prevailing assumption: whatever Wales decides, DHCW will be fine. If Wales builds, DHCW builds it. If Wales buys, DHCW procures, deploys, and supports it. Either way, the 1,200 jobs are safe.

This assumption rests on DHCW being competent at procurement, deployment, and support. The evidence says otherwise:

Procurement: The RISP radiology contract (£47-56 million) was awarded to a supplier whose name has never been publicly disclosed, encountered implementation setbacks within two years, and no competitive tender notice was found on public procurement platforms. The Channel 3 / Aire Logic NTA contract has no disclosed value and no visible procurement trail. DHCW described its own consultancy spend as £757,000 while employing 23 off-payroll workers costing an estimated £1.5-4.5 million per year. The GP systems procurement forced Welsh practices onto one vendor's system, then forced them back to another — a double migration costing an undisclosed sum.

Deployment: Nine programmes under Level 3 intervention. OpenEyes: seven years behind schedule. LINC/LIMS: eight years, one lab partially live, original supplier terminated. WCCIS: eleven years, organisations trying to leave. NHS Wales App: "mired in delay, non-delivery." EPS: a decade behind England. Not one major programme delivered on time.

Support: WPAS linked to patient death. WCCG running on 8-year unsupported technology. eMPI outage that mixed up patient records across Wales. 21 system outages in seven months. The Cyber Resilience Unit: four people.

If this is the organisation that will procure, deploy, and support a commercial EHR from Epic or Oracle — systems that cost £100-450 million and require sophisticated programme management, clinical engagement, and technical infrastructure — the outcome is predictable. Cambridge University Hospitals deployed Epic in 2014 and the system became so unstable that ambulances were diverted, the CQC put the trust in special measures, and the CEO and CFO were forced to resign. That was with a competent technical leadership team that made mistakes. DHCW would be attempting the same thing with a leadership team that has never delivered anything at scale.

The Clock Is Ticking

Every month that passes without a decision widens the gap:

  • England: Over 90% of trusts have an EPR. £10 billion committed to digital transformation by 2028/29. 39 million NHS App users. The Federated Data Platform — controversial but operational — provides data integration across trusts.

  • Northern Ireland: Epic deployed across all five trusts by May 2025 for £275 million. A population of 1.9 million — smaller than Wales — with a unified EHR across the entire health system.

  • Scotland: NearMe video consultations stabilised at 33,000 per month. Active Digital Health and Care Strategy. NHS National Services Scotland providing shared components.

  • Estonia: 99% health data digitised since 2015. E-prescriptions at 99%. National EHR running since 2008. Annual savings equivalent to 2% of GDP.

  • Wales: Nine programmes failing. NTA programme in Phase 2 producing a Strategic Investment Plan. No EHR vendor selected. No build-or-buy decision made. No timeline. The CEO admits she cannot demonstrate a return on investment. The Cabinet Secretary describes the oversight framework as "complex, data-heavy, burdensome, lacks transparency and does not drive improvement."

Northern Ireland: 1.9 million people, unified EHR across all trusts. Wales: 3.2 million people, nine programmes failing, no EHR timeline.

Every month of delay is not neutral. Clinical systems deteriorate. Staff work around broken systems. Patients experience delays. The technology gap between Wales and every comparator widens. And the cost of catching up increases — because the vendors Wales would eventually need to negotiate with know that Wales has no alternative and no urgency.

What Wales Could Be

This is what makes the current situation not just a failure but a tragedy. Wales is perfectly positioned to build something world-class: 3.1 million people (larger than Estonia when it built its national EHR), seven health boards under a single government (no fragmented commissioner structure), an established national body with 1,200 staff, and a stated policy intent for "all core health services consolidated into a single all-Wales electronic health record application" by 2030.

The right size. The right structure. The right policy. The wrong leaders.

With a competent executive team — recruited internationally, with proven EHR delivery credentials — Wales could build a national system on open standards, own it, control it, and deliver it for a fraction of what fragmented individual procurements would cost. Our Technical Strategy Reform proposals set out exactly how: FHIR UK Core interoperability, a Welsh Health Data Exchange modelled on Estonia's X-Road, open APIs, and genuine vendor competition.

Under DHCW's current leadership, none of this will happen. The question is not whether it is possible. It is whether Wales will change the people who will decide.

The NPfIT Warning

The United Kingdom has been here before. In 2002, the government launched the National Programme for IT — the most ambitious health IT programme in the world. It was going to give every NHS patient a unified electronic record. It was centrally designed, centrally procured, centrally imposed.

It cost £12.7 billion. It was dismantled after nine years. The Public Accounts Committee called it "one of the worst and most expensive contracting fiascos in the history of the public sector." NAO found that by March 2012, only £3.7 billion of a forecasted £10.7 billion in benefits had been realised.

The reasons it failed read like a description of DHCW:

  • Top-down imposition without clinician engagement. Clinicians felt "devalued, marginalised, and ignored." DHCW publishes zero whistleblowing data and has allegedly dismissed the senior technologists who raised concerns.
  • Leaders who didn't understand technology. NPfIT was led by administrators and consultants, not by people who had delivered technology at scale. DHCW's executive team has no published track record of technology delivery.
  • Suppression of criticism. The government was described as showing "apparent reluctance to audit and evaluate the programme." DHCW blocks this website from NHS Wales networks and refuses FOI requests about its spending.
  • Culture of optimism over realism. NPfIT consistently overpromised and underdelivered. The Senedd warned DHCW against "over-optimism and a focus on celebrating successes at the expense of realistically assessing what needs to be achieved" — in 2023. The warning was ignored.

The lesson of NPfIT is not that national health IT is impossible. It is that national health IT led by the wrong people, with the wrong culture, and the wrong accountability is catastrophically expensive. Wales risks reproducing several of the conditions that caused NPfIT to fail, at a scale where the damage — while smaller in absolute terms — would be proportionally devastating.

The Decision That Cannot Wait

The Senedd election is on 7 May 2026. Every candidate should be asked:

Do you understand that the EHR decision is the single most consequential technology decision Wales will make in the next decade — and that the organisation responsible for delivering it has failed at every programme it has attempted?

Do you support replacing DHCW's leadership with executives recruited through open international competition — people who have actually delivered national-scale health technology — before the build-or-buy decision is made?

Will you ensure that the EHR decision is made transparently, with published options appraisals, independent technical assessment, and genuine clinical engagement — not behind closed doors by an executive team that marks its own homework?

The clinicians of Wales deserve an EHR that works. The patients of Wales deserve records that follow them. The taxpayers of Wales deserve a decision made by people who know what they are doing.

Wales has a window. The right size, the right structure, the right policy intent. What it lacks — the only thing it lacks — is leadership that can deliver. Fix that, and everything else becomes possible. Fail to fix it, and Wales will either build another failed national system at enormous cost, or fragment into a patchwork of commercial vendors that destroys the national advantage forever.

This is the billion-pound question. And the people currently responsible for answering it have never delivered a programme on time.

Right of Reply: DHCW was invited to respond prior to publication. No response has been received.

Source Note

EHR market data from 6B Health, Digital Health, and Signify Research (February 2026). NPfIT costs from NAO, PAC, and IEEE Spectrum. Epic and Oracle implementation costs from Digital Health and HTN. Estonia data from e-Estonia and Healthcare IT News. Denmark data from Healthcare Denmark and sundhed.dk. Northern Ireland from Digital Health. Welsh data from DHCW published strategy (2024-2030), Senedd Joint Committee Report (July 2023), Welsh Government Digital and Data Strategy (2023), DHCW Level 3 Escalation Framework (May 2025), and DHCW Public Accountability Meeting (January 2026). Imperial College fragmentation study published in BMJ Open. BBC EPR patient safety investigation (2024). All DHCW programme failure data from sources cited throughout this site.

What You Can Do

This evidence exists because someone looked. You can help make sure it leads to change:

  • Write to your MS — ask what they will do about DHCW's leadership before the EHR decision is made
  • Share this page — the more people who understand what is at stake, the harder it is to get wrong
  • Submit what you know — if you have evidence about the EHR debate inside DHCW or the health boards
  • Support the FOI campaign — 57 Freedom of Information requests targeting DHCW's hidden data

Related pages: