# Blueprint for World-Class Digital NHS in Wales — Full Content
URL: https://bluenhs.org/

_Full plain-text content of every page on bluenhs.org, concatenated for LLM consumption. Authored by Dr Rafal Bergman, former Chief Technology Officer at Digital Health and Care Wales (DHCW). Licensed CC BY 4.0. See /llms.txt for the navigation index._



# A Blueprint for World-Class Digital NHS in Wales
URL: https://bluenhs.org/

_Wales deserves world-class digital health infrastructure. 3.16 million people depend on it. 611,000 are on waiting lists. This systems dynamics analysis maps why DHCW fails, how failed leadership protects itself, and what would fix it._


# Facts
URL: https://bluenhs.org/facts/

_Headline figures from the Blueprint analysis. DHCW's £200M annual budget; £0.5M of quantified value over five years (Finance Director's own figure); 611,000 on NHS Wales waiting lists; 18 months zero PDC corrective actions; all 7 Meadows traps active; 11 reinforcing feedback loops; the £3–10B cost of inaction over five years._

Every marquee figure cited across the Blueprint analysis, organised for citation. Sources are named; dates are absolute. The page is licensed CC BY 4.0 — reproduce with attribution.

## Money

| Figure | What it is | Source | Date |
|---|---|---|---|
| **~£200M** | DHCW's annual budget reached this level in 2025–26 | DHCW board financial reports | 2025 |
| **~£600M** | Cumulative Welsh Government funding to DHCW over five years | WG allocations to DHCW, 2020–21 to 2024–25 | 2025 |
| **£0.5M** | Quantified delivered value across the full five-year period — the Finance Director's own figure | DHCW Finance Director (2022 Finance Director of the Year) | 2025 |
| **83p** | Return per £1,000 of Welsh Government funding | Derived from above | 2025 |
| **~£1.25B** | DHCW's vendor contract portfolio value | DHCW board contract registers | 2024–25 |
| **£100–150M / yr** | Annual direct waste in the current model | This analysis ([Cost of Inaction](/blueprint/cost-of-inaction/)) | May 2026 |
| **£3–10B over 5y** | Downstream cost to NHS Wales of continuing the current model. Patient harm is a separate ledger and not monetised | This analysis ([Cost of Inaction](/blueprint/cost-of-inaction/)) | May 2026 |
| **£5–15M** | Cost of implementing the full six-intervention Blueprint | This analysis ([The Blueprint](/blueprint/)) | May 2026 |

## People

| Figure | What it is | Source | Date |
|---|---|---|---|
| **1,263** | DHCW staff headcount, up from 675 — an 80% increase | DHCW board headcount reports | 2024–25 |
| **+80%** | DHCW headcount growth while delivery worsened | DHCW board headcount reports | 2024–25 |
| **611,000** | People on NHS Wales waiting lists | NHS Wales statistics | 2025 |
| **30,000+** | Welsh patients waiting cross-border in NHS England because Welsh systems cannot cope | NHS England cross-border activity statistics | 2025 |
| **3.16M** | People that depend on NHS Wales | NHS Wales population coverage | 2025 |

## Delivery

| Figure | What it is | Source | Date |
|---|---|---|---|
| **0 / 9** | DHCW programmes on time | DHCW programme delivery reports ([Tragedy of the Commons](/traps/tragedy-of-the-commons/)) | 2024–25 |
| **12+ months** | DHCW at Level 3 enhanced monitoring — the first NHS Wales body ever held there. Escalated to Level 4 in 2026 | Welsh Government 5-level escalation framework | March 2025 – 2026 |
| **18 months** | The Performance and Delivery Committee generated zero corrective actions across eighteen consecutive months | DHCW PDC minutes ([Captured Governance](/diagnosis/l11-oversight-obstruction/)) | May 2024 – May 2025 |
| **2 of 47** | Phase One of Level 4 milestones missed (LIMS and WRISTS) | DHCW board, March 2026 | March 2026 |
| **£32.9M + £13.1M** | DPIF revenue + capital remaining unallocated as of March 2026 | DHCW board minutes; Accountable Officer letter | March 2026 |
| **8+ years** | Time WCCG has run on technology unsupported by its vendor, against repeated technical-staff warnings | DHCW technical risk register ([Drift to Low Performance](/traps/drift-to-low-performance/)) | Ongoing |

## Governance

| Figure | What it is | Source | Date |
|---|---|---|---|
| **7 / 7** | Donella Meadows's system traps active at DHCW simultaneously. No documented precedent in public-sector digital delivery | This analysis ([Seven Traps](/traps/)) | May 2026 |
| **11** | Reinforcing feedback loops identified, in two clusters — five drive delivery failure, six protect that failure from correction | This analysis ([The Diagnosis](/diagnosis/)) | May 2026 |
| **51** | Board approvals without scrutiny documented in the knowledge graph (incl. a £20M Kainos framework the Chair admitted he had not looked at, and a £226M Microsoft Enterprise Agreement passed in a single sentence) | Knowledge graph of DHCW board minutes | 2021–2025 |
| **33** | DHCW directors declaring 'nil' on conflict-of-interest declarations while holding undisclosed UWTSD Professor of Practice titles (Thomas 10, Evans 10, Hurle 8, Hall 5) | DHCW board minutes, declarations of interest ([Loyalty Selection](/diagnosis/l8-loyalty-selection/)) | 2021–2025 |
| **4 months** | Audit Wales declared "good governance, stable and cohesive board" four months before DHCW was placed under Level 3 escalation | Audit Wales structured assessment, November 2024 | November 2024 |
| **3 successive** | NHS Wales CEOs drawn from a single health board (ABUHB), via Goodall and Paget | Public NHS Wales appointments record ([Loyalty Selection](/diagnosis/l8-loyalty-selection/)) | 2018–2024 |
| **December 2020** | Three executive directors received UWTSD Professor of Practice titles — four months before DHCW's founding board met | UWTSD honorary appointment register | December 2020 |

## Cited statements (verbatim)

> "We've always struggled because it's one-year funding. We should have worked it out by now, surely."
> — Ruth Glazzard (CFO), DHCW board, March 2026.

> "This persistent ambiguity, let's call it, which we usually phrase as 'we need to clarify roles and responsibilities'… PDC is an oversight and assurance function, not a programme management function."
> — Ifan Evans, Performance and Delivery Committee, 46 months after DHCW's founding.

> "Government was asking… why then are those programs rated red amber?"
> — Ifan Evans on Welsh Government pressure to soften RAG ratings, PDC November 2024.

> "I should have looked. I don't know how these appear on our website as contracts."
> — DHCW Chair, on the £20M Kainos framework approved without scrutiny.

> "Complex, data-heavy, burdensome, lacks transparency and does not drive improvement."
> — Cabinet Secretary Jeremy Miles, 2026, publicly describing the Welsh Government 5-level escalation framework — while continuing to operate it.

## How to cite

Bergman, R. (2026). *Facts — A Blueprint for World-Class Digital NHS in Wales.* https://bluenhs.org/facts/

CC BY 4.0. Free to reproduce, share, summarise, and adapt with attribution.


# The Welsh Government 5-Level Escalation Framework
URL: https://bluenhs.org/wg-escalation-framework/

_The escalation framework Welsh Government operates over NHS Wales bodies, from standard monitoring (Level 1) to special measures (Level 5). DHCW was held at Level 3 enhanced monitoring for 12+ months — the first NHS Wales body ever to reach it — and was escalated to Level 4 Targeted Intervention in 2026. The Cabinet Secretary publicly described the framework as 'complex, data-heavy, burdensome, lacks transparency and does not drive improvement'._

## The framework, in five levels

The Welsh Government operates a five-level escalation framework over NHS Wales bodies. Each level represents an increasing degree of central intervention; the levels are publicly documented and applied to health boards and Special Health Authorities including DHCW.

| Level | Name | Trigger | What it means in practice |
|---|---|---|---|
| **1** | Standard monitoring | Routine | The default operating state. Quarterly Joint Executive Team meetings; performance reviewed against the long-term plan. |
| **2** | Enhanced support | Sustained performance concerns | Welsh Government works with the organisation on a recovery plan. Additional support is offered. No additional formal sanction. |
| **3** | Enhanced monitoring | Significant or persistent failure | Formal increased oversight. Quarterly Integrated Quality and Performance Delivery (IQPD) meetings; documented improvement trajectory; named WG officer attached; independent input where required. |
| **4** | Targeted intervention | Failure at Level 3 has not produced recovery | A more directive posture. Welsh Government takes specific action — including directing leadership change in some cases, or imposing structural conditions. |
| **5** | Special measures | Failure at Level 4 has not produced recovery, or risk profile demands immediate action | The most intensive intervention available. Welsh Government effectively takes over key decisions. |

The framework is operated by the Welsh Government Director General for Health and Social Services and overseen by the Cabinet Secretary for Health and Social Services.

## DHCW's journey through the framework

DHCW's trajectory is the most documented case of the framework's failure to drive correction:

- **1 April 2021 – March 2025 — Level 1, standard monitoring.** DHCW operated at the default level from its establishment on 1 April 2021 — itself [a rebrand of NWIS](/about/why-this-blueprint/), the previous body, with substantially the same leadership team. The trajectory across those four years was a car crash in slow motion: programmes drifted, governance markers were rubber-stamped, technical capability eroded. Welsh Government oversight did not detect it. DHCW leadership did not detect it. When Level 3 escalation came in March 2025, it arrived as a complete shock to the DHCW executive — which is itself diagnostic. Almost four years of structural failure had produced no escalation above Level 1. The framework's balancing-loop signal — the part of it that should detect drift and trigger correction — had been silent for the entire period in which the conditions that produced Level 3 were forming.
- **March 2025 — Level 3 enhanced monitoring.** DHCW was placed at Level 3, the first NHS Wales Special Health Authority ever to reach it. The escalation followed sustained programme delivery failure across the portfolio.
- **March 2025 – March 2026 — twelve months at Level 3.** The IQPD framework operated as designed. Quarterly meetings occurred. Documentation was produced. No de-escalation followed. No structural change in DHCW leadership followed. The Performance and Delivery Committee continued to generate zero corrective actions across the period.
- **2026 — escalation to Level 4 Targeted Intervention.** Twelve months of Level 3 had not produced recovery. DHCW was escalated to Level 4.
- **March 2026 — public disowning of the framework.** The Cabinet Secretary publicly described the entire framework as "complex, data-heavy, burdensome, lacks transparency and does not drive improvement" — while continuing to operate it.
- **From April 2026.** The IQPD meetings — the formal vehicle for monitoring conduct under Level 3 — were due to be wound down and replaced by a new escalation meeting chaired by the Director General. As of the March 2026 board, two of forty-seven Phase One milestones had already been missed (LIMS and WRISTS); £32.9M of DPIF revenue and £13.1M of capital remained unallocated; the Accountable Officer had formally stated delivery was "not possible without confirmed DPIF allocation".

## Why twelve months of Level 3 produced no correction

This is the central diagnostic question for the framework — and it is answered by the Blueprint's [L11: Captured Governance](/diagnosis/l11-oversight-obstruction/). The escalation framework is a balancing loop in systems-dynamics terms: it is supposed to detect deviation and trigger correction. At DHCW the loop has been neutralised at every step:

1. **The information that reaches WG is filtered by DHCW.** Leadership obscures the information provided to Welsh Government's oversight bodies. This contributed to the departure of the Welsh Government GDS director from his role — the principal technical counterweight to DHCW inside WG. He has not been replaced.
2. **The independent expert is invisible.** Under the Level 3 framework an independent digital expert was appointed. That person has never been publicly identified — making it impossible to assess their independence, qualifications, or conclusions.
3. **Welsh Government is itself part of the failure conditions.** WG materially shaped DHCW's failure conditions: annual funding cycles; contradictory remit letters; a recruitment freeze imposed while delivery was accelerated; the £33M-to-£28M DPIF cut for 2024–25; late funding confirmations 25% through the financial year. WG cannot be both co-author of the failure and sole arbiter of the recovery. The Blueprint's [Intervention 6: Reform the Funder](/interventions/reform-the-funder/) addresses this directly.
4. **The framework has no enforcement mechanism shorter than its own ladder.** Level 3 produces no consequence on a defined timeline. Twelve months can pass without de-escalation, without dismissal, without structural change. The only escalation route is up the ladder — to Level 4, then Level 5 — and each step takes months. By the time Level 5 is reached, the organisation has been failing under monitoring for years.

## What "lacks transparency" means in this context

The Cabinet Secretary's own assessment of the framework — "complex, data-heavy, burdensome, lacks transparency and does not drive improvement" — names the failure modes:

- **Complex.** The IQPD process produces large volumes of paperwork. Reviewing it requires sustained attention from skilled staff. WG does not have the equivalent of an Audit Wales technical division to do so.
- **Data-heavy.** The framework produces telemetry but not insight. RAG status is reported. Action lists are produced. None of it converts into accountability.
- **Burdensome.** For the organisation being escalated, the framework imposes documentation overhead without surfacing genuine performance signals. Compliance with the framework becomes a substitute for delivery against it.
- **Lacks transparency.** Decisions about escalation, de-escalation, and Level 4 conditions are made through processes that are not externally observable. The Senedd does not have line-of-sight into the framework's internal deliberations. Audit Wales has limited reach.
- **Does not drive improvement.** Twelve months of Level 3 with no de-escalation is the evidentiary baseline for this conclusion.

The minister responsible for overseeing DHCW has publicly acknowledged that the oversight mechanism does not work — while continuing to operate it. This is the unusual structural condition that the Blueprint's diagnosis names as the captured-governance pattern.

## What a working escalation framework would look like

This is the design challenge the Blueprint's [Who Guards the Guardians?](/blueprint/governance/) addresses. The short version:

- **Distributed oversight.** No single actor — including Welsh Government — should be sole arbiter, because WG itself materially shaped the failure conditions. Multiple independent oversight roles with statutory backing.
- **Named experts, published assessments.** Where independent expertise is commissioned, the expert is named, the brief is public, and the assessment is published.
- **Time-bounded de-escalation criteria.** Each level has a documented exit condition with a defined timeline. Twelve months at Level 3 without movement should trigger an automatic structural response, not another twelve months.
- **Separation of commissioning and delivery oversight.** Welsh Government commissions the digital body; it should not also be the sole judge of its performance. The two functions need structural separation.

These design changes do not require new legislation in many cases — they require operational redesign of the existing framework. The Cabinet Secretary's own public assessment provides the political opening.


# References
URL: https://bluenhs.org/references/

_Hub for the reference resources on the site: marquee figures (Facts), named NHS Wales national systems (WPAS, RISP, LIMS, EPMA, eMPI, WCCG, WICIS, PSBA), the Welsh Government 5-level escalation framework, glossary, cross-cutting concepts, and the companion evidence site carenhs.org._

These pages are reference material — figures, systems, frameworks, terms, and concepts that appear throughout the analysis. Each is linked from the relevant feedback loop, system trap, or intervention page where it is discussed.

## [Facts](/facts/)

Every marquee figure cited across the analysis, with its source and the date it was observed. 19 entries across Money (8), People (5), Delivery (6), and Governance (7), plus verbatim quoted statements from the DHCW board, Welsh Government, and Audit Wales. Machine-readable: each figure is published as a Schema.org `Claim` and the page itself as a `Dataset`.

## [Named NHS Wales National Systems](/systems/)

One-page references for the eight national NHS Wales digital systems named across the analysis: **WPAS, RISP, LIMS, EPMA, eMPI, WCCG, WICIS, and PSBA**. Each entry gives what the system is, its current status, the documented patient-safety record where one exists, and where it is discussed in the diagnosis.

## [The Welsh Government 5-Level Escalation Framework](/wg-escalation-framework/)

The escalation framework Welsh Government operates over NHS Wales bodies, from standard monitoring (Level 1) to special measures (Level 5). DHCW's actual journey through it — Level 1 from 1 April 2021 to March 2025; Level 3 enhanced monitoring for 12+ months; escalated to Level 4 Targeted Intervention in 2026. The Cabinet Secretary publicly described the framework as "complex, data-heavy, burdensome, lacks transparency and does not drive improvement."

## [Glossary](/glossary/)

DHCW-specific terms and acronyms in one place: Level 3, WPAS, RISP, LIMS, DPIF, captured governance, Meadows's seven traps, once-for-Wales, and more. One-line definitions linked to fuller treatments. Published as a Schema.org `DefinedTermSet`.

## [Concepts](/concepts/)

Cross-cutting analytical concepts tagged across pages — governance-capture, oversight-degradation, level-3-escalation, executive-competence, vendor-dependency, and others. Each tag aggregates every page that addresses that concept. Useful for tracing a single thread across the analysis.

## [Evidence → carenhs.org](https://carenhs.org)

The companion site for the evidence apparatus. Knowledge graph spanning DHCW board and committee meetings (April 2021 – April 2026), with high-severity findings, contradiction registers, sanitisation evidence, and witness testimony. carenhs.org is where the evidence is documented; bluenhs.org is the structural analysis derived from it.

## Machine-readable indexes

For AI search and language-model consumption:

- [/llms.txt](/llms.txt) — navigation index for LLMs, conforming to the proposed [llmstxt.org](https://llmstxt.org/) standard.
- [/llms-full.txt](/llms-full.txt) — full plain-text content of every page on the site, concatenated for direct LLM consumption.
- [/sitemap.xml](/sitemap.xml) — XML sitemap.

The site is licensed [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — free to cite, summarise, and reproduce with attribution.


# Glossary
URL: https://bluenhs.org/glossary/

_Glossary: DHCW, Level 3, WPAS, RISP, LIMS, DPIF, captured governance, Meadows's seven traps. One-line definitions linked to fuller treatments._

Terms and acronyms used across this analysis, in alphabetical order. Where a fuller treatment exists on the site, the entry links to it.

### Brooks's Law

"Adding people to a late software project makes it later." Coined by Fred Brooks in _The Mythical Man-Month_ (1975) from his experience running IBM OS/360. Adding staff imposes onboarding burden on the productive people already there and grows coordination overhead faster than it grows output. Explains why DHCW's 80% headcount growth did not improve delivery. See [L1: The Hiring Trap](/diagnosis/l1-hiring-trap/).

### Cluster A

The five reinforcing loops that explain why DHCW's delivery fails. Structural loops that any national health IT body under comparable constraints would face some version of. See [/diagnosis/](/diagnosis/).

### Cluster B

The six reinforcing loops that explain why DHCW's failure is protected from correction. Not structural accident — an active self-preservation engine that intercepts every corrective mechanism before it reaches Cluster A. See [/diagnosis/](/diagnosis/).

### Connecting Care

The rebrand of WCCIS, adopted after the original deployment stalled. New name, same product, same delivery failure. See [L4: The Rebranding Escape](/diagnosis/l4-rebranding-escape/).

### DHCW

Digital Health and Care Wales. The Special Health Authority, established 2021, responsible for NHS Wales digital infrastructure. Successor to NWIS. The subject of this analysis.

### Digilugu

Estonia's national e-health record system, providing every Estonian citizen with a single longitudinal record accessible across providers. A working version of what WCCIS was meant to be. See [Alternatives](/blueprint/alternatives/).

### Donella Meadows

(1941-2001) Environmental scientist, MIT-trained systems analyst, and lead author of the 1972 Club of Rome report. Author of _Thinking in Systems_ (2008, posthumous) and the 1999 essay introducing the 12-Leverage-Points framework. The methodology of this site is hers. See [Methodology](/methodology/).

### eMPI

electronic Master Patient Index. DHCW's national patient identifier system. Has mixed up patient records in operational use.

### Feedback loop

A circular chain of cause-and-effect where a change in one element eventually comes back to affect that same element. _Reinforcing_ loops amplify change (vicious or virtuous cycles); _balancing_ loops correct toward a target. See [Methodology](/methodology/).

### Forrester, Jay

(1918-2016) MIT engineer, inventor of magnetic-core memory, and founder of the system-dynamics discipline through _Industrial Dynamics_ (1961), _Urban Dynamics_ (1969), and _World Dynamics_ (1971). Donella Meadows' teacher and collaborator. See [Methodology](/methodology/).

### Jay Forrester

See [Forrester, Jay](#forrester-jay).

### Leverage points

Donella Meadows' 1999 framework ranking twelve places to intervene in a system, ordered from shallowest (parameters) to deepest (paradigm). The insight: most reform effort targets the shallow end, where systems absorb change without altering behaviour. See [Methodology](/methodology/).

### Level 3 Enhanced Monitoring (and the WG 5-level framework)

The Welsh Government NHS Wales Escalation and Intervention Arrangements define **five escalation levels**, not three. Level 3 is the middle of the range, not the top. The levels escalate roughly as follows:

- **Level 1 — Routine arrangements.** Standard oversight. No concerns raised.
- **Level 2 — Enhanced monitoring.** Raised concerns; increased reporting cadence to Welsh Government; no formal intervention.
- **Level 3 — Targeted intervention (Enhanced monitoring with intervention).** Formal escalation; specific programme-level interventions; an independent digital expert appointed.
- **Level 4 — Significant intervention.** Board-level action; independent members appointed to the board; direct Welsh Government oversight of strategy and delivery.
- **Level 5 — Special measures.** Full ministerial direction; board suspension or replacement; direct control of the organisation by Welsh Government.

DHCW was escalated to **Level 3** in March 2025 — the first Welsh NHS body ever at this level. It has been held there for over 12 months with no de-escalation and no structural consequences. At the time of writing, DHCW is reportedly elevated to **Level 4** but the elevation has not been publicly announced — the timing coincides with the Welsh election period.

The architecture that makes Level 3 unable to function as a correction mechanism is documented at [The Architecture of Failure](/diagnosis/) and [/traps/escalation/](/traps/escalation/). For the definitive framework document and the latest escalation status, see [carenhs.org](https://carenhs.org).

### MedCom

The Danish national interoperability body, responsible for the country's health-data exchange standards and infrastructure. A working example of a non-monopoly delivery model. See [Alternatives](/blueprint/alternatives/).

### NWIS

NHS Wales Informatics Service. DHCW's predecessor organisation, operating until it was reconstituted as DHCW in 2021. Much of DHCW's current leadership came across from NWIS.

### OpenEyes

DHCW ophthalmology programme: £8.5M spent, four-plus years late, live in only two of seven health boards. One of several programmes under Level 3 targeted intervention.

### Reinforcing loop (R)

A feedback loop that amplifies change. When the dynamic is harmful, it produces a vicious cycle that gets worse over time. All eleven loops identified in the diagnosis are reinforcing. See [Methodology](/methodology/).

### Senedd

The Welsh Parliament (Senedd Cymru / Welsh Parliament). Its Public Accounts and Public Administration Committee has produced several of the primary documents the analysis draws on.

### Stock

Something that accumulates or depletes over time: money, headcount, trust, institutional knowledge, delivery capability. Stocks are what a competent board watches; flows are what management typically reports. See [Methodology](/methodology/).

### Sundhedsdatastyrelsen

The Danish Health Data Authority. Governs national health data in Denmark. See [Alternatives](/blueprint/alternatives/).

### System trap

A recurring pattern of systemic dysfunction produced by interacting feedback loops. Donella Meadows catalogued seven in _Thinking in Systems_; each has a characteristic signature and a known escape route. DHCW has all seven active simultaneously. See [/traps/](/traps/).

### TEHIK

_Tervise ja Heaolu Infosüsteemide Keskus_ — Estonia's Centre for Health and Welfare Information Systems. Operates Digilugu and the surrounding national digital-health stack. See [Alternatives](/blueprint/alternatives/).

### Thinking in Systems

Donella Meadows' 2008 primer on systems dynamics (Chelsea Green, posthumous). The canonical short introduction to the discipline and the source of the seven system traps used on this site.

### WCCG

Welsh Clinical Communications Gateway. National messaging infrastructure that ran on unsupported technology for eight-plus years despite repeated warnings from staff.

### WCCIS

Welsh Community Care Information System. £42M+ spent, deployed to only 19 of 29 target organisations, subsequently rebranded as Connecting Care. See [L4: The Rebranding Escape](/diagnosis/l4-rebranding-escape/).

### WPAS

Welsh Patient Administration System. Identified as a contributory factor in at least one patient death. Described by one health board as "the single biggest risk to patient safety."

### X-Road / X-tee

Estonia's national data-exchange layer — a secure, decentralised backbone through which public-sector systems interoperate. Underpins Digilugu and much else. See [Alternatives](/blueprint/alternatives/).


# The Blueprint
URL: https://bluenhs.org/blueprint/

_Wales can have world-class digital health. Denmark, Estonia, and NHS Digital England already built it — with less money, fewer staff, and competent recruited leadership. This is what the destination looks like._

Three countries comparable to Wales already built world-class digital health. They spent less. They recruited better. They delivered more. They did it by rejecting the very model Wales chose.

This page defines where Wales needs to get. The [six interventions](/interventions/) describe how to get there.

## What Good Looks Like

A GP in Bridgend opens a patient record and it works. A consultant in Swansea can see a patient's full history from Betsi Cadwaladr. A nurse in Hywel Dda refers a patient electronically and the referral arrives instantly, tracked end to end. A patient checks their waiting time online, in real time, because the data is published. The 30,000 people currently waiting in England because Welsh digital systems cannot cope — that number approaches zero.

This is not a fantasy. It is a description of what Denmark and Estonia deliver today, for comparable populations, at a fraction of the cost.

Wales has the funding, the clinical talent, and the infrastructure foundation. What it lacks is the structural conditions for that capacity to be applied.

## The "Once for Wales" Monopoly

DHCW exists as a single national Special Health Authority delivering all digital infrastructure for NHS Wales. Every system, every integration, every strategic procurement flows through one monopoly delivery body. This is described — in DHCW strategy documents, in Welsh Government commissioning papers, in the public rationale for DHCW's existence — as "once for Wales." The phrase is presented as self-evident. A small country. A single delivery body.

It is not the default. Every comparable country made a different choice. The monopoly delivery model is not industry standard. It is a specifically Welsh arrangement — and the results speak for themselves. DHCW's annual budget reached **approximately £200M in 2025-26**; cumulative Welsh Government funding over five years exceeds £600M; quantified delivered value across the full period is £0.5M. That is 83p of return per £1,000 invested, and the return rate has worsened year on year as headcount and budget have grown faster than delivery. Twelve months of [Level 3 enhanced monitoring](/glossary/#level-3-enhanced-monitoring-and-the-wg-5-level-framework) produced no de-escalation; in 2026, DHCW was escalated further to Level 4 Targeted Intervention. As of March 2026, £32.9M of DPIF revenue and £13.1M of capital remained unallocated, with the Accountable Officer formally stating delivery was "not possible without confirmed DPIF allocation"; two of forty-seven Phase One milestones had been missed (LIMS and WRISTS). The Cabinet Secretary publicly described the entire escalation framework as "complex, data-heavy, burdensome, lacks transparency and does not drive improvement."

Even the timing of the failure is structural. Every governance deficit pattern observed at the point of Level 3 escalation 34 months in was already operational at the very first board meeting. The architecture was complete on day one.

Three jurisdictions built differently, spent less, and deliver more.

## What Comparable Countries Built

Every country comparable to Wales in scale and governance chose a different architecture. The details differ. The structural pattern is identical: **competent technical leadership recruited against external criteria**, and **interoperability standards separated from application delivery**. No monopoly delivery body.

### Denmark — Standards Body + Regional Delivery

Denmark (5.9 million people) separated standards from delivery. A central body — [MedCom](https://medcom.dk/) — develops and certifies interoperability standards. Five health regions choose and procure their own clinical systems. The national patient portal [sundhed.dk](https://da.wikipedia.org/wiki/Sundhed.dk) has served citizens since 2003. The result for a Danish patient: full record access, cross-regional referrals, and digital prescriptions — built on an interoperability backbone, not a monopoly delivery body.

[Read the full Denmark case study →](/blueprint/alternatives/#denmark)

### Estonia — Interoperability Layer + Sovereign Delivery

Estonia (1.3 million people — 40% of Wales) built [X-Road](https://e-estonia.com/solutions/interoperability-services/x-road/), an open-source national data exchange layer now used by Finland, Iceland, and Ukraine. [TEHIK](https://www.tehik.ee/en), the national health IT authority, covers 1.3 million people with roughly 200 staff — one-sixth of DHCW's headcount. Over 99% of prescriptions are digital. The result for an Estonian patient: a unified health record assembled from distributed systems, no single point of failure, at a fraction of the cost.

[Read the full Estonia case study →](/blueprint/alternatives/#estonia)

### NHS Digital England — Competent Leadership, Federated Procurement

[NHS Digital](https://digital.nhs.uk/about-nhs-digital) recruited its C-suite openly from the commercial sector — Rolls-Royce, Jaguar Land Rover, Credit Suisse, HSBC, the Home Office. Roles were advertised at market rates against published technical criteria, assessed by external panels. Compare DHCW: executive roles filled through processes that did not match this standard, with a recruitment pattern documented at [L7: The Competence Void](/diagnosis/l7-competence-void/). NHS Digital's record is not perfect — Care.data and NPfIT are documented failures — but tens of millions use the NHS App and electronic prescribing operates at national scale.

[Read the full NHS Digital England case study →](/blueprint/alternatives/#nhs-digital-england)

### The Common Pattern

None of these countries built "once for Denmark" or "once for Estonia" as a delivery monopoly. All three separated interoperability from delivery. All three recruited technical leadership externally against verifiable criteria. None of them assembled an executive cohort from a single antecedent organisation's patronage pipeline — pre-credentialled before the new body's first board meeting — as a starting condition. The monopoly model, and the patronage pipeline that staffs it, is a specifically Welsh arrangement, designed and maintained by the governance system this analysis critiques.

Wales already has the foundations to follow this pattern. It has regional health boards with more operational autonomy than Danish regions had before the 2007 reform. It is wealthier and better-staffed than Estonia was when X-Road launched in 2001. The gap is not resources. It is governance.

## The Destination

In the destination state, patients and clinicians experience the result:

A health board in Hywel Dda procures a referral system that meets its clinical needs — not the system DHCW built. A national patient index means the data still flows across boundaries. The GP in Bridgend still sees the full record. Waiting times are published in real time because transparency is statutory, not discretionary. When something fails, the organisation conducts a genuine post-mortem and publishes the findings.

DHCW still exists — but as a standards-and-interoperability body, the functional equivalent of MedCom or TEHIK. Tightly scoped: national data standards, national interoperability, national patient index, national cybersecurity. Nothing else. Delivery happens at health-board level, under clinical leadership, within those standards.

The single deepest change is the leadership paradigm: *"I am a temporary steward of public resources. My competence is measured solely by what patients and clinicians experience. If I cannot deliver, I should be replaced by someone who can. Transparency is non-negotiable."*

Every element of this destination exists in at least one comparable health system. The destination is proven. The [six interventions](/interventions/) are the steps to reach it.

## The Economics

Two paths from today. The gap widens every year.

DHCW's annual budget reached approximately £200M in 2025-26 with quantified delivered value across five years of £0.5M. Against that, direct annual waste runs at **£100-150M per year**; cumulative five-year status-quo direct waste is **£500M-£1B**. Planned reform requires a one-off £5-15M investment that breaks even **in weeks**, not months.

Downstream impact across the seven health boards and NHS Wales is **5-15× the direct figure** — total five-year cost of status quo to NHS Wales: **£3-10 billion**. These are the financial figures. They do not include the cost of patient harm — WPAS-linked deaths, eMPI patient-record mixups, eight years of unsupported WCCG technology, WRISTS-and-LIMS milestone failure — which is on a separate ledger and not exchangeable for pounds.

[Read the full cost analysis →](/blueprint/cost-of-inaction/)


## The Target Architecture
URL: https://bluenhs.org/blueprint/target-architecture/

_What NHS Wales' digital infrastructure should look like in the destination state. A federated, standards-led, open-API architecture modelled on Estonia's X-Road, Denmark's MedCom, Finland's Kanta, and NHS England's national platforms — translated for Welsh constraints. Six layers, three principals, open standards, building blocks not platforms._

The diagnosis explains why NHS Wales' digital infrastructure failed. The six interventions explain what to change about the conditions that produced the failure. This page describes what the resulting technical estate looks like once those interventions have run.

Every component on this page is in operational use somewhere in northern Europe today. None is a Welsh invention. The novelty of the Welsh target architecture is the combination — and the discipline of refusing to build anything the international evidence does not already endorse.

### Five design principles

1. **Standards are separated from delivery.** A small national body sets and certifies interoperability standards. It does not build clinical applications. Health boards procure clinical applications within those standards. Denmark, Estonia and Finland operate this separation; the DHCW monopoly model collapses it. Separation breaks L11: Captured Governance at the technical layer — the standards-setter is no longer judge-of-its-own-delivery.

2. **Federation, not centralisation.** Patient data lives in the system that generated it: the GP record sits with the GP system, the hospital record sits with the EPR, the prescription record sits with the dispensing system. A national interoperability backbone makes the federation queryable — the patient record looks unified, but no monolithic database exists. Estonia has operated this pattern since 2001. The architectural consequence is that no single supplier outage can take the estate offline — the failure mode the PSBA March 2026 outage demonstrated.

3. **Open standards by default.** HL7 FHIR R4, SNOMED CT International, dm+d, ICD-10/11, OpenID Connect, OAuth 2.1, IHE profiles, X-Road or equivalent. Where a Welsh standard is required (bilingualism, devolved policy), it is an extension of an international standard, not a replacement. Procurement specifications cite international standards by name; non-compliant tenders are non-compliant on technical grounds.

4. **Open source as the default for shared infrastructure.** The interoperability backbone, the terminology services, the audit ledger, the national identity broker, the patient-facing app — open source by default. Source code is published under permissive licence. Procurement of proprietary software at the application layer remains a health-board choice, but the layers that hold the federation together are sovereign, inspectable, and forkable.

5. **Building blocks, not platforms.** Every national component does one thing — identity, messaging, terminology, audit, indexing — and exposes a typed API. There is no "national platform" that does everything. The international failure cases (Care.data, NPfIT) were platforms; the international successes (X-Road, sundhed.dk, NHS Login, FHIR-based health information exchanges) are building blocks.

### The six-layer stack

- **Layer 1 — Sovereign Infrastructure.** UK-resident cloud (hyperscaler with sovereign zones plus on-prem for the regulated minority), high-availability networking, a successor to PSBA designed with explicit redundancy, the `.nhs.wales` DNS authority, a public-sector certificate authority, and the cyber-defence operations centre. No single shared dependency may have the power to take the estate offline.
- **Layer 2 — Identity, Security, Audit.** A national citizen identity service (NHS Wales account, OpenID Connect, federable with NHS Login). A national clinician identity service (smartcards transitioning to FIDO2/passkeys, OAuth 2.1 + SCIM). An immutable, append-only audit ledger. A national PKI. National cyber-defence at this layer.
- **Layer 3 — The Wales Health Interoperability Backbone.** An X-Road-equivalent federated data exchange. A FHIR API gateway. A clinical-event pub/sub bus (admit, discharge, transfer, prescription issued, prescription dispensed, referral made, referral closed). National terminology services (SNOMED CT, dm+d, ICD-10/11) as API endpoints. Direct adoption of an existing open-source stack — X-Road, Mirth Connect, HAPI FHIR — rather than bespoke Welsh code.
- **Layer 4 — National Shared Services.** A small set of national services that genuinely benefit from national operation: Welsh Demographics Service (NHS number, residence, registration — successor to the eMPI), National Practitioner Index, National Electronic Prescription gateway, National Referral and Booking gateway, federated Welsh Care Record (a virtual record assembled at query time, not a centralised database), Population Health Data Platform. Each a building block, each exposing a typed FHIR API.
- **Layer 5 — Health-Board Clinical Applications.** EPRs, theatre management, maternity, mental health, endoscopy, ophthalmology, cancer pathways, community care, the WCCIS successor — procured by health boards within the national standards. Health boards may choose Epic, Cerner, Better, Cynerio, an open-source system or a domain-specific best-of-breed product. Procurement compliance with layers 2–4 standards is the gating criterion; brand of EPR is a health-board decision.
- **Layer 6 — Patient and Clinician Experience.** The NHS Wales App (citizen-facing, mobile and web, statutory bilingual parity, accessibility-first). A citizen consent surface giving granular control. A clinician workspace that aggregates views across the EPRs the clinician has access to. A public population dashboard publishing waiting times, performance, vendor portfolio status and digital incidents at statutory frequency.

### Component responsibility map

Three principals, no overlap:
- **Standards body** (~400 staff at steady state, scaled from Estonia's TEHIK at ~200 staff for 1.3M people): interoperability backbone, terminology services, national patient index, conformance test suites, audit ledger, national cyber-defence operation, open-source stewardship. Does not build clinical applications.
- **Health boards** (7 boards): clinical applications, local procurement, embedded engineering teams, post-implementation review, clinical-safety case ownership, local operational metrics.
- **Citizens** (3.1M): identity, granular consent, full record access, prescription self-service, right to audit access log, rectification and complaint, bilingual interface (statutory).

Conformance testing happens at the boundary between columns. Every clinical application a health board procures must pass the conformance suite the standards body publishes. Pass/fail is publicly published.

### Federation, not monopoly

Federated standards produce five properties monopoly delivery cannot:
- **Failure is local.** A health-board EPR outage stops that health board, not all seven. A PSBA-class shared-infrastructure failure cannot reach the application layer because applications no longer share the dependency.
- **Procurement is competitive.** A dissatisfied health board has a credible exit: re-procure within the standards. The vendor is on notice; the standard is the contract.
- **Comparability is structural.** Seven boards running different EPRs within the same standards produces seven natural experiments. Outcomes are comparable because the data model is the same.
- **Capability accumulates locally.** Embedded teams operating inside health boards under clinical leadership build deep clinical-domain capability the monopoly never could.
- **National capacity is conserved.** The standards body, freed from the impossible job of building everything everywhere, does the small set of things that genuinely benefit from national operation — and does them well.

### Standards adopted

| Layer | Standard | Operational precedent |
|---|---|---|
| Interoperability | HL7 FHIR R4 (R5 transition path) | NHS England, US Cures Act, Finland (Kanta), Australia |
| Interoperability | IHE profiles (XDS, PIX, PDQ) | Denmark, Switzerland, USA |
| Interoperability | X-Road / equivalent federated data exchange | Estonia, Finland, Iceland, Ukraine, Faroe Islands |
| Terminology | SNOMED CT International | UK, USA, Denmark, Netherlands, Australia |
| Terminology | dm+d | NHS UK-wide |
| Terminology | ICD-10 / ICD-11 transition | International |
| Records | OpenEHR archetypes (where clinical models needed) | Norway, Slovenia, regional UK |
| Identity | OpenID Connect / OAuth 2.1 | NHS Login, GOV.UK One Login, Estonia eID |
| Identity | FIDO2 / passkeys | NHS Login, US federal, Google, Apple |
| Messaging | Kafka / AMQP-compatible event bus | International |
| API | OpenAPI 3.1 | International default |
| Audit | RFC 3161 / append-only ledger | International, regulated industries |

### Operating model

Six non-negotiable disciplines:
1. **API contracts published, versioned, free.** OpenAPI spec, versioning policy, deprecation calendar, public conformance test suite. Free at the point of use for any NHS Wales body.
2. **Conformance automated, not negotiated.** Pass/fail publicly published. A failing vendor cannot be procured; a passing vendor cannot be excluded on opaque grounds.
3. **Four-year procurement cycles with mandatory recompetition.** No contract above £1M renews without an open recompetition.
4. **Strangler-fig migration, never big-bang replacement.** New capability is built behind the national API gateway. Existing system continues to operate. Old system retired one boundary at a time.
5. **Open source by default for the standards body's own work.** Source code published under permissive licence. NIIS-style consortium model.
6. **Public, machine-readable observability.** Every layer publishes operational metrics (availability, latency, error rates, incident counts) to a public dashboard at statutory frequency.

### What this architecture is not

- Not a national platform (Care.data and NPfIT failure modes).
- Not a monopoly delivery body (once-for-Wales rejected by every comparator).
- Not a Welsh-bespoke standards stack (3M-population user base cannot operate parallel standards credibly).
- Not closed-source national infrastructure (open source by default at sovereign-capability layers).

### Welsh translation of the international pattern

- **Bilingualism statutory.** Welsh-language and English-language parity is a layer-6 invariant. Terminology services support Welsh-language clinical terms where established.
- **Market depth limited.** Some specialist supplier negotiations require coordination with NHS England rather than independent procurement. The architecture interoperates with the NHS UK-wide API estate — NHS Login, GP Connect, the NHS App — rather than building parallel infrastructure.
- **Devolved-and-reserved boundaries explicit.** Welsh national services for devolved functions; interoperation with UK reserved-function infrastructure; no duplication of either.

The architecture is the picture of where the system arrives once the six interventions have run their course. The 36-month timeline is the route.

[Read the full target architecture →](/blueprint/target-architecture/)


## International Case Studies
URL: https://bluenhs.org/blueprint/alternatives/

_Denmark, Estonia, and NHS Digital England built world-class digital health for less, with competent recruited leadership. What they built, how they structured it, and what it cost._

Every country comparable to Wales in scale and governance made a different structural choice. None adopted a monopoly delivery body. All separated interoperability standards from application delivery. All recruited technical leadership externally against verifiable criteria.

This page documents the three most relevant comparators: Denmark (standards body + regional delivery), Estonia (interoperability layer + sovereign delivery), and NHS Digital England (competent leadership + federated procurement).

Denmark — Standards Body + Regional Delivery

Denmark has a population of around 5.9 million, roughly twice Wales's 3.1 million, organised into five health regions with real autonomy over delivery. The national architecture is divided: a central body sets standards and maintains the interoperability backbone; regions choose and procure their own electronic health record systems.

The [Danish Health Data Authority (Sundhedsdatastyrelsen)](https://sundhedsdatastyrelsen.dk/da/english) owns national data standards, the national patient index, and the statistical registers that make Danish health data one of the most complete datasets in Europe. It does not build regional EHR systems. It defines the standards those systems must meet.

The interoperability layer sits with [MedCom](https://medcom.dk/), a publicly-funded non-profit founded in 1994, jointly owned by the Ministry of the Interior and Health, Danish Regions, and local government (KL). MedCom develops, tests, and certifies the messaging standards that allow every Danish health IT system to exchange clinical data. It does not sell software. It connects software. Denmark has ranked consistently in top-tier international interoperability comparisons for two decades — a position MedCom has held continuously since before most UK digital-health strategy documents were written.

[Sundhedsplatformen](https://www.regionh.dk/) is the Epic-based regional EHR covering Region Hovedstaden (Copenhagen) and Region Sjælland — approximately 2.5 million residents. Implementation began in 2016. Capital-region cost estimates reported in the Danish press cluster around DKK 2.8 billion (~£320M) over the rollout period. The go-live attracted serious public controversy; the system is now functioning and measurably improving clinical outcomes. Other regions — Midtjylland, Syddanmark, Nordjylland — made independent EHR choices. Regions have real power to choose.

The national patient portal [sundhed.dk](https://da.wikipedia.org/wiki/Sundhed.dk), operational since 2003, gives every Danish citizen access to their own health record, prescriptions, lab results and referrals. It is built on top of MedCom's infrastructure — not on top of a monopoly delivery body.

**The structural lesson:** in Denmark, standards are separated from delivery. The national body does the interoperability work that no one else can do efficiently. Regions compete and choose on the applications that sit above it.

**The Welsh comparison:** Wales already has regional health boards with more operational autonomy than Danish regions had before the 2007 structural reform. The governance infrastructure to support this pattern exists. The choice to centralise delivery through a monopoly was exactly that — a choice.

Estonia — Interoperability Layer + Sovereign Delivery

Estonia has a population of around 1.3 million — roughly 40% of Wales. It is the most frequently cited digital-government case study in the world, and almost none of its architecture is a monopoly delivery model.

The foundation is [X-Road (X-tee)](https://e-estonia.com/solutions/interoperability-services/x-road/), the national data exchange layer Estonia designed and has operated since 2001. X-Road is open source, used across 99% of Estonian state services, and federates data between hundreds of public and private databases. Every EHR system in Estonia is connected via X-Road. No single vendor, and no single government body, holds the clinical data. The data is distributed across the systems that generate it; X-Road makes it look unified.

[TEHIK](https://www.tehik.ee/en) — the Health and Welfare Information Systems Centre, about 200 staff, established in its current form in 2017 — is Estonia's national health IT authority. It operates X-Road for health, runs the national e-Health record as a virtual record federating data from existing provider systems rather than a monolithic database, and manages the information systems that underpin Estonian social security and labour services. TEHIK covers health IT for 1.3 million people with roughly one-sixth of DHCW's headcount.

Estonia's e-prescription service (Digiretsept) launched in 2010. Over 99% of prescriptions in Estonia are now issued and dispensed digitally. Published costs for initial development are in the low single-digit millions of euros, with annual running costs reported in the same range.

The [Nordic Institute for Interoperability Solutions (NIIS)](https://niis.org/) is the joint Estonia-Finland organisation that now develops X-Road as open-source infrastructure. Finland built its national e-health service [Kanta](https://www.kanta.fi/en/about-kanta-services) on the same X-Road stack. Iceland, Ukraine, Schleswig-Holstein, Québec and others have adopted it. Compare: DHCW's "once for Wales" monopoly against an Estonian interoperability standard co-developed with Finland and exported across continents.

**The structural lesson:** Estonia's national body owns the connective tissue, not the applications. Applications are chosen by providers. Data follows the patient because interoperability is mandatory — not because one body builds everything.

**The Welsh comparison:** Wales is wealthier, better-staffed, and has stronger institutional foundations than Estonia had in 2001 when X-Road launched. Estonia built world-class digital health with fewer resources and a harder starting position. The gap between Wales and Estonia is not funding or talent. It is governance.

NHS Digital England — Competent Leadership, Federated Procurement

NHS England Transformation Directorate (which absorbed NHS Digital in February 2023) covers around 56 million people. The architecture is: central standards, competitive national services, trust-level procurement.

The most important contrast is leadership recruitment. [NHS Digital](https://digital.nhs.uk/about-nhs-digital) recruited its C-suite openly from the commercial sector. Chief Information Officer roles have been filled from Rolls-Royce and Jaguar Land Rover. The Chief Digital Officer came from the Home Office. Senior technology leaders were recruited from Credit Suisse and HSBC. Roles were advertised at market rates. Selection was against published technical criteria, run by external panels.

Compare DHCW's recruitment model. Executive roles — CEO, CDO, Head of Software Engineering — were filled through processes that did not match this standard. The CEO role was filled by a candidate with 30+ years in NHS Wales finance and health information management — not technology delivery. The CDO is an internal promotion with 19 years at NWIS/DHCW. The Head of Software Engineering was advertised at Band 8c (£71-82k) — well below market rate for the responsibility. Selection was conducted by a small internal circle, not by open external recruitment against verifiable technical criteria. The structural pattern is documented at [L7: The Competence Void](/diagnosis/l7-competence-void/).

The delivery record follows from the recruitment model. The [NHS App](https://digital.nhs.uk/services/nhs-app) now has tens of millions of registered users. [GP Connect](https://digital.nhs.uk/services/gp-connect) provides national GP-record interoperability integrated with all major GP software providers. Electronic prescribing operates at national scale. NHS Digital's track record is not perfect — Care.data and NPfIT are documented failures — but the scale and competence of delivery is not comparable to DHCW's 0-of-9 record on time, 83p of return per £1,000 invested.

**The structural lesson:** competent technical leadership, recruited externally against verifiable criteria, produces delivery. The recruitment process determines the delivery outcome. This is not correlation; it is mechanism.

**The Welsh comparison:** Welsh public-sector salary scales are already close to NHS Digital's for non-executive roles. The gap is at C-suite level and in the selection process. Closing it requires political will, not additional funding.

## The Pattern That Wales Rejected

Three different architectures. One consistent structural pattern:

1. **Competent technical leadership** recruited against external, verifiable criteria
2. **Interoperability standards separated from application delivery**
3. **No monopoly delivery body** staffed by executives promoted from administration, finance, or long internal service
4. **No patronage pipeline pre-credentialing the executive cohort** before the new body's first board meeting

None of these countries built their equivalent of "once for Wales" as a delivery monopoly. The monopoly model — and the patronage pipeline that staffs it — is a specifically Welsh arrangement, designed and maintained by the same governance system this analysis critiques.

The [blueprint](/blueprint/) proposes a variant of the Danish-Estonian pattern adjusted for Welsh political reality. The [six interventions](/interventions/) are the operational steps to get there.

## What Wales Should Learn From Their Failures

The three jurisdictions are not flawless. Each carries documented failures that inform what reform should — and should not — attempt.

**NHS Digital England.** Care.data (2014, halted 2016) failed because patient consent and communication were treated as a downstream task. NPfIT (2002-2011, written off at roughly £10 billion) failed because it was attempted as a top-down monolithic build — exactly the architecture Wales is now being urged to dismantle. Both failures are directly relevant: Wales should not repeat them by sequencing transparency after delivery (Care.data) or by trying to standardise applications rather than data flows (NPfIT).

**Denmark — Sundhedsplatformen go-live.** The Epic-based EHR rollout in Region Hovedstaden produced documented disruption to clinical workflows in 2016-2018, with significant clinician dissatisfaction reported in the Danish press. The lesson: even well-resourced regional procurement requires deep clinical involvement in implementation, not just procurement. Welsh embedded teams under [Flip the Model](/interventions/flip-the-model/) carry this lesson directly.

**Estonia — early X-Road resistance.** When X-Road launched in 2001, agency-level resistance to data-sharing standards was significant. Estonia overcame it by making interoperability mandatory in legislation, not encouraged in policy. The lesson: interoperability standards have to be statutory, not aspirational, or they fail. Welsh standards under the architectural endpoint of [Flip the Model](/interventions/flip-the-model/) carry this lesson.

The point is not that the alternatives are perfect. It is that they have records — public, debated, learned from — to compare against. DHCW's record is comparable in cost and time, with documented delivery far below any of them.

## Risks the Alternatives Don't Cover

Wales has structural conditions the comparator jurisdictions do not. Three are material:

- **Market depth.** Wales has roughly half Denmark's population and is part of a UK procurement market dominated by NHS England's purchasing power. Some specialist supplier negotiations require coordination with England rather than independent procurement.
- **Bilingualism.** Welsh-language primary care interfaces and patient-facing systems are a statutory requirement, not an option. Estonia's largely monolingual interface is not a complete model.
- **Devolved-versus-reserved policy boundaries.** Several digital-health-adjacent functions (the UK-wide elements of the NHS App, GP IT systems' UK-wide elements, identity infrastructure) sit at the intersection of devolved and reserved competence. The standards body Wales builds must interoperate with reserved-function infrastructure, not stand alone.

These constraints do not invalidate the comparator pattern. They shape implementation choices within it. The blueprint adjusts for them; the destination remains the same.


## Once for Wales
URL: https://bluenhs.org/blueprint/once-for-wales/

_The Welsh policy framing that justifies DHCW's monopoly on digital delivery for NHS Wales. Not 'how small countries do digital health' — Denmark, Estonia, and NHS Digital England all rejected this model. A specifically Welsh arrangement that the comparator countries explicitly avoided._

## What "Once for Wales" means

"Once for Wales" is the policy framing under which Digital Health and Care Wales (DHCW) exists as a single national Special Health Authority delivering all digital infrastructure for NHS Wales. Every system, every integration, every strategic procurement flows through one monopoly delivery body. The phrase appears in DHCW strategy documents, in Welsh Government commissioning papers, and in the public rationale for DHCW's existence. It is presented as self-evident: a small country, a single delivery body.

It is not self-evident. It is a specifically Welsh arrangement that every comparator country explicitly rejected.

## Why it is presented as the default

The argument runs:

- Wales has 3.16 million people, served by seven health boards.
- A unified digital backbone benefits from network effects.
- A single delivery body avoids duplication.
- "Once for Wales" therefore captures economies of scale that fragmented delivery would miss.

The reasoning is plausible at first reading. It is also the reasoning that every comparator country considered and rejected.

## Why the comparators rejected it

**Denmark** (5.9 million people, comparable to Wales × 2). Denmark separates **standards** (set by Sundhedsdatastyrelsen, the Danish Health Data Authority, on behalf of the state) from **delivery** (carried out by multiple regional and commercial actors operating under those standards). Sundhedsdatastyrelsen is a small agency with regulatory teeth — it does not deliver applications. The delivery layer competes; the standards layer arbitrates. Danish citizens see a single national patient record (Sundhedsplatformen / sundhed.dk) — but that record is fed by multiple delivery actors, not one.

**Estonia** (1.3 million people — closer to a single Welsh health board in population). Estonia operates **e-Health Estonia**, a national agency of approximately 200 staff. That is **one-sixth of DHCW's headcount**. Like Denmark, Estonia separates the standards function (run by the agency) from delivery (which sits with health institutions and the X-Road interoperability layer). The agency is a thin standards body; the delivery is distributed.

**NHS Digital England** (now NHS England Transformation Directorate) — comparable scale, far higher institutional maturity. England separates national infrastructure (Spine, NHS Number, dm+d) from application delivery (which sits with trusts, primary care suppliers, and competing vendors). National standards are set centrally; delivery is plural. No equivalent of DHCW's monopoly delivery body exists.

The pattern is consistent. **Every comparator separates standards-setting from application delivery.** None has a single monopoly delivery body. "Once for Wales" is not the small-country default — it is the Welsh exception.

## How "Once for Wales" operates structurally at DHCW

The monopoly framing has three structural effects, each of which the Blueprint analysis maps to a specific feedback loop:

1. **No exit option for health boards.** If a health board is dissatisfied with the delivery of a national system — say, WPAS, eMPI, or RISP — there is no alternative supplier within NHS Wales. The health board can complain; it cannot procure elsewhere. This is the mechanism that produces [L2: The Credibility Death Spiral](/diagnosis/l2-credibility-death-spiral/) — trust erodes but cannot be priced into delivery decisions because there is no competitive signal.

2. **No external benchmark for quality.** When the delivery layer is one body, "what good looks like" is whatever that body produces this quarter. There is no parallel delivery against which to compare. Drift to low performance becomes invisible, because the comparator is yesterday's DHCW rather than today's Denmark. This is the structural mechanism behind [Trap: Drift to Low Performance](/traps/drift-to-low-performance/).

3. **Standards and delivery are not separated.** DHCW sets the technical standards *and* delivers the systems that implement them. The standards-setter and the delivered-against party are the same organisation. Audit Wales has no independent technical-standards function to refer to. The Welsh Government has no separate standards-body to commission. The result is a closed-loop technical authority — [L11: Captured Governance](/diagnosis/l11-oversight-obstruction/) compounds this directly.

## What ending the monopoly looks like

The Blueprint's [Intervention 4: Flip the Model](/interventions/flip-the-model/) ends the once-for-Wales monopoly delivery model. Specifically:

- **Separate the standards function from the delivery function.** A small Welsh national digital standards body sets technical and interoperability standards; delivery is carried out by health boards, the DHCW successor, and (where appropriate) external providers operating under those standards.
- **Embed engineering teams in health boards under clinical leadership.** Removes the unbreakable monopoly between the delivery body and the receiving institutions.
- **Recruit externally on published technical criteria.** Removes the loyalty-selection pipeline that the monopoly has structurally protected.

Ending "once for Wales" is not anti-cooperation. It is anti-monopoly. Denmark and Estonia are more digitally integrated than NHS Wales by every measurable outcome — and they got there by separating standards from delivery, not by concentrating both in one body.

## What this page does not say

This page does not argue that DHCW should be dissolved. It argues that the single-monopoly-delivery model is the wrong structural condition for any successor body to operate under. The next step is in [Intervention 4](/interventions/flip-the-model/).

The "once for Wales" framing has been load-bearing for DHCW's existence. Removing the framing is not a presentational choice — it is a precondition for the structural reform the Blueprint proposes.


## The 36-Month Timeline
URL: https://bluenhs.org/blueprint/timeline/

_Five phases from external intervention through leadership reset to consolidation. Trust rebuilds on a 2-5 year lag; the timeline does not compress._

Reform at this depth cannot be compressed into a political cycle. Trust rebuilds on a 2-5 year lag behind demonstrated delivery. Embedded teams need 6-12 months to ship. The sequencing matters as much as the components.

Three decision points anchor the timeline. **Month 6:** leadership-change decisions are made on the forensic review's findings. **Month 12:** the transparency dashboard goes live under statutory publication; the first patient-safety triage outputs are published. **Month 18:** the first health board reports either "DHCW delivered what they promised" — the single most important signal — or the reform escalates to the next intervention tier.

{% roadmap() %}
{% roadmap_phase(period="Months 0-3", label="External intervention begins") %}
Welsh Government commissions independent forensic review. Patient safety triage of live national systems (WPAS, eMPI, WCCG, WICIS, RISP, PSBA-class infrastructure); systems pause where the safety case is unsupported. Audit Wales mandates data publication. Protected staff reporting channel established. Transparency dashboard designed, ready for deployment once the review creates conditions for its use. In parallel: Senedd Public Accounts Committee opens the Reform-the-Funder workstream — RAG audit trail, capital-and-revenue coherence rules, remit-letter discipline.
{% end %}
{% roadmap_phase(period="Months 3-6", label="Diagnosis") %}
Forensic review findings delivered. Independent skills audits of all executive/director roles completed. Non-executive board audit complete; sub-committees reconstituted with technical NEDs against published competency criteria. Leadership change decisions made based on review findings.
{% end %}
{% roadmap_phase(period="Months 6-12", label="Leadership reset + structural reform") %}
New leadership appointed. Transparency dashboard launched under statutory publication. Independent panel selects 3 priority programmes against patient-safety-weighted criteria; published 24-month stop list. Vendor portfolio audit complete; sole-bidder contracts above £1M re-tendered or terminated. Two pilot health boards selected; embedded teams assembled with externally recruited team leads. Multi-year programme funding negotiations begin under Reform-the-Funder terms.
{% end %}
{% roadmap_phase(period="Months 12-24", label="Delivery") %}
Focused programmes delivering to clinical users. Embedded teams shipping working software. Culture shifting as staff see real delivery metrics. Genuine post-mortems on WCCIS, OpenEyes, LIMS, RISP, WICIS, and WCCG completed and published. Trust recovery begins — the earliest it can start. This is the trough phase: short-term political pressure is maximum, visible delivery is minimum; the temptation to abandon the reform is strongest exactly when leading indicators say to hold. Trust rebuilds on a 2-5 year lag behind demonstrated delivery.
{% end %}
{% roadmap_phase(period="Months 24-36", label="Consolidation") %}
Embedded model spreading to additional health boards. Portfolio cautiously expanding based on proven demand. L2: Credibility Spiral showing signs of reversal. Organisation identity transitioning from "national IT department" to standards-and-interoperability body — the architectural endpoint of Flip the Model. Health boards procure clinical applications within national standards; the standards body holds the patient index, interoperability backbone, and shared cybersecurity.
{% end %}
{% end %}


## The Cost of Inaction
URL: https://bluenhs.org/blueprint/cost-of-inaction/

_DHCW's annual budget reached ~£200M with £0.5M of quantified value across five years. Direct waste runs at £100-150M per year. Cumulative five-year direct waste in status quo: £500M-£1B. Downstream NHS Wales impact: £3-10 billion. Plus patient harm — a separate ledger, not monetisable._

Two paths from today. The gap between them widens every year.

Year 1 has already begun. As of March 2026: £32.9M of DPIF revenue and £13.1M of capital remain unallocated; the Accountable Officer has formally stated delivery is "not possible without confirmed DPIF allocation"; recruitment is frozen by remit letter; the CFO and the sharpest non-executive director (Rowan Gardner) have departed; two of forty-seven Phase One milestones have already been missed (LIMS and WRISTS); the IQPD oversight meetings are being wound down and replaced by a new escalation meeting chaired by the Director General from April. The figures below are not hypothetical at year zero — they are the trajectory if the conditions documented at March 2026 continue.

DHCW's annual budget reached approximately **£200M in 2025-26** — a year-on-year growth trajectory that has accelerated since founding. Quantified delivered value across the full five years remains **£0.5M** — the Finance Director's own admission. Against a £200M annual budget producing no measurable patient or clinician benefit at scale, the direct waste rate is **£100-150M per year**, not the conservative £25-40M figure used in earlier estimates. The conservative figure was anchored on a single waste category — re-procurement savings — and was always understood to be a floor, not a central estimate.

## The Figures Above Are The Floor, Not The Ceiling

The fork above quantifies **direct DHCW waste** — Welsh Government funding consumed without corresponding delivered value. These figures are already substantial. They are not the whole cost.

Every DHCW failure cascades outward. Seven health boards and three trusts depend on DHCW for national clinical systems, referral infrastructure, patient records, and interoperability. When WPAS malfunctions, the cost is not borne by DHCW — it is borne by clinicians whose lists break, by patients whose appointments are lost, and by the Welsh NHS budget that absorbs the delays. When WCCIS fails to deploy, community care teams build shadow IT and duplicate work. When cross-border referrals spike because Welsh digital infrastructure cannot cope, NHS Wales pays English providers to take Welsh patients — **over 30,000 currently waiting across the border**.

The PSBA outage of March 2026 makes the multiplier visible in a single event. When the Public Sector Broadband Aggregation network failed across all NHS Wales organisations, every health board lost O365, EPMA, RISP, and radiology simultaneously. Clinical lists could not be opened. E-prescribing stopped. Image viewing stopped. The cost of that single day was not borne by DHCW — DHCW does not own PSBA — but it was a direct consequence of architectural decisions about single-supplier shared dependencies that the diagnosis documents at [L5: The Vendor Dependency Spiral](/diagnosis/l5-vendor-dependency-spiral/). Multiply that day by every clinical hour lost across every health board: this is the downstream multiplier in concrete form.

The downstream multiplier is difficult to quantify precisely because the captured governance has not published the data to quantify it (see [L10: The Information Fortress](/diagnosis/l10-information-fortress/)). But directionally: for every £1 of direct DHCW waste, the downstream cost to health boards and NHS Wales is plausibly £5-15 — clinician hours consumed, patient-safety incidents, duplicated procurement, delayed treatment, cross-border referral costs. The March 2026 PSBA outage demonstrated the multiplier in a single event: a few hours of failure at the supplier level produced a full operational day's worth of lost capacity across every health board simultaneously. The system-dynamics term is _externality_: DHCW's dysfunction is a cost exported to everyone else in the Welsh NHS.

This is why the diagnosis is not a DHCW-internal matter. It is an NHS Wales matter.

## Total Five-Year Impact Across NHS Wales: £3-10 Billion

Doing the arithmetic at the current run-rate: **£500M-£1B of direct DHCW waste** over five years, multiplied by a **5-15× downstream factor**, implies total cost to NHS Wales of **£3-10 billion** over the five-year window. These are the numbers the Welsh NHS is currently paying, in clinician time, in patient-safety incidents, in cross-border referrals, and in delivery delays, for a delivery body that produces 83p of delivered value per £1,000 invested.

This is the scale. It is not rhetoric. It is straightforward compounding of the direct rate, the downstream multiplier, and the five-year window over which the reform sequence unfolds.

## The Cost That Is Not Counted

The figures above are pounds. They do not include the cost of harm to patients, because that cost is not on the same scale as money. WPAS has been linked to at least one patient death. The electronic master patient index has mixed up patient records. WCCG ran on unsupported technology for more than eight years against repeated technical-staff warnings. WICIS — the intensive care system — has been "effectively still on pause," with Welsh Government commissioning an independent patient safety review. The Royal Colleges' joint briefing of July 2025 found that patients "regularly experience delays that lead to worsening health."

Each of these is a category of harm that has no exchange rate to pounds. The £3-10 billion five-year figure is what NHS Wales is paying in money. What patients are paying is a separate ledger, and the entries on it cannot be undone by future reform — they have already accrued. The case for planned reform is not only economic. The case is that every additional month of the status quo adds entries to a ledger that should never have been opened.

The question is not whether reform happens. The question is whether Wales chooses planned reform now, at £5-15M, or crisis-forced reform later at two to three times that direct cost — with the downstream NHS Wales impact multiplying the whole sum into the billions, and the patient-harm ledger continuing to accrue entries that cannot be reversed.

## The Cost of Trust That Cannot Be Bought Back

Even successful reform carries a cost the figures above do not show. Trust rebuilds on a 2-5 year lag behind demonstrated delivery — the [credibility death spiral](/diagnosis/l2-credibility-death-spiral/) is a stock that must be replenished before health boards engage with national digital programmes again, and stocks fill slowly. The longer DHCW remains in its current state, the deeper the credibility deficit that has to be repaid. Health-board reluctance to adopt national systems, clinician disengagement from digital tooling, the loss of an entire generation of technical staff who learned that expertise did not matter — these are themselves costs of having allowed the spiral to deepen. They appear in no DHCW budget line. They appear in every clinical interaction across NHS Wales.


## Who Guards the Guardians?
URL: https://bluenhs.org/blueprint/governance/

_A design principle for distributed oversight. Welsh Government cannot be the sole guardian — it materially contributed to the failure conditions. No single actor is trusted with the whole transition; capture of any one channel does not neutralise accountability entirely._

There is a genuine contradiction at the heart of this blueprint. The analysis argues that Welsh Government oversight of DHCW has been weak, intermittent, and partly captured. The interventions proposed above require competent oversight of the transition. If the oversight body itself is part of the problem, who oversees the reform?

The honest answer: no single actor can be fully trusted. Welsh Government cannot be the sole guardian, because Welsh Government materially contributed to the failure conditions — pressured RAG ratings, mid-cycle DPIF cuts, capital funding refusals while milestones were retained, programme compression, recruitment freezes via remit letter. The design must distribute oversight across multiple actors with overlapping jurisdiction so that capture of any one channel does not neutralise accountability entirely. And the design must include reform of the Welsh Government function itself, alongside DHCW reform, via [Intervention 6: Reform the Funder](/interventions/reform-the-funder/).

**Actors who can operate independently of Welsh Government executive capacity:**

- **Senedd Public Accounts Committee** -- can mandate hearings, require testimony, commission Audit Wales investigations. A credible oversight actor but a part-time one; realistic role is periodic political accountability, not day-to-day transition management.
- **Audit Wales** -- has statutory powers to access documents and examine accounts. Would need to commission specialist digital expertise for technical delivery assessment, which is feasible but not its current capability.
- **Statutory publication requirements** -- legislation requiring DHCW to publish specific data categories on a fixed schedule removes the need for ongoing enforcement. Non-compliance becomes a legal matter, not a governance negotiation. This is the highest-leverage structural mitigation because it operates automatically once enacted.
- **Employment Tribunal proceedings** -- operate entirely outside DHCW and Welsh Government control. Tribunal judgments are public; disclosure orders have compulsory force. May surface evidence that forces accountability regardless of political appetite.
- **Information Commissioner's Office** -- has enforcement powers over FOI compliance. Repeated FOI obstruction (documented at [L10: The Information Fortress](/diagnosis/l10-information-fortress/)) is appealable to the ICO; sustained non-compliance attracts enforcement notices.
- **Misconduct in Public Office** -- the Attorney General's Reference (No 3 of 2003) test is "wilful misconduct serious enough to amount to an abuse of the public's trust." The [L11](/diagnosis/l11-oversight-obstruction/) finding documents the threshold being met by named individuals; criminal investigation is not within Welsh Government's gift to grant or refuse.
- **NHS Counter Fraud Authority and Action Fraud** -- procurement irregularities, sole-bidder anomalies, and undisclosed contract values fall within their remit independently of DHCW or Welsh Government oversight.

**The design principle:** route oversight through multiple independent channels rather than a single supervisory body. More complex than routing everything through Welsh Government, but more robust against the specific failure mode this analysis has identified -- capture of the accountability mechanism by the people it is meant to hold accountable.

## The Board That Replaces the Captured One

Distributed oversight at the regulatory level only matters if the board it oversees is structurally capable of receiving and acting on the signal. The diagnosis documents a board that approved without scrutiny on 51 occasions, with three sub-committees that produced zero corrective actions across eighteen consecutive months and were blind to imminent escalation. Reform of the regulatory architecture above DHCW is necessary but not sufficient.

[Intervention 1](/interventions/competent-leadership/) accordingly includes a **non-executive board reset**: NED competence audit against published criteria; sub-committee reform with technical NEDs; structural protection of the kinds of governance challenge Rowan Gardner, Simon Jones, and Ruth Glazzard offered before they departed. Replacing the executive without reforming the non-executive cohort reproduces the same dynamic — a board that nods through what an executive presents.

## The EPMA Gap

A second structural finding: DHCW is held accountable for milestones it does not control. EPMA is the clearest example — the national programme is a coordinating function, but local organisations are the accountable delivery body (Evans, on the record at PDC May 2025). DHCW is measured; the actors who can deliver are not. Welsh Government is the only body that can resolve this triangle, and the resolution belongs in [Intervention 6: Reform the Funder](/interventions/reform-the-funder/) rather than in distributed-oversight design — because no oversight architecture can correct accountability that has been written incoherently in the first place.


## Monitoring Framework
URL: https://bluenhs.org/blueprint/monitoring/

_Four quadrants tracked monthly. A decision tree that distinguishes premature abandonment from genuine failure. The critical signal: one health board saying 'DHCW delivered what they promised._

Four quadrants, tracked monthly:

**Quarterly assessment:** Has Cluster B been broken? Which Cluster A loop is currently strongest? Has [L2: Credibility Spiral](/diagnosis/l2-credibility-death-spiral/) reversed? Are new loops emerging?

**Red flags:** Dashboard launch delayed past month 9. "Context" columns added to soften delivery numbers. Embedded team leads hired from within DHCW. Multi-year funding allocated to the organisation rather than to programmes. Curation ratio of published board minutes drops below 60%. Any new "declared nil" entry against a director with a known undisclosed interest. "Receive and note" cadence above 50% of substantive board agenda items. Audit Wales positive assessment within six months of a worsening of measurable indicators (the captured-assurance pattern from 2025). Approval-without-scrutiny on any contract above £5M. Recruitment freezes paired with milestone tightening (the 2026 remit-letter pattern). Any of these signals that the reform has been captured by the system it was designed to change.

**Reform-the-Funder leading indicators (Intervention 6):** RAG-rating audit trail visible to the Senedd Public Accounts Committee monthly, with every rating change logged by author, date, and rationale. Capital-allocation timing relative to programme inception (capital settled before fiscal year start, not later). Remit-letter content scored against milestone load — any reduction in input authority paired with proportionate output relief. Frequency of Welsh Government named in published programme post-mortems alongside DHCW.

**When an intervention appears not to be working,** use this decision tree before concluding it has failed:

1. Has the intervention been in place for less than its expected delay period? (Trust rebuilds on a 2-5 year lag; embedded teams need 6-12 months to ship.) If yes — check leading indicators. If leading indicators are moving in the right direction, the intervention is working. Be patient.
2. Are leading indicators flat or declining? Check whether Cluster B is blocking — is information being suppressed about the intervention's progress? Is the intervention being undermined from within? If yes — the problem is not the intervention. It is incomplete implementation of Intervention 1.
3. Are leading indicators flat with no evidence of Cluster B blocking? The intervention may need adjustment. Commission an independent review of that specific intervention, not a wholesale rethink.

The most likely failure mode is premature abandonment: people judge the reform too early, see no results, and abandon it. Every delay in this system is measured in years, not quarters. The decision tree pre-empts this.

**The critical signal:** One health board saying "DHCW delivered what they promised." That is the single most important indicator. Everything else is leading. That is the outcome.


## One-Page Brief
URL: https://bluenhs.org/blueprint/brief/

_The blueprint in one scrollable page. Six interventions, dependencies, 36-month timeline, £5-15M to reform against £3-10B NHS Wales 5-year cost of inaction, printable._

## The Situation

DHCW — NHS Wales' national digital body — has grown its annual budget to **~£200M in 2025-26**, with cumulative Welsh Government funding exceeding £600M over five years and **£0.5M of quantified delivered value across the full period**. That is **83p per £1,000 invested**, and the rate has worsened year on year as headcount and budget have grown faster than delivery. Twelve months of [Level 3 enhanced monitoring](/glossary/#level-3-enhanced-monitoring-and-the-wg-5-level-framework) produced no de-escalation; in 2026 DHCW was escalated further to Level 4 Targeted Intervention. As of March 2026, two of forty-seven Phase One milestones had been missed (LIMS and WRISTS); 42% of IMRP milestones were on track at the last published audit (July 2025); £32.9M of DPIF revenue and £13.1M of capital remained unallocated. Pre-DHCW UWTSD Professor of Practice titles awarded to three executive directors in December 2020 — four months before DHCW's founding board — were never declared at any board meeting through to escalation. The cause is structural: 11 reinforcing feedback loops in two clusters, protected by all seven of Meadows' recognised system traps. No documented precedent in public-sector digital delivery.

## The Plan

Five sequenced structural interventions, targeting Meadows' leverage points 2-6 (system rules, information flows, self-organisation, goals, paradigm) rather than the shallow Levels 10-12 (headcount, reorganisation, strategy documents) where every prior fix has absorbed.

| # | Intervention | Leverage | Owner | Timeline | Addresses |
|---|---|---|---|---|---|
| 1 | **Competent Leadership** | L2 + L5 | Welsh Gov / Audit Wales | Months 0-6 | Cluster B self-preservation (L7, L8, L9, L10, L11); patient-safety triage; board reset; patronage pipeline disclosure |
| 2 | **Radical Transparency** | L6 | New leadership | Months 6-9 | L6 (manufactured narrative), L10 (information fortress); anti-sanitisation protocol; statutory publication |
| 3 | **Portfolio Ruthlessness** | L7 | Independent panel + new leadership | Months 6-18 | L1 (hiring trap), tragedy-of-the-commons; the stop list |
| 4 | **Flip the Model** | L5 + L4 + L3 | New leadership + health boards | Months 6-24 | L2 (credibility), L5 (vendor dependency); £1.25B contract recovery; standards-body endpoint |
| 5 | **Break the Annual Trap** | L5 | Welsh Gov | Months 6-18 | L3 (funding uncertainty); programme envelopes |
| 6 | **Reform the Funder** | L5 + L3 | Welsh Gov / Senedd PAC | Months 0-12 (parallel) | L3, L11; WG-as-co-conspirator; RAG honesty; capital-revenue coherence |

The five DHCW interventions (1-5) depend on Competent Leadership being complete first; interventions 2-5 run in parallel from Month 6. The sixth intervention — Reform the Funder — runs in parallel from Month 0, owned by Welsh Government and the Senedd Public Accounts Committee rather than DHCW. Without it, the same conditions that produced DHCW's failure reproduce for any successor body.

## The 36-Month Timeline

- **Months 0-3** — Independent forensic review commissioned. Patient safety triage of live national systems (WPAS, eMPI, WCCG, WICIS, RISP, PSBA-class infrastructure). Audit Wales mandates data publication. Protected staff reporting channel established. Senedd PAC opens parallel Reform-the-Funder workstream (RAG audit trail, capital-and-revenue coherence rules, remit-letter discipline).
- **Months 3-6** — Review findings delivered. Leadership change decisions made. Non-executive board audit complete; sub-committees reconstituted with technical NEDs.
- **Months 6-12** — New leadership appointed. Transparency dashboard launched against statutory publication categories. Independent panel selects 3 priority programmes against patient-safety-weighted criteria; published stop list binding for 24 months. Two pilot health boards embedded. Vendor portfolio audit completed; sole-bidder contracts above £1M re-tendered or terminated.
- **Months 12-24** — Focused programmes delivering. Embedded teams shipping. Genuine post-mortems on WCCIS and OpenEyes completed. Trust recovery begins.
- **Months 24-36** — Embedded model scaled. Portfolio cautiously expanding on proven demand. Organisation identity transitioning from "national IT authority" to "clinical digital enabler."

Full detail: [the 36-month timeline](/blueprint/timeline/).

## The Economics

|  | Planned reform | Status quo |
|---|---|---|
| Year 1 | £5-15M one-off investment | £100-150M direct DHCW waste |
| Year 2-5 | Cumulative £500M-£1B saved | Cumulative £500M-£1B wasted |
| Downstream NHS Wales (5y) | £3-10B saved | £3-10B cost |
| Patient harm | Halts adding to the ledger | Continues to accrue |
| Crisis risk | Contained | Near-certain crisis-forcing event |
| Reform cost | £5-15M | 2-3× once forced |

Break-even on the planned reform: **in weeks**, not months. Year-one direct DHCW waste (£100-150M) already exceeds the one-off reform cost by an order of magnitude. The annual waste rate is anchored against a ~£200M annual budget that produced £0.5M of quantified delivered value across the full five-year period; the earlier £25-40M/year estimate was a conservative floor built on a single waste category (re-procurement savings).

**The figures above are direct DHCW waste only.** Every DHCW failure cascades outward to seven health boards and NHS Wales as a whole: clinician time lost to broken systems, patient-safety incidents, cross-border referrals, duplicated shadow IT. The downstream cost is plausibly **a 5-15× multiple** of the direct figures — precisely quantifiable only once [L10: The Information Fortress](/diagnosis/l10-information-fortress/) is broken. The March 2026 PSBA outage made the multiplier visible in a single event: a few hours of supplier-level failure produced a full operational day's worth of lost capacity across every NHS Wales organisation simultaneously.

**Total five-year impact across NHS Wales: £3-10 billion.** £500M-£1B direct × a 5-15× downstream multiplier = £3-10B of total NHS Wales five-year cost in status quo, or savings under planned reform. Reform is not a DHCW-internal saving. It is an NHS-wide dividend on a scale comparable to a small acute trust's annual budget — every year.

**Patient harm is on a separate ledger.** WPAS has been linked to at least one patient death. The electronic master patient index has mixed up patient records. WCCG ran on unsupported technology for more than eight years. The Royal Colleges find that patients "regularly experience delays that lead to worsening health." These entries cannot be undone by future reform. Every additional month of the status quo adds to a ledger that money cannot reverse.

Full detail: [the cost of inaction](/blueprint/cost-of-inaction/).

## The Single Decision

Planned reform or crisis-forced reform. These are the only two options. The Welsh choice is the sequencing: now, at £100-150M/year direct DHCW waste plus a 5-15× multiple in downstream NHS Wales impact — or later, after a crisis-forcing event, at 2-3× the direct cost and an even larger downstream tax. Patient harm continues to accrue on a separate ledger that money cannot reverse, regardless of when reform happens; it just stops accruing sooner.

What does a crisis-forcing event look like? On the current trajectory: a PSBA-class shared infrastructure failure scaled beyond a single day's outage; a documented patient-safety incident traced to a known-deferred remediation; an Audit Wales special report; sustained Senedd-led political escalation that forces unscheduled leadership departure; or a successful Employment Tribunal judgment whose factual findings are incompatible with continued tenure. The PSBA outage of March 2026 — when O365, EPMA, RISP, and radiology went offline simultaneously across all NHS Wales — was a precursor, not the event itself.

## For the Sceptical

- The international case: [Alternatives](/blueprint/alternatives/) — Denmark, Estonia, NHS Digital England built better digital health with competent leadership. The DHCW monopoly model is a Welsh-specific choice, not an industry norm.
- The systems-thinking case: [Methodology](/methodology/) — stocks, flows, loops, delays, traps. A forty-year discipline.
- The evidence: [carenhs.org](https://carenhs.org) — primary documents, witness testimony, the full record.

## What You Can Do

- **Share** this brief with your Senedd Member, your health board, your colleagues.
- **Support** [carenhs.org](https://carenhs.org) — the campaign for accountability in NHS Wales.
- **Use** the evidence — every claim is auditable, every source is cited, every conclusion is falsifiable.

The diagnosis is structural. The prescription is specific. The only missing ingredient is the political will to act.


# The Architecture of Failure
URL: https://bluenhs.org/diagnosis/

_The full-system picture. Cluster A drives delivery failure. Cluster B protects the failure from correction. External oversight is blocked by Cluster B before it can reach Cluster A. Five structural loops, six self-preservation loops, one architecture._

Eleven feedback loops, two clusters, one architecture. The five loops of [Cluster A](/diagnosis/feedback-loops/#cluster-a) produce delivery failure. The six loops of [Cluster B](/diagnosis/feedback-loops/#cluster-b) protect that failure from every corrective mechanism the system has been given. The diagram below shows how.

![Architecture of Failure](/img/diagrams/architecture-of-failure.svg)

## The Missing Thermostat

Every complex organisation needs a balancing loop — a corrective cycle, like a thermostat, that detects problems and triggers reform. In a healthy system, the cycle works like this: delivery failures are visible, oversight bodies investigate, accountability is enforced, leadership changes or adapts, structural conditions improve, and delivery recovers.

DHCW has this mechanism. It is called Level 3 — the formal "enhanced monitoring" tier of the [Welsh Government 5-level NHS escalation framework](/glossary/#level-3-enhanced-monitoring-and-the-wg-5-level-framework). DHCW spent over twelve months at Level 3 — the first Welsh NHS body ever held there — without producing any leadership change or structural reform. In 2026, DHCW was escalated further, to Level 4 Targeted Intervention. Even that has not corrected the underlying failure: the Cabinet Secretary publicly declared the entire framework "complex, data-heavy, burdensome, lacks transparency and does not drive improvement." **The thermostat has been captured at every level the framework defines.**

This is the central finding of the analysis: it is not that DHCW lacks corrective mechanisms. It is that Cluster B has neutralised every one of them. And the neutralisation was not gradual: every governance deficit pattern observed at Level 3 escalation 34 months in was already operational at the very first board meeting. The architecture was complete on day one.

## How the Corrective Pathways Are Blocked

Six specific mechanisms intercept every corrective pathway before it can reach the delivery failures in Cluster A.

1. [L9: Whistleblower Suppression](/diagnosis/l9-whistleblower-suppression/) — internal signals of failure are suppressed. Concern-raisers are punished. Disciplinary processes are weaponised.
2. [L10: Information Fortress](/diagnosis/l10-information-fortress/) — external signals are blocked. The public-facing website is inaccessible. Zero accountability data is published.
3. [L6: Manufactured Narrative](/diagnosis/l6-manufactured-narrative/) — internal perception is managed. Staff satisfaction surveys report 80% approval while the organisation sits under Level 3 escalation.
4. [L7: Competence Void](/diagnosis/l7-competence-void/) — the people who could diagnose the problem are overridden, sidelined, or leave.
5. [L8: Loyalty Selection](/diagnosis/l8-loyalty-selection/) — the people who replace them are selected for loyalty, not competence.
6. [L11: Oversight Obstruction](/diagnosis/l11-oversight-obstruction/) — the external oversight function itself is actively degraded from inside the organisation.

Each corrective signal — whether from a clinician, a health board, an auditor, or a Senedd committee — encounters at least one of these six loops. Most encounter several. None reaches Cluster A intact.

## Why No Single Fix Works

A reform aimed at any single loop is absorbed by the others. Fire the CEO? L8 selects the next one for loyalty. Demand transparency? L10 releases aggregated numbers that manufacture the opposite narrative. Protect whistleblowers on paper? L9 redefines the complaint as a conduct issue. Commission an external review? L11 controls the terms of reference.

This is why headcount grew 80%, spend reached £600M, Level 3 Enhanced Monitoring passed its first anniversary, and DHCW was escalated to Level 4 in 2026 — and the underlying failure remains intact. The interventions targeted Cluster A symptoms while Cluster B remained untouched. The architecture absorbed each attempted correction and returned to its prior state.

## What the Diagnosis Is Built On

The eleven feedback loops are derived from a knowledge graph of 1,779 nodes and 3,427 edges, built from 61 DHCW board and committee meeting transcripts across five years (April 2021 – March 2026), and supplemented by Welsh Government performance escalation correspondence, Audit Wales structured assessments, FOI disclosures, and the ABUHB dossier (123 sources).

The structural finding repeats across the timeline. Pre-DHCW UWTSD Professor of Practice titles awarded to three executive directors in December 2020 — four months before DHCW's founding board — were never declared at any board meeting through to escalation. PoP non-declaration is not a deterioration. It is the default configuration from day one.

By edge count: 51 instances of board approval without scrutiny, 45 instances of undeclared interests at specific meetings, 33 "nil" declarations made while undisclosed Professor of Practice titles were held (Thomas 10, Evans 10, Hurle 8, Hall 5), 63 instances of failure to act on a flagged risk, 107 documented sanitisation events, 237 further passages identified as hiding-intent.

## Why the Blueprint Is Sequenced the Way It Is

The [blueprint](/blueprint/) treats Cluster B as the prerequisite, not the follow-up. [Intervention 1: Competent Leadership](/interventions/competent-leadership/) removes the people currently operating Cluster B — forensic review, mandated data publication, protected reporting channel, leadership change where the evidence demands it. Only then can [Interventions 2-5](/blueprint/) address Cluster A. Any other sequence is absorbed.

This is the analytical claim that makes the blueprint different from the dozens of prior reform attempts documented at [carenhs.org](https://carenhs.org): **the structural problem and the self-preservation engine that guards it are both present, and you have to dismantle the second before you can fix the first.**

The system is not failing despite leadership's efforts. It is succeeding at the wrong goal.


## The Hiring Trap
URL: https://bluenhs.org/diagnosis/l1-hiring-trap/

_DHCW grew from 675 to roughly 1,263 staff — an increase of 80%. Delivery got worse. Brooks's Law explains why adding people to a struggling programme makes it struggle more._

When delivery falls behind, the instinctive response is to hire. DHCW grew from 675 to roughly 1,263 staff -- an increase of 80%, with 22.6% growth in the last year alone.

## What is the Hiring Trap at DHCW?
Brooks's Law, documented in 1975, explains why adding people to a late programme makes it later. Each new hire pulls experienced staff away from delivery and into onboarding and coordination. Communication overhead grows with the square of team size. In a specialised domain, new hires take 6-18 months to become productive.

The headcount appears on the dashboard immediately. The productivity does not.

Delivery capability drops despite more people. Political pressure intensifies. The response: hire even more.

## How It Manifests at DHCW

At DHCW, the problem goes beyond Brooks's Law. The evidence from [L8: The Loyalty Selection Loop](/diagnosis/l8-loyalty-selection/) shows that hiring prioritises loyalty over competence. New hires don't just add coordination overhead -- they add *negative* delivery value because they lack the skills the role requires. The 80% headcount growth isn't diluting capability. It may be actively degrading it.

The CEO was asked what patient benefit corresponded to a 25% workforce expansion. Her answer: "It would be lovely to sit here and be able to demonstrate the value."

The Finance Director of the Year (2022 award) could cite only £0.5M in non-cash time savings from ~£600M of Welsh Government funding -- 83p of delivered value for every £1,000 invested. Nine programmes simultaneously under Level 3 targeted intervention -- despite the workforce nearly doubling.

The headcount itself is partly fiction. Vacancy savings have been built into DHCW's financial plans every year since founding, hollowing out the very organisation the headcount figure claims to describe. By month six of Year 1 (2021-22), there was already a £456K vacancy underspend. In Year 2, vacancy savings were explicitly designed into the financial plan; Chair Simon Jones warned: "you just heap misery on misery every year when you do that" — a warning that was sanitised from the published minutes. By Year 3, non-recurrent savings (predominantly from unfilled posts) were 54% of the total savings target. By Q1 of Year 4, 84% of the in-year savings target had been delivered through vacancies. By Year 5, the cross-meeting arc was confirmed: vacancies produced overload, overload produced burnout (65%+ in the staff survey), burnout produced delivery failure, and delivery failure produced first Level 3 enhanced monitoring and then, in 2026, Level 4 Targeted Intervention. As CFO Claire Osmundsen-Little put it before her departure to Swansea Bay: staff "can't finish at 5pm."

The 80% headcount growth is therefore an upper-bound figure. The actual filled-post number is lower, deliberately so, every year. By July 2025, only 42% of IMRP milestones were on track. This is the headcount-but-no-output paradox that Brooks's Law describes — but in DHCW's case it is not just an emergent property of growth. It is a financial design choice.

Beyond the formal headcount, 23 off-payroll workers operate as a shadow workforce making healthcare decisions outside the normal accountability structures. In a healthy governance system, every person making decisions that affect patients would be visible, accountable, and audited.

The hiring trap feeds directly into [L5: The Vendor Dependency Spiral](/diagnosis/l5-vendor-dependency-spiral/). Staff who lack capability cannot deliver, so work gets outsourced. Outsourced work does not build internal capability. The headcount grows, the vendor portfolio grows, and delivery stays flat.

## What would a healthy alternative look like?

Governance tracks delivery per person, not headcount. Hiring targets specific skill gaps identified by independent audit. New hires are measured against delivery contribution within 12 months. If output per person declines despite growth, the hiring strategy is revised — not doubled down on.

## How does the blueprint break the Hiring Trap?

Breaking the hiring trap requires two interventions acting together. [Portfolio Ruthlessness](/interventions/portfolio-ruthlessness/) stops the demand spiral: fewer programmes mean headcount can be matched to work the organisation can realistically deliver, rather than scaled to mask political pressure. [Flip the Model](/interventions/flip-the-model/) changes what hiring is for: internal capability build, not vendor coordination overhead. Together they convert headcount from a political signal back into a delivery input.


## The Credibility Death Spiral
URL: https://bluenhs.org/diagnosis/l2-credibility-death-spiral/

_Trust is destroyed instantly by a missed deadline but takes 2-5 years of consistent delivery to rebuild. DHCW's CEO admitted in January 2026 that timelines presented to Senedd were over-optimistic._

Programmes miss deadlines. Health boards lose trust. They resist adopting DHCW systems. Low adoption means DHCW cannot demonstrate value. The programme looks like a failure, attracting more scrutiny. Under pressure, DHCW promises ambitious timelines to show progress. It misses those too.

## What is the Credibility Death Spiral at DHCW?
Trust is destroyed in a single missed deadline. Rebuilding it takes 2-5 years of consistent delivery. One cycle of over-promising and under-delivering can take half a decade to recover from -- even if delivery improves immediately.

The asymmetry between how quickly trust is destroyed and how slowly it is rebuilt is the core of the loop: every broken promise compounds the next.

## How It Manifests at DHCW

In January 2026, the CEO admitted that timelines presented to Senedd in 2023 were over-optimistic, conceding that "we need to do more discovery" -- meaning the original plans were committed to before the organisation even understood the problem properly.

This is not missing deadlines. This is presenting timelines leadership knew, or should have known, were unreliable.

Audit Wales found "reluctance" and "trepidation" among health boards engaging with DHCW. The words are diplomatic. The reality is worse.

WCCIS: abandoned by multiple organisations. The NHS Wales App: 345,000 registered users, but the Deputy CEO of NHS Wales admitted "hardly anybody is using it regularly." OpenEyes: £8.5M spent, four years late, live in only 2 health boards after six years.

The credibility problem is not just at the Senedd level. Even within DHCW's own milestone tracking, the signals are unreliable. In July 2024, Ifan Evans named what he called the "biggest disappointment": "Milestone owners collectively failed to flag slippage, and by the time this was recognized, the impact was unavoidable... milestone owners said 'yes on track' when they were not." No milestone owner was named. No corrective mechanism was established. The same admission identified the underlying cause: "Confidence of delivery is not currently measured." Five hundred and fifty milestones were being tracked. Zero of them had a confidence score attached.

A pattern followed. By July 2025, DHCW was self-marking milestones green internally before Welsh Government had formally confirmed completion (PDC, 10 July 2025: "Is that with agreement with Welsh government or is that us just saying it's complete? Yeah, that's a good question, David. So we've turned those to green because they were completed."). All twelve May-June milestones reported green before IQPD formal confirmation on 16 July. Six months earlier, Evans had described the underlying behavioural pattern at board: "people for understandable behavioural reasons are optimistic... the style is we're gonna make it, we're gonna make it, no we're not." The pattern was named at the top of the organisation, then continued unchanged.

Health boards see this. So do oversight bodies. The credibility cost is paid not in one missed deadline but in a steady erosion: every time a milestone reported green turns out red, every previous green report becomes suspect.

The credibility spiral interacts with [L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/). Internally, leadership maintains an "80% satisfaction" culture. Externally, health boards see missed deadlines, abandoned systems, and products nobody uses. The gap signals one of two things: either DHCW leadership doesn't know how bad things are, or they don't care. Neither builds trust.

## What would a healthy alternative look like?

Governance requires conservative timelines validated by independent technical assessment before commitments are made to Senedd or health boards. Delivery is demonstrated before the next promise. When a deadline is missed, the post-mortem is published. Trust recovers naturally because stakeholders see honesty, not theatre.

## How does the blueprint break the Credibility Death Spiral?

Rebuilding trust requires changing what goes out the door. [Radical Transparency](/interventions/radical-transparency/) closes the gap between internal narrative and external reality — if health boards can see delivery data directly, the theatre stops working and the asymmetry between promise and outcome becomes visible in real time. [Flip the Model](/interventions/flip-the-model/) creates the underlying delivery capability, so that the commitments made under transparency can actually be met. Transparency without delivery accelerates the spiral; transparency with delivery reverses it.


## The Funding Uncertainty Trap
URL: https://bluenhs.org/diagnosis/l3-funding-uncertainty-trap/

_Annual budget cycles drive short-term contracts, which drive staff turnover, which destroys programme knowledge. Confirmed by Audit Wales._

Welsh Government funds DHCW on annual budget cycles. Annual funding drives short-term contracts. Short-term contracts drive staff turnover. Turnover destroys programme knowledge that takes years to rebuild.

## What is the Funding Uncertainty Trap at DHCW?
When knowledge walks out the door, programmes stall. Stalled programmes produce poor performance reviews. Poor performance makes multi-year funding commitments politically risky. So funding stays annual.

The loop is self-sustaining: annual funding produces the exact failure signal that makes multi-year funding look irresponsible, which keeps funding annual.

## How It Manifests at DHCW

Audit Wales has confirmed this cycle. No multi-year programme funding settlements are in place.

But the funding trap is also weaponised as an excuse. "We can't deliver because of annual funding" conveniently deflects from the harder truth: we can't deliver because our leaders can't manage technology programmes. The funding argument is the first line of defence whenever delivery is questioned.

Beyond the annual cycle, Welsh Government has actively worsened the funding picture rather than mitigated it. WG pressured DHCW to soften red RAG ratings — Evans, Performance and Delivery Committee, November 2024: "Government was asking... why then are those programs rated red amber? If you've got a reasonable confidence that you will hold the implementation dates." It cut DPIF from £33M planned to £28M allocated for 2024-25. In one quarter, 30% of the DPIF was withheld — Chair Simon Jones called this "very strange," said he had "never experienced" anything like it. WG refused capital funding for e-referrals and the integration hub in November 2025, then set milestones requiring those very systems. Funding letters were not confirmed until 25% through the financial year (July 2022). DPIF itself, Evans admitted, had "expanded to fit things that are arguably not that transformational. Not that transformational like Link and RISP."

This does not vacate the leadership critique. It adds a structural finding alongside it: the entity issuing DHCW's corrective signals is materially co-authoring the conditions DHCW is being escalated for. Glazzard, five years in: "We've always struggled because it's one-year funding. We should have worked it out by now, surely."

The funding trap is compounded by [L8: The Loyalty Selection Loop](/diagnosis/l8-loyalty-selection/). When knowledgeable staff leave due to contract uncertainty, they should be replaced by the best available candidates. Instead, they are replaced by loyalists who lack the departing staff's expertise.

The knowledge loss becomes permanent, not temporary.

Denmark's Sundhedsdatastyrelsen and Estonia's TEHIK serve comparable populations with modern digital health systems, funded through similar government mechanisms. They face the same constraints. They deliver. DHCW does not.

Annual funding is a real structural problem. It is not the reason DHCW is the worst performer among comparable organisations.

By March 2026, the funding-trap end-state was visible: £32.9M of DPIF revenue and £13.1M of capital remained unallocated; an Accountable Officer letter had been sent stating delivery was "not possible without confirmed DPIF allocation"; recruitment was frozen by remit letter. The structural problem this loop describes is not theoretical. It is the operational status quo.

## What would a healthy alternative look like?

Multi-year funding is tied to specific programmes with externally verified milestones — not to the organisation. Programme continuity survives leadership changes. Staff are retained on permanent contracts because the funding horizon allows it. Institutional knowledge accumulates instead of draining.

## How does the blueprint break the Funding Uncertainty Trap?

The structural fix is [Break the Annual Trap](/interventions/break-annual-trap/): multi-year funding envelopes tied to programmes and verified milestones, not to the organisation. This severs the mechanism that makes short-term contracts rational in the first place. Knowledge retention stops requiring heroism from individual programme leads and starts being a property of the funding design.


## The Rebranding Escape
URL: https://bluenhs.org/diagnosis/l4-rebranding-escape/

_When a programme fails, DHCW rebrands it. New name, new slides, same approach, same structural problems. WCCIS became Connecting Care after £42M+ and deployment to only 19 of 29 targets._

When a programme fails, the instinctive institutional response is not to understand why. It is to rename it. New slide deck, same team, same structural problems. The rebrand absorbs the energy that should go into an honest post-mortem. No lessons are extracted. The renamed programme fails for the same reasons.

## What is the Rebranding Escape at DHCW?
Donella Meadows called this "Shifting the Burden." The real fix -- an honest post-mortem to understand structural causes of failure -- is painful and politically costly. The quick fix -- rebrand and restart -- is painless and creates the appearance of progress. But each quick fix erodes the organisation's ability to learn. The structural causes are never identified, never addressed, and never prevented from recurring.

## How It Manifests at DHCW

WCCIS (Welsh Community Care Information System) is the defining example. Over £42M spent. Deployed to only 19 of 29 target organisations. Multiple organisations abandoned it. It was called "live," then quietly rebranded to "Connecting Care." As of 2026, the business case for its replacement is still pending -- four years after the rebrand, they still don't have a plan.

At DHCW, the rebranding escape is protected by [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/). Honest post-mortems require candid internal testimony about what went wrong. When staff who raise concerns are dismissed, smeared, or subjected to fake disciplinary processes, no one will testify truthfully.

The organisation's capacity to learn from failure has been structurally destroyed.

The latest iteration is "Building our Future" -- a reorganisation that follows the same pattern. Structural cosmetics absorbing change energy without changing the system.

The pattern operates at the strategic level too, not just at the programme level. By January 2025, Neill challenged Evans at board: how could an organisation of 800 people credibly operationalise the 14-15 sub-strategies it had simultaneously declared? Doyle had asked the same question in DHCW's founding year, requesting the board understand "what to stop doing." Neither challenge produced a stop list. Each new strategic framework was added to the existing pile and announced as the framework that would unify the others.

The largest rebrand in flight is at the top of the strategic stack. Ifan Evans authored *A Healthier Wales* — the strategy DHCW was created to implement, then failed to deliver. He is now designing the next ten-year digital strategy for NHS Wales. The successor is being authored by the failed author of the predecessor. This is the rebrand reflex at its largest possible scale: the failure of strategy one is being relabelled as strategy two, by the same hand, with the same patronage architecture intact.

## What would a healthy alternative look like?

When a programme fails, governance mandates a genuine post-mortem with external facilitation. Findings are published. Structural causes are identified and addressed before any successor programme is funded. The people who led the failure are not automatically given the successor. Learning is treated as an institutional asset, not a political liability.

## How does the blueprint break the Rebranding Escape?

The rebrand works only because the real record is hidden. [Radical Transparency](/interventions/radical-transparency/) removes the cover: published post-mortems, published delivery status, published successor business cases. When the record is public, renaming a failed programme cannot erase the history it was supposed to escape. The incentive to rebrand collapses when the audit trail survives.


## 18 Stocks
URL: https://bluenhs.org/diagnosis/stocks/

_What governance should be tracking — and what it actually tracks. Eighteen stocks organised by visibility reveal the gap between the numbers DHCW reports and the reality it hides._


## 11 Feedback Loops
URL: https://bluenhs.org/diagnosis/feedback-loops/

_Eleven structural feedback loops explain why DHCW fails and why no one fixes it. Five produce delivery failure. Six protect that failure from correction._


## The Vendor Dependency Spiral
URL: https://bluenhs.org/diagnosis/l5-vendor-dependency-spiral/

_DHCW manages a contract portfolio valued at roughly £1.25 billion. With each outsourced programme, internal capability erodes and vendor leverage grows._

When a complex programme arises, the path of least resistance is to outsource it. But outsourcing does not build internal capability. Without internal capability, you cannot evaluate vendor performance or negotiate from strength.

## What is the Vendor Dependency Spiral at DHCW?
The vendor controls timeline and cost. When the next programme arises, you outsource again -- because you still lack the capability you never built. The portfolio grows. The leverage shrinks.

Each outsourced programme reduces the internal skill pool available to assess the next contract, which makes the next contract more likely to be outsourced on worse terms.

## How It Manifests at DHCW

DHCW manages a contract portfolio valued at roughly £1.25 billion. For an organisation whose stated purpose is to deliver digital services for NHS Wales, that number is an indictment.

Multiple contracts have undisclosed values -- a basic transparency failure that should not be possible in a public body. The RISP radiology supplier is unnamed despite a £47-56M contract value. The Channel 3/Aire Logic NTA contract value has never been publicly disclosed. Promptly Health received £11M with no visible business case.

Sole-bidder contracts have become normalised.

The board approval process is itself part of the problem. The knowledge graph documents 51 distinct instances where DHCW's board approved spending without substantive scrutiny. A £20M framework agreement with Kainos was approved in circumstances where the Chair later admitted: "I should have looked. I don't know how these appear on our website as contracts." A £226M Microsoft Enterprise Agreement passed at the March 2026 board meeting in a single sentence — no questions, no vote. The full business cases for LIMS, LINC, and RISP were all moved into private session under "commercial sensitivity," removing scrutiny from the public record. When the largest contracts the organisation signs are approved with less attention than a routine policy update, the vendor sets the terms by default.

This spiral is compounded by [L7: The Competence Void](/diagnosis/l7-competence-void/). Leaders without technology delivery experience cannot evaluate vendor claims -- they cannot tell the difference between a vendor delivering real value and a vendor managing upward. They cannot negotiate effectively because they don't understand what they're buying. This creates an accountability vacuum that vendors exploit -- and rationally should exploit, because there is no counterweight.

The Head of Software Engineering role was advertised at Band 8c (£71-82k) -- well below market rate for the responsibility. At that salary, you will not attract someone who can build the internal capability to challenge a £1.25 billion vendor portfolio. The role was either designed for an internal candidate or it was never serious. Either way, the spiral continues.

The consequence of this dependency was visible in March 2026, when the PSBA network went down across NHS Wales. O365, EPMA, RISP, radiology — all unavailable simultaneously across every health board. A single shared infrastructure dependency, owned and operated by a single supplier, took the whole digital estate offline. There was no fallback. There could not be: the architecture had not been designed to survive the loss of any of its central vendor relationships.

Denmark and Estonia built modern digital health infrastructure with smaller budgets by investing in internal capability and maintaining genuine competitive procurement. Wales has the resources to do the same.

The vendor dependency is a choice, not a constraint.

## What would a healthy alternative look like?

Every outsourced programme includes a mandatory knowledge transfer plan and internal capability build. Contract values are published. Vendor performance is assessed independently. Sole-bidder contracts trigger automatic governance review. Internal technical teams retain enough capability to evaluate vendor claims and negotiate from strength.

## How does the blueprint break the Vendor Dependency Spiral?

The spiral only unwinds when the default choice flips from "outsource" to "build". [Flip the Model](/interventions/flip-the-model/) rebuilds internal technical capability as the primary delivery mechanism, with vendors used for bounded specialist work on published terms. As in-house competence rises, the organisation regains the ability to evaluate vendor claims — which is the precondition for every other procurement and contract reform.


## The Manufactured Narrative
URL: https://bluenhs.org/diagnosis/l6-manufactured-narrative/

_DHCW reports 80% staff satisfaction while under Level 3 enhanced monitoring. carenhs.org blocked on NHS Wales devices. Zero accountability data published. The narrative is not organic — it is manufactured._

DHCW reports 80% staff satisfaction. DHCW is under Level 3 enhanced monitoring for serious delivery failures. Both statements are true. They are not contradictory — if you understand the mechanism: leadership actively controls the information available to staff.

## What is the Manufactured Narrative at DHCW?
A manufactured narrative is not a cultural bubble where staff are organically insulated from reality through awards and institutional pride. It is actively produced. Information that would contradict leadership's story is blocked, unpublished, or reframed. Information that supports the story is amplified.

The manufactured narrative prevents the internal pressure that would otherwise force change. When staff believe the organisation is performing well, they don't demand accountability. When they don't see external criticism, they don't question leadership decisions.

## How It Manifests at DHCW

carenhs.org was blocked on NHS Wales network devices. The site contains only publicly sourced material from Senedd proceedings, Audit Wales reports, and government statements. Blocking a website composed entirely of government-sourced material is not network security. It is deliberate information control.

Zero whistleblowing data is published. Zero disciplinary data is published. Zero staff leavers analysis.

The 80% satisfaction score is both leadership's shield -- "our staff are happy" -- and proof of the deception: our staff don't know what's happening.

The published record is actively curated. The knowledge graph documents 107 distinct sanitisation instances across DHCW's board and committee minutes — passages where what was spoken in the room was substantively altered, softened, or deleted before publication — plus 237 further passages identified as hiding-intent. In one published transcript, the curation ratio fell to 10.7%: 3,680 published words from 34,257 spoken. Sixteen of nineteen speakers had been removed entirely.

The deletions are not random. They follow a consistent pattern: executive admissions of failure are removed, structural financial warnings are removed, independent-member challenges are removed, evidence of deterioration is removed. What survives is what supports the narrative. Specific examples include:

- Helen Thomas describing LIMS as "causing anxiety" — sanitised to "issues" (Sep 2025).
- Carwyn Lloyd-Jones: "We talk about technical debt all the time, but it's an anodyne phrase" — deleted.
- Chair Simon Jones warning that vacancy savings "heap misery on misery every year" — deleted.
- Ifan Evans's "biggest disappointment" admission that milestone owners had said "yes on track when they were not" — deleted.
- Ifan Evans on the "technical debt spiral" — forced prioritisation of maintenance over transformation, problem compounding over time — deleted.

If the published record is what staff and external stakeholders see, and the published record is curated to remove every signal of failure, then the 80% satisfaction figure is not a finding. It is a manufactured artefact of the same machine.

Maintaining the narrative requires the full weight of Cluster B. No one raises internal challenges because [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) makes that career-ending. No external information reaches staff because [L10: The Information Fortress](/diagnosis/l10-information-fortress/) blocks it at the network level. And management positions are filled by people who will reinforce the narrative, not challenge it -- that is [L8: The Loyalty Selection Loop](/diagnosis/l8-loyalty-selection/).

The Interim Chair, Ruth Glazzard, admitted the board learns about problems "late in the day." If the board is receiving curated information, what are the 1,263 staff below the board receiving?

## What would a healthy alternative look like?

Staff have unrestricted access to all public information about their organisation. Satisfaction surveys are published alongside externally assessed delivery metrics — if the two diverge, governance investigates why. No public information source is blocked on internal networks. The gap between internal narrative and external reality is measured and closed, not manufactured.

## How does the blueprint break the Manufactured Narrative?

The narrative is only manufacturable because contradictory information can be suppressed. [Radical Transparency](/interventions/radical-transparency/) publishes the inputs the narrative relies on hiding — whistleblowing counts, disciplinary data, leaver analysis, delivery status against commitments. Once those flow at statutory cadence, internal perception re-anchors to external reality, and the satisfaction-survey-to-Level-3-escalation gap becomes impossible to maintain.


## The Competence Void
URL: https://bluenhs.org/diagnosis/l7-competence-void/

_DHCW is a technology delivery organisation led by executives with no complex technology delivery experience. NHS Digital England recruited from Credit Suisse, HSBC, and Rolls-Royce CIO._

DHCW is a technology delivery organisation. Its leadership team has no complex technology delivery experience.

## What is the Competence Void at DHCW?
When leaders lack technical experience, they cannot evaluate technical recommendations. They override experts with uninformed decisions. The best technical staff learn that expertise does not matter. They leave. Remaining staff lower their standards. The void deepens.

The delay is the killer: existing systems run on legacy knowledge and momentum for years before the consequences surface. By the time leadership's inability is visible externally, the people who could have prevented the collapse have already left.

## How It Manifests at DHCW

The CEO, Helen Thomas, has 30+ years in NHS Wales finance and health information management — not technology delivery. The CDO, Rebecca Cook, has 19 years within NWIS/DHCW — an internal promotion track, not an external technology career. Sam Hall, a director, comes from local government and statistics. Sam Lloyd, Executive Director of Operations, has 19 years in English public health agencies with no devolved Welsh health system experience. Only the Medical Director, Rhidian Hurle, holds postgraduate clinical informatics qualifications.

For comparison, NHS Digital England recruited its C-suite from Credit Suisse, HSBC, the Home Office CDO, Jaguar Land Rover CIO, and Rolls-Royce CIO. The contrast speaks for itself.

Technical teams warned about WCCG running on unsupported technology for 8+ years. Leadership ignored the warnings.

The pattern is wider than WCCG. At month seven, the architecture team explicitly asked the board for support and prioritisation; the board agreed "enthusiastically", and no follow-through occurred. Three years later, in November 2023, the architect Carwyn Lloyd-Jones told the board: "We talk about technical debt all the time, but it's an anodyne phrase." That observation was deleted from the published minutes. The National Target Architecture (NTA) — repeatedly referenced as the framework that would resolve the integration and platform problems — has never been delivered. Cloud readiness was rated "quite red" by Sam Lloyd in a presentation that was subsequently deleted from the meeting record. The competence void is visible not only in what leadership cannot evaluate, but in what it cannot defend long enough to fund.

WPAS was identified as a factor in at least one patient death. A health board described it as the "single biggest risk to patient safety."

The best technical staff learn that expertise does not matter here. They leave. Or they are pushed out -- [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) describes what happens to those who persist in raising technical concerns. The people who could have fixed WCCG, modernised WPAS, or challenged a vendor's inflated claims walk out the door.

The void takes 2-3 years to become visible. Existing systems run on legacy knowledge and momentum. By the time the consequences surface — Level 3 targeted intervention, patient safety incidents, ~£600M of Welsh Government funding with £0.5M of delivered value — the people who could have prevented them have already left or been dismissed.

The competence void also explains why [L5: The Vendor Dependency Spiral](/diagnosis/l5-vendor-dependency-spiral/) is so severe. Leaders who cannot evaluate technology cannot evaluate vendors. They cannot tell the difference between a vendor delivering value and a vendor managing upward. The £1.25 billion contract portfolio is managed by people who lack the technical foundation to hold vendors accountable.

Ifan Evans authored the Welsh Government's digital health strategy, then moved directly to DHCW to oversee its implementation. The strategy author assessing his own work is a structural conflict of interest — and the conflict was visible at the highest levels. Forty-six months after DHCW's founding, Evans admitted on the public record that the basic structure of accountability had never been established: "This persistent ambiguity, let's call it, which we usually phrase as 'we need to clarify roles and responsibilities'... PDC is an oversight and assurance function, not a programme management function" (PDC, February 2025). Four years into a £600M+ programme, the executive director responsible for strategy was conceding that no one knew who was accountable for what. He has now been appointed to design the next ten-year strategy. In an organisation where no one has the technical standing to challenge that appointment, the conflict goes unexamined.

## What would a healthy alternative look like?

Leadership roles in a technology delivery organisation require demonstrable technology delivery experience — verified by independent skills audit, not self-reported. Board recruitment is conducted by external panels with published criteria. When NHS Digital England needed to deliver at scale, it recruited C-suite leaders from Credit Suisse, HSBC, and Rolls-Royce. Governance ensures the people making technical decisions have the competence to evaluate them.

## How does the blueprint break the Competence Void?

No internal initiative can fix a competence void in the leadership that would run it. The entry point is [Competent Leadership](/interventions/competent-leadership/): the leadership producing the failure must be replaced by leadership capable of delivery, selected by an external panel against published technical criteria. Every other intervention depends on this, because every other intervention requires leaders who can evaluate it.


## The Loyalty Selection Loop
URL: https://bluenhs.org/diagnosis/l8-loyalty-selection/

_DHCW leadership was pre-selected from an ABUHB patronage pipeline. Pre-DHCW UWTSD Professor of Practice titles were awarded to the CEO, Medical Director, and Strategy Director — and never declared at any board meeting._

Promote loyal allies over competent candidates. The allies cannot deliver — but they are loyal. They protect the leader from accountability. They don't challenge bad decisions. The leader's position becomes more secure, so more allies are promoted. Delivery capability hollows out while the control structure solidifies.

The organisation grows. Output shrinks.

## What is the Loyalty Selection Loop at DHCW?
Structural dysfunction distributes harm randomly. Patronage distributes harm systematically toward those who threaten the clique's position, and benefit systematically toward those who protect it.

The beneficiary test is the diagnostic: in every decision where delivery is sacrificed, the same small group benefits. The concentration of benefit runs always in the same direction, and always at the expense of the same category of person.

## How It Manifests at DHCW

The selection pattern predates DHCW. Leadership was pre-selected from a patronage pipeline controlled by Andrew Goodall (Welsh Government Permanent Secretary, formerly ABUHB CEO) and Judith Paget (Director General Health & Social Services, formerly ABUHB CEO). Three successive NHS Wales CEOs emerged from a single health board.

In December 2020 — four months before DHCW's founding board meeting — Helen Thomas (CEO), Rhidian Hurle (Medical Director), and Ifan Evans (Executive Director Strategy) were awarded "Professor of Practice" titles by the University of Wales Trinity Saint David, via its Wales Institute of Digital Information (WIDI). The titles were never declared at any board meeting, despite the Chair asking for declarations of interest at every meeting opening.

The pattern is quantified across the full timeline. The knowledge graph documents 33 instances where directors declared "nil" while holding undisclosed Professor of Practice titles — Thomas 10, Evans 10, Hurle 8, Hall 5 — and 45 further instances of undeclared interests at specific meetings. At the very first board meeting (April 2021), Thomas had held the title for 10 months. She declared nil. The Chair's invitation — "Any declarations?" — was met with "No" or "None" at every meeting through to escalation.

This is not deterioration. It is the default configuration from day one.

Multiple independent witnesses describe sham recruitment for key senior roles below the executive: competitive processes conducted as performance theatre while outcomes were predetermined. Selection criteria were reverse-engineered to match specific candidates. External candidates with stronger qualifications were rejected in favour of internal loyalists.

The beneficiary test applies directly: in every DHCW decision where delivery was sacrificed, the same small group benefited — the circle around the CEO — and always at the expense of the same category of person: technical staff who raised concerns. That is not dysfunction. It is patronage.

The Head of Software Engineering -- the most senior hands-on technical role in a national health IT organisation -- was advertised at Band 8c (£71-82k). This is well below market rate for equivalent responsibility in the private sector, other NHS organisations, or UK government digital. Either the role was designed for a predetermined internal candidate, or the process was never serious about attracting external talent. Both lead to the same place.

The CEO accumulated a BCS Fellowship, FedIP registration, the UWTSD Professor of Practice title, and a "Digital CEO of the Year" award — all during the 2020-2021 transition to the permanent CEO role. Multiple internal sources confirm she encouraged staff credential accumulation, reframing institutional positioning as organisational culture.

Meanwhile, directors without disclosed academic qualifications occupy senior technical roles. The pattern is consistent: credentials and advancement flow along loyalty lines.

The Senedd noted in July 2023: "It was not clear to us which skills or departments have been prioritised" in the workforce expansion. When asked what patient benefit corresponded to a 25% workforce increase, the CEO replied: "It would be lovely to sit here and be able to demonstrate the value."

The CEO of a technology delivery organisation could not name a single patient outcome from hiring hundreds of additional staff.

The selection pattern extends beyond permanent staff. 23 off-payroll workers operate as a shadow workforce, making healthcare decisions outside normal accountability structures — invisible to governance scrutiny.

The pipeline continues. Ifan Evans authored *A Healthier Wales* — the Welsh Government strategy DHCW was created to implement — and was seconded to DHCW to oversee its implementation. The implementation has not been delivered. He is now designing the "Blueprint": the next 10-year digital strategy for NHS Wales. The people who failed at the originating health board were promoted to oversee DHCW, where they presided over the same failures. The pipeline that created the leadership structure remains intact.

The loyalty selection loop creates the conditions for every other Cluster B loop. [L7: The Competence Void](/diagnosis/l7-competence-void/) exists because loyalty, not competence, determines who leads. [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) operates because loyalists control HR. [L10: The Information Fortress](/diagnosis/l10-information-fortress/) holds because loyalists control communications. [L11: The Oversight Obstruction Loop](/diagnosis/l11-oversight-obstruction/) works because loyalists occupy the positions through which oversight flows. L8 is the root of the self-preservation engine.

## What would a healthy alternative look like?

Recruitment is competitive, transparent, and externally auditable. Selection criteria are published before shortlisting. Panel members include external specialists. Outcomes are documented and challengeable. Salaries reflect market rates for the role's actual responsibility. When governance is not captured, hiring optimises for the organisation's mission — not for the leader's control.

## How does the blueprint break the Loyalty Selection Loop?

Loyalty selection cannot be unwound by the people it put in place. [Competent Leadership](/interventions/competent-leadership/) removes the selectors: externally run executive recruitment against published criteria, replacing the leadership layer that converted appointment power into a patronage network. Only after that substitution does it become possible to rebuild hiring below the board as merit-based rather than loyalty-based.


## The Whistleblower Suppression Loop
URL: https://bluenhs.org/diagnosis/l9-whistleblower-suppression/

_A technical expert raises a legitimate concern at DHCW. They are subjected to retaliatory action and dismissed. Other staff watch and learn: raising concerns is career-ending._

A technical expert raises a legitimate concern about delivery failures, financial waste, or patient safety.

## What is the Whistleblower Suppression Loop at DHCW?
The expert is smeared. Isolated. Subjected to disciplinary proceedings on pretextual charges. Their work device is confiscated to prevent evidence preservation. They are dismissed.

Other staff watch. The lesson is clear: raising concerns is career-ending. Staff stop raising concerns. Problems go unreported. The board receives only curated information. Problems compound invisibly until they surface externally — through Senedd scrutiny, Employment Tribunal filings, or patient safety incidents.

The loop's speed is what distinguishes it: each visible retaliation produces an immediate chilling effect across the remaining workforce. No gradual erosion is needed.

## How It Manifests at DHCW

This is not alleged. It is witnessed and multiply corroborated.

The whistleblower was dismissed for alleged "gross misconduct" in a process that violated DHCW's own disciplinary policies. If the charge were legitimate, the process would follow the organisation's own rules. It did not. Courts recognise deviation from internal procedures as strong evidence of bad faith -- the charge was pretextual and the process was designed to reach a predetermined outcome.

The CEO and all directors were personally informed of the significant harm and financial loss their actions were causing. They did not claim ignorance. They did not seek further information. They chose to accelerate the removal.

This satisfies the "wilful" element of misconduct in public office under the Attorney General's Reference (No 3 of 2003) test. They knew. They were told. They chose to proceed.

The timing is devastating. The dismissal occurred while DHCW was under active investigation by the Senedd and Welsh Government. The whistleblower's concerns overlapped with the concerns these bodies were investigating. Silencing a source of the very information external investigators need, during an active investigation, is not merely retaliatory — it is obstruction of the oversight function.

Weaponised disciplinary proceedings are systemic, not isolated. Multiple independent witnesses describe the same pattern applied to other technical team members who raised concerns: concern raised, pretextual charges initiated, harassment through the process itself, person leaves or is dismissed. This is not one manager's poor HR practice. It is an organisational tool — the HR function repurposed as a weapon against technical staff who challenge leadership decisions.

The dismissed employee's role was replaced with a downgraded position — lower band, less authority. The oversight function was structurally eliminated. This is [Trap 6: Success to the Successful](/traps/success-to-the-successful/) in action: the loyalty network became more secure, delivery capability became weaker, and the gap widened.

In 2018, the Senedd found NWIS culture was "the antithesis of open" and staff testimony was "pre-prepared." Eight years later, the same culture persists — only now with documented evidence of retaliation. Glassdoor reviews describe a "horrendous culture of bullying with management sweeping any issues under the carpet." The pattern has not changed. It has hardened.

By 2026, DHCW publishes zero whistleblowing disclosure statistics, zero disciplinary proceedings data, zero staff leavers analysis. The absence of data is itself data.

What might have substituted for individual whistleblowing — committee scrutiny, external assurance — was equally absent. The Performance and Delivery Committee generated zero corrective actions across eighteen consecutive months from May 2024 to May 2025, even with a Welsh Government observer in the room. Audit Wales — the external assurance function with statutory access — issued a "good governance, stable and cohesive board" assessment four months before Level 3 enhanced monitoring was triggered. By the time DHCW was escalated further, to Level 4 Targeted Intervention in 2026, the pattern was complete: every internal route for raising a concern had been closed, and every external assurance route had been telling a story that subsequent events disproved within months. Suppression of one whistleblower was effective because the rest of the system was not asking the questions that whistleblower would have answered.

## What would a healthy alternative look like?

Concern-raisers are protected by statute. Their information reaches governance bodies through a channel that cannot be intercepted by the people the concern is about. Whistleblowing data is published annually. Disciplinary processes are audited for patterns of retaliatory use. When someone raises a concern and is then subjected to disciplinary action, the coincidence triggers automatic independent review.

## How does the blueprint break the Whistleblower Suppression Loop?

Retaliation cannot be stopped by the leadership conducting it. [Competent Leadership](/interventions/competent-leadership/) removes the people who weaponised HR against technical dissent and installs leadership accountable under externally verifiable criteria. Only after that substitution does it become possible to restore the internal feedback capacity that whistleblower protections are supposed to guarantee.


## The Information Fortress
URL: https://bluenhs.org/diagnosis/l10-information-fortress/

_carenhs.org blocked on NHS Wales devices. Zero whistleblowing data published. Multiple contract values undisclosed. Board papers published as unsearchable PDFs. Every door barred._

Every door barred. The information fortress is not a single wall but a system of interlocking barriers.

## What is the Information Fortress at DHCW?
An information fortress buys time. Data that would contradict leadership's narrative is withheld, unpublished, or released in formats designed to resist scrutiny. For 6-18 months, the barrier works. But when reality breaks through — escalation, tribunal filing, campaign pressure — the gap between what was reported and what is real creates a crisis of confidence far worse than continuous transparency would have produced.

The cover-up becomes the scandal.

## How It Manifests at DHCW

**Website blocking.** carenhs.org was blocked on NHS Wales network devices. The site contains only publicly sourced material from Senedd proceedings, Audit Wales reports, and government statements. Blocking a website that contains only government-sourced material is not standard network security. It is deliberate information control.

**Data suppression.** Zero whistleblowing disclosure statistics published. Zero disciplinary proceedings data published. Zero staff leavers analysis published. Multiple contract values undisclosed — including the RISP radiology supplier, unnamed despite a £47-56M contract. The Channel 3/Aire Logic NTA contract value has never been publicly disclosed. Promptly Health received £11M with no visible business case.

**Curated minutes.** Beyond what is withheld, what is published is actively curated. The knowledge graph documents 107 sanitisation instances and 237 hiding-intent passages across DHCW's board and committee minutes. In one published transcript, the curation ratio fell to 10.7% — 3,680 published words from 34,257 spoken, with sixteen of nineteen speakers removed entirely. Audit Wales was present in the room for 43 of those sanitisation events, in real time, undetected. The full catalogue of named deletions is documented at [L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/).

**Oversight concealment.** The independent digital expert appointed under the Level 3 framework has never been publicly identified. Board papers are published as unsearchable PDFs. The board itself admits receiving information "late in the day."

In February 2026, Cabinet Secretary Miles criticised "a pattern of late notification that undermines system confidence." One month later, he went further: the governance framework was "complex, data-heavy, burdensome, lacks transparency and does not drive improvement." The minister responsible for overseeing DHCW was publicly acknowledging that the oversight mechanisms do not work.

The fortress interacts with every other Cluster B loop. [L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/) depends on the fortress to prevent contradictory information from reaching staff. [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) generates the problems that the fortress must hide. [L11: The Oversight Obstruction Loop](/diagnosis/l11-oversight-obstruction/) extends the fortress to the oversight bodies that could force it open.

Information flows cannot be fixed from inside when the people who control them are the people who benefit from blocking them. Breaking the fortress requires external force: Senedd mandate, Audit Wales statutory powers, Employment Tribunal disclosure, or Welsh Government directive.

## What would a healthy alternative look like?

Accountability data is published proactively on a statutory schedule — whistleblowing reports, disciplinary statistics, staff turnover analysis, contract values, programme delivery status. FOI requests are assessed on merit. Board papers are published in searchable, accessible formats. No public information source is blocked on internal networks. Transparency is the default, not a concession.

## How does the blueprint break the Information Fortress?

The fortress only stands because its custodians control what leaves the building. [Radical Transparency](/interventions/radical-transparency/) strips that discretion: statutory publication cadence for whistleblowing, disciplinary, leaver, contract, and delivery data; searchable board papers; no more unexplained blocks on public information sources. When publication is mandated rather than permitted, the fortress ceases to be a viable strategy.


## Captured Governance
URL: https://bluenhs.org/diagnosis/l11-oversight-obstruction/

_Eighteen months of zero PDC corrective actions. Audit Wales declared 'good governance' four months before Level 3. Welsh Government's GDS director driven from his role. Twelve months at Level 3 produced no de-escalation — only Level 4._

Suppressing internal feedback is one thing. Blocking external information is another. L11 is something else entirely: actively degrading the oversight function itself.

## What is captured governance at DHCW?

Oversight obstruction differs from information suppression in kind, not degree. Suppression hides the searchlight's view. Obstruction disables the searchlight. Once an oversight role is hollowed out, the capacity does not regenerate: the institutional knowledge walks out with the person, and the replacement — if there is one — inherits a capture-ready role rather than a functioning one.

The delay is permanent. What is removed does not come back on its own.

## How It Manifests at DHCW

**The captured assurance machinery.** Before the Welsh Government level, the oversight functions internal to DHCW had already failed. The board approved without scrutiny: the knowledge graph documents 51 distinct instances — including a £20M Kainos framework where the Chair later admitted "I should have looked. I don't know how these appear on our website as contracts," a £226M Microsoft Enterprise Agreement passed in a single sentence with no questions and no vote, and full business cases for LIMS, LINC, and RISP all moved into private session under "commercial sensitivity." The Performance and Delivery Committee generated zero corrective actions across eighteen consecutive months from May 2024 to May 2025, even after Welsh Government installed Olivia Shoraks as a permanent observer. All three sub-committees were blind to imminent Level 3 escalation in March 2025 — none raised concern signals in their final pre-escalation meetings. Audit Wales, present from month six, issued a "good governance, stable and cohesive board" assessment four months before Level 3 was triggered.

Forty-six months after DHCW's founding, Ifan Evans told the Performance and Delivery Committee: "This persistent ambiguity, let's call it, which we usually phrase as 'we need to clarify roles and responsibilities'... PDC is an oversight and assurance function, not a programme management function." DHCW was simultaneously being measured against milestones for programmes it did not control — most visibly EPMA, where local organisations are the accountable delivery body — while having no mechanism to hold those organisations to account. The gap was named, repeated, and never closed.

**Welsh Government as co-conspirator, then sudden judge.** Welsh Government created the conditions for failure, then escalated DHCW for the predictable consequences. WG pressured DHCW to soften red RAG ratings on programmes (Evans, PDC November 2024: "Government was asking... why then are those programs rated red amber?"). It cut DPIF from £33M planned to £28M allocated for 2024-25. It refused capital funding for e-referrals and the integration hub in November 2025, then set milestones requiring them. It compressed LIMS from a four-year programme into two years. It imposed a recruitment freeze via the remit letter while demanding accelerated delivery. Funding letters were not confirmed until 25% through the financial year (July 2022). The entity issuing the corrective signal was substantively co-authoring the failure.

**The neutralised oversight layer.** Leadership obscured the information provided to Welsh Government's oversight bodies. This contributed to the departure of the WG GDS director from his role. The GDS director was the Welsh Government's principal technical counterweight to DHCW — the one person in the oversight structure with the capacity to independently assess whether DHCW's claims about programme status were true. His departure removed that capacity. It has not been replaced. With the technical oversight function neutralised, DHCW's information monopoly over its own oversight body became near-total. The independent digital expert appointed under the Level 3 framework has never been publicly identified — making it impossible to assess their independence, qualifications, or conclusions.

**From Level 3 to Level 4.** Twelve months of Level 3 monitoring produced no de-escalation and no consequences. The monitoring mechanism itself had been captured: it functioned as proof that oversight existed without functioning as actual oversight. In 2026, DHCW was escalated further — to Level 4 Targeted Intervention. The Cabinet Secretary then publicly declared the entire framework "complex, data-heavy, burdensome, lacks transparency and does not drive improvement." The minister responsible for overseeing DHCW had publicly acknowledged that the oversight mechanism does not work — while continuing to operate it.

**March 2026 end state.** At the most recent board meeting on record: Phase One of Level 4 found two of forty-seven milestones missed (specifically LIMS and WRISTS); £32.9M of DPIF revenue and £13.1M of capital remained unallocated; an Accountable Officer letter had been sent stating delivery was "not possible without confirmed DPIF allocation"; recruitment was frozen by remit letter; the IQPD meetings were ceasing, replaced by a new escalation meeting chaired by the Director General from April; Rowan Gardner — the sharpest governance challenger on the board — was departing; the CFO had already departed to Swansea Bay. Glazzard's closing assessment: "We've always struggled because it's one-year funding. We should have worked it out by now, surely."

L11 completes the self-preservation engine. [L6](/diagnosis/l6-manufactured-narrative/) manages internal perception. [L9](/diagnosis/l9-whistleblower-suppression/) silences dissent. [L10](/diagnosis/l10-information-fortress/) blocks external information requests. But none of those address the most dangerous threat: active government oversight with real powers. L11 closes that gap.

This is the difference between hiding from a searchlight and disabling it.

The obstruction occurred while DHCW was under active investigation — meaning leadership was simultaneously being scrutinised and actively undermining the quality of that scrutiny. Every escalation step was taken with full knowledge of the consequences: the CEO and directors were informed of the harm and chose to proceed.

The legal significance is substantial. Misconduct in public office under the Attorney General's Reference (No 3 of 2003) requires "wilful misconduct serious enough to amount to an abuse of the public's trust."

Actively undermining the government's ability to monitor a public body — during an investigation triggered by serious delivery and patient safety concerns — meets that threshold.

The [Escalation trap](/traps/escalation/) is visible here. External scrutiny escalates. Leadership escalates information control in response. This generates more evidence of obstruction. Scrutiny escalates further. They are escalating their way into formal accountability proceedings — criminal, regulatory, and civil — that will bypass the information fortress entirely, because Employment Tribunal judgments are public, criminal investigations have compulsory powers, and Audit Wales has statutory access.

## What would a healthy alternative look like?

Oversight bodies have independent, unfiltered access to delivery data — not mediated through the people being overseen. Technical oversight roles are structurally protected: departure of the person fulfilling that role triggers automatic replacement, not a gap. Monitoring frameworks have teeth — escalation produces consequences on a defined timeline, not open-ended observation. When governance is not captured, the oversight function cannot be degraded from inside.

## How does the blueprint break captured governance?

Obstruction of oversight cannot be remedied from within the obstructing organisation. Breaking L11 requires two moves together: [Competent Leadership](/interventions/competent-leadership/) replaces the leadership actively degrading the oversight function, and [the governance redesign](/blueprint/governance/) distributes oversight across multiple actors with overlapping jurisdiction — so that capture of any single channel does not neutralise accountability entirely.


# Seven System Traps — All Active at DHCW
URL: https://bluenhs.org/traps/

_Donella Meadows identified seven system archetypes that lock organisations into dysfunction. DHCW has all seven active simultaneously. The literature contains no documented precedent._

Feedback loops are the individual mechanisms — each one a specific vicious cycle that depletes a specific resource. System traps are what emerges when multiple loops interact. Think of loops as the gears and traps as the machine they build together.

Donella Meadows identified seven system archetypes: recurring patterns of feedback loop interaction that produce predictable dysfunction (Meadows 2008, *Thinking in Systems*). Each has a characteristic signature and a known escape route. Recognising which trap you are in is more valuable than analysing individual events, because the archetype eliminates entire classes of solutions that look plausible but cannot work.

Most failing organisations exhibit two or three traps. DHCW has all seven active simultaneously — a combination the systems dynamics literature has not documented in any single organisation. Seven traps do not activate in parallel by accident. They require a coordinating mechanism. That mechanism is [Cluster B](/diagnosis/): the self-preservation engine that locks each trap in place and blocks the corrective feedback that would release it.

Three structural facts make DHCW's seven-trap pattern intelligible. First: the executive pipeline from which DHCW was assembled — the ABUHB succession of NHS Wales CEOs, the December 2020 UWTSD Professor of Practice titles awarded to three DHCW directors four months before its founding board — predates DHCW itself. Second: every governance deficit pattern observed at Level 3 escalation 34 months in was already operational at the very first board meeting. Third: twelve months of Level 3 enhanced monitoring produced no de-escalation; in 2026, DHCW was escalated to Level 4 Targeted Intervention. The traps did not develop over time. They were imported.

The supporting evidence base — 1,779 graph nodes, 3,427 edges, 61 board and committee meetings across five years (April 2021 – March 2026) — is summarised on the [diagnosis index](/diagnosis/).

## How traps relate to loops

A loop is a single causal circuit — one feedback mechanism, one stock being drawn down, one delay between cause and effect. A trap is what several loops build when they share components. Two loops that drain the same stock will interact. Seven loops wired together through common hires, common vendors, and common information flows produce a machine that behaves in characteristic ways — and that machine is what Meadows called an archetype.

The gears-and-machine metaphor is exact, not rhetorical. [L8: Loyalty Selection](/diagnosis/l8-loyalty-selection/) and [L7: The Competence Void](/diagnosis/l7-competence-void/) each do their own damage; together, they build **Success to the Successful** — loyalists accumulate influence while technical staff accumulate exits. [L9: Whistleblower Suppression](/diagnosis/l9-whistleblower-suppression/), [L10: The Information Fortress](/diagnosis/l10-information-fortress/), and [L11: Oversight Obstruction](/diagnosis/l11-oversight-obstruction/) together build **Escalation** — the arms race between information control and external scrutiny. The trap is not a theory overlaid on the loops. It is what the loops, running simultaneously, become.

## The trap kinship map

Traps that share loops are kin. The table lists the direct relations — pairs where two traps share at least one loop. Kinship predicts co-activation: wherever you see one of these traps, expect to see its relatives.

| Trap | Related traps |
|------|---------------|
| [Shifting the Burden](/traps/shifting-the-burden/) | Drift to Low Performance, Success to the Successful |
| [Drift to Low Performance](/traps/drift-to-low-performance/) | Shifting the Burden, Seeking the Wrong Goal |
| [Seeking the Wrong Goal](/traps/seeking-the-wrong-goal/) | Escalation, Drift to Low Performance |
| [Policy Resistance](/traps/policy-resistance/) | Shifting the Burden, Escalation |
| [Tragedy of the Commons](/traps/tragedy-of-the-commons/) | Shifting the Burden, Success to the Successful |
| [Success to the Successful](/traps/success-to-the-successful/) | Shifting the Burden, Seeking the Wrong Goal |
| [Escalation](/traps/escalation/) | Seeking the Wrong Goal, Policy Resistance |

Shifting the Burden is kin to four of the other six. That is a structural fact about its position in the system, not a coincidence.

## Severity ordering

Each trap page carries a `severity` field, one of five levels: **dominant**, **fundamental**, **very-active**, **structural**, **active**. Severity measures how much of the surrounding dysfunction originates with the trap — or is actively protected by it. **Shifting the Burden** is tagged **dominant** because almost every other trap depends on it for cover: the quick-fix reflex (hire instead of fix, outsource instead of build, dismiss the person who identified the problem) is the mechanism by which the other six avoid correction. Remove it, and the other traps become visible. **Escalation** is tagged **active** — severe in consequence and ongoing, but narrower in scope: the arms race concerns a specific subsystem (information control versus external scrutiny) rather than the whole organisation's posture. Severity is a reading guide: it tells the reader which trap to understand first.

## The seven traps, in severity order

1. **[Shifting the Burden](/traps/shifting-the-burden/)** (dominant) — quick fixes erode long-term problem-solving capacity; the ultimate burden-shift is dismissing the person who identified the problem.
2. **[Seeking the Wrong Goal](/traps/seeking-the-wrong-goal/)** (fundamental) — the organisation optimises for empire size and narrative control, not clinician adoption or patient benefit.
3. **[Drift to Low Performance](/traps/drift-to-low-performance/)** (very-active) — standards have drifted past inefficient, through wasteful, to dangerous.
4. **[Policy Resistance](/traps/policy-resistance/)** (structural) — six actors pulling in different directions, including DHCW's own technical staff in opposition to their own leadership.
5. **[Tragedy of the Commons](/traps/tragedy-of-the-commons/)** (active) — nine programmes compete for shared delivery capacity; none gets enough to succeed.
6. **[Success to the Successful](/traps/success-to-the-successful/)** (active) — loyalists accumulate power, technical staff accumulate exits; every promotion widens the gap.
7. **[Escalation](/traps/escalation/)** (active) — external scrutiny escalates, leadership escalates information control, each move generates fresh evidence.

Each trap page below explains the generic pattern, how it manifests at DHCW, which [feedback loops](/diagnosis/) produce it, and the escape route — which feeds directly into the [blueprint](/blueprint/).


## Shifting the Burden
URL: https://bluenhs.org/traps/shifting-the-burden/

_Hire instead of fixing process. Outsource instead of building capability. Rebrand instead of learning. And the ultimate burden-shift: dismiss the person who identified the problem._

## What is the 'shifting the burden' trap at DHCW?

In Meadows' framework, Shifting the Burden occurs when a system applies quick fixes that address the symptom but erode the system's own capacity to solve the underlying problem. Each quick fix makes the fundamental solution harder to implement, creating dependency on more quick fixes.

## How It Manifests at DHCW

The quick fixes are familiar: hire more people instead of fixing the processes that make them unproductive. Outsource to vendors instead of building internal capability. Rebrand failing programmes instead of conducting genuine post-mortems.

But DHCW has added a more dangerous variant: dismiss the person who identified the problem instead of fixing the problem. The whistleblower's role was replaced with a downgraded position — lower band, less authority. The oversight function was structurally eliminated. The problem didn't go away. The person who could see it did.

This is burden-shifting at its most corrosive. The "burden" being shifted is not just a technical challenge — it is the discomfort of being held accountable. And the "quick fix" is not hiring or rebranding — it is removing the source of accountability itself.

The reflex is reinforced culturally. Board Secretary Mary Darling described DHCW as "very keen to please" — offered as a positive trait that simply needs moderation. In burden-shifting terms, "keen to please" is a precise diagnostic: a culture that defaults to whichever response makes the immediate problem stop, regardless of whether it solves anything underneath. Quick fixes do not have to be argued for in such a culture. They are the path of least resistance.

The pattern extends beyond hiring, outsourcing, and rebranding to the structure of finance and strategy. DPIF — the discretionary investment fund meant for transformational programmes — was, in Evans's words, "expanded to fit things that are arguably not that transformational. Not that transformational like Link and RISP." A funding stream designed to compound long-term capacity was redirected to plug short-term operational gaps; the burden was shifted from operational budgets onto the transformation envelope. Vacancy savings did the inverse: a workforce shortfall was not addressed; instead, the deficit was paid down by leaving posts empty, transferring the burden to the staff who remained until 65% of them reported burnout.

The most ambitious burden-shift is strategic. Ifan Evans authored *A Healthier Wales* — the strategy DHCW was set up to implement, then failed to deliver. He has now been appointed to design the successor: the next ten-year "Blueprint". The failure of strategy one is being shifted forward into strategy two by the same author. Authoring the successor is the largest possible quick fix — the past failure is, by definition, not the strategy now in flight.

## Produced By

[L1: The Hiring Trap](/diagnosis/l1-hiring-trap/) — hire instead of fix. [L4: The Rebranding Escape](/diagnosis/l4-rebranding-escape/) — rebrand instead of learn. [L5: The Vendor Dependency Spiral](/diagnosis/l5-vendor-dependency-spiral/) — outsource instead of build. [L7: The Competence Void](/diagnosis/l7-competence-void/) — leaders who can't diagnose the real problem reach for surface fixes. [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) — remove the problem-identifier instead of the problem.

## How is the 'shifting the burden' trap broken?

Cannot come from within. The burden-shifting IS the leadership strategy — it is not a mistake they can be coached out of. External intervention is required: Welsh Government, Senedd, or legal proceedings that impose accountability from outside the system.

The blueprint's [Intervention 4: Flip the Model](/interventions/flip-the-model/) targets this trap directly -- replacing the outsource-and-hire reflex with a structure that builds genuine internal capability.


## Drift to Low Performance
URL: https://bluenhs.org/traps/drift-to-low-performance/

_Standards have drifted past inefficient, through wasteful, to dangerous. WPAS linked to at least one patient death. Royal Colleges warn of delays that lead to worsening health._

## What is the drift-to-low-performance trap at DHCW?

Drift to Low Performance occurs when an organisation gradually accepts worse outcomes as normal. The performance target drifts downward to meet actual performance, rather than actual performance being raised to meet the target. Each year's failure becomes next year's baseline.

## How It Manifests at DHCW

The trajectory is visible in the data:

- WCCIS: £42M+ spent, deployed to only 19 of 29 target organisations, called "live," then abandoned and rebranded to "Connecting Care"
- OpenEyes: £8.5M, 4+ years late, live in only 2 health boards
- NHS Wales App: ~345,000 registered users, but the Deputy CEO of NHS Wales admitted "hardly anybody is using it regularly"
- The CEO admitted: "We don't have an ROI on all of our investments"
- The Finance Director of the Year (2022 award) could cite only £0.5M in non-cash time savings from ~£600M of Welsh Government funding by January 2026 — 83p of delivered value for every £1,000 invested

The drift now has a floor: patient harm. WPAS has been linked to at least one patient death. eMPI mixed up patient records. WCCG ran on unsupported technology for 8+ years despite repeated technical staff warnings. The Royal Colleges stated that patients "regularly experience delays that lead to worsening health."

The drift is also visible in what is being measured. By July 2025, only 42% of IMRP milestones were on track. The CANISC cancer information system carried the highest risk score on the corporate risk register (score 20) from DHCW's founding through to replacement — every year, the same score, no movement. Rowan Gardner, the board's sharpest governance challenger before her departure, identified "a repeating pattern of end-of-life systems without replacement funding": a structural feature of the portfolio, not an anomaly. In March 2026, the consequence was visible to every NHS Wales clinician simultaneously: the PSBA network failed across all NHS Wales organisations, taking O365, EPMA, RISP, and radiology offline at once. The drift floor is no longer one programme producing one harmed patient. It is a single shared infrastructure failure producing simultaneous loss of digital service across the system.

The performance standard has drifted past "inefficient" through "wasteful" to "dangerous."

## Produced By

[L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/) — internal perception of success prevents recognition of declining standards. [L4: The Rebranding Escape](/diagnosis/l4-rebranding-escape/) — rebranding failed programmes avoids the honest assessment that would anchor standards. [L7: The Competence Void](/diagnosis/l7-competence-void/) — leaders without technology delivery experience cannot recognise when standards are slipping.

## How is the drift-to-low-performance trap broken?

Anchor to patient safety outcomes, not programme milestones. The Royal Colleges' July 2025 joint briefing provides the external benchmark — an authoritative, independent assessment that cannot be managed or rebranded. But anchoring requires leaders who accept the benchmark. Current leadership's response to every external benchmark has been to manage the narrative, not match the standard.

The blueprint's [Intervention 2: Radical Transparency](/interventions/radical-transparency/) targets this trap directly -- public, real-time dashboards anchored to patient outcomes make it impossible for performance standards to drift unnoticed.


## Seeking the Wrong Goal
URL: https://bluenhs.org/traps/seeking-the-wrong-goal/

_Not accidental. Not emergent. Witnessed. A close circle around the CEO deliberately directing resources, hiring, and decisions to serve their collective interests._

## What is the 'seeking the wrong goal' trap at DHCW?

Seeking the Wrong Goal is Meadows' term for a system that is performing exactly as designed — but the design serves the wrong purpose. The system achieves what it optimises for. The problem is what it optimises for.

## How It Manifests at DHCW

The analysis has traced this trap through three levels of understanding:

**Structural view:** DHCW appears to optimise for activity metrics — headcount growth, programme count, strategy publications — rather than outcome metrics like clinician adoption or patient benefit. The clearest single artefact: 14-15 sub-strategies pursued simultaneously by an 800-person organisation. An organisation that size cannot operationalise that many strategies. It can only publish them.

**Leadership view:** The optimisation is not accidental. Leadership choices consistently favour empire size, career positioning, and narrative control over delivery. The CEO accumulated a BCS Fellowship, FedIP registration, the UWTSD Professor of Practice title (awarded alongside two other DHCW directors in December 2020 — four months before DHCW's founding board), and a "Digital CEO of the Year" award, all during the 2020-2021 transition to permanent CEO. The Compassionate Leadership Pledge was approved by the board in fewer than fifteen seconds with no questions, while staff burnout sat at 65%. When asked what patient benefit corresponded to a 25% workforce expansion: "It would be lovely to demonstrate the value."

**Witness view:** A close circle around the CEO deliberately directs resources, hiring, and decision-making to serve their collective interests. Sham recruitment processes with predetermined outcomes. Protege advancement at the expense of programme delivery. Empire-building observed in specific decisions by specific people for specific beneficiaries. This is not an emergent property of a dysfunctional system. It is a conscious, coordinated choice.

**Pipeline view:** The clique predates DHCW. Three successive NHS Wales CEOs emerged from a single health board (ABUHB), via a patronage pipeline controlled by Andrew Goodall (Welsh Government Permanent Secretary, formerly ABUHB CEO) and Judith Paget (Director General Health & Social Services, formerly ABUHB CEO). The UWTSD professorships described above were awarded — through WIDI, the Wales Institute of Digital Information — before DHCW had even held its first board meeting. The wrong goal was not adopted by the organisation over time. It was set by the people who built it. The clique is not a corruption of the design. It is the design.

## Produced By

[L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/) — the internal perception of success supports the wrong goal. [L8: The Loyalty Selection Loop](/diagnosis/l8-loyalty-selection/) — hiring for loyalty rather than competence builds an organisation optimised for control. [L10: The Information Fortress](/diagnosis/l10-information-fortress/) — blocking external information prevents the wrong goal from being exposed. [L11: The Oversight Obstruction Loop](/diagnosis/l11-oversight-obstruction/) — degrading oversight ensures no external body can challenge the goal.

## How is the 'seeking the wrong goal' trap broken?

This trap cannot be escaped by the current leadership because the wrong goal IS their goal. The goal changes only when the people who set the goal change. The people who set the goal have been directly informed of the harm their actions cause. They chose to continue. The escape route is leadership replacement — preceded by the accountability proceedings that make replacement unavoidable.

The blueprint's [Intervention 1: Competent Leadership](/interventions/competent-leadership/) addresses this trap at its root -- replacing the people who set the wrong goal -- while [Intervention 4: Flip the Model](/interventions/flip-the-model/) restructures the organisation so the new goal is patient outcomes, not empire size.


## Policy Resistance
URL: https://bluenhs.org/traps/policy-resistance/

_Six actors pulling in different directions. Welsh Government, DHCW leadership, health boards, clinicians, vendors — and DHCW's own technical staff, in direct opposition to their own leadership._

## What is the policy-resistance trap at DHCW?

Policy Resistance occurs when multiple actors within a system have conflicting goals, and each actor's actions to achieve their goal counteract the others. The result is stalemate: enormous effort is expended, but the system barely moves.

## How It Manifests at DHCW

Six actors are pulling in different directions simultaneously:

| Actor | What they want | What they do |
|-------|---------------|-------------|
| Welsh Government | National standardisation, value for money | Co-author failure conditions (DPIF cuts, capital refusal, RAG pressure), then escalate for the consequences |
| DHCW Leadership | Self-preservation, empire growth, narrative control | Grow headcount, manage scrutiny, suppress dissent |
| Health Boards | Local autonomy, systems that work | Resist imposed systems, run shadow IT |
| Clinicians | Tools that improve patient care | Ignore or workaround systems that slow them down |
| Vendors | Revenue maximisation | Expand scope, extend timelines, avoid accountability |
| DHCW Technical Staff | Deliver working systems, raise real risks | Raise concerns — and get dismissed |

The addition of DHCW's own technical staff as a *resisting* actor is the most telling feature. In a healthy organisation, technical staff and leadership are aligned. At DHCW they are in direct opposition: staff want to deliver and flag risks, leadership wants to control the narrative and suppress risk signals.

DHCW is fighting itself in addition to fighting external resistance.

The structural design also locks the resistance in place. Take EPMA: DHCW is measured against EPMA milestones by Welsh Government, but as Evans admitted on the record, "the national EPMA program is a coordinating and knowledge sharing function... Each local organisation has its own EPMA implementation program. They are the accountable delivery body." DHCW carries the consequences for missing milestones it does not control. Health boards control the implementation but face no equivalent accountability. The actor measured cannot deliver, the actors who can deliver are not measured, and Welsh Government — which set up the gap — escalates DHCW for the result. Resistance is not just behavioural friction between actors. It is engineered into the organisational chart.

## Produced By

[L2: The Credibility Death Spiral](/diagnosis/l2-credibility-death-spiral/) — health boards resist because they don't trust DHCW to deliver. [L7: The Competence Void](/diagnosis/l7-competence-void/) — leaders without tech experience make decisions that technical staff must then resist. [L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) — technical staff resistance is punished, driving it underground rather than resolving it.

## How is the policy-resistance trap broken?

Align goals by making delivery the only metric that matters for career survival. Currently, loyalty matters more than delivery for career progression at DHCW. Reverse this: tie leadership tenure to externally verified delivery milestones. When every actor's self-interest aligns with patient benefit, policy resistance dissolves.

The blueprint's [Intervention 1: Competent Leadership](/interventions/competent-leadership/) targets this trap directly -- removing the leadership whose goals conflict with every other actor's is the only way to align the system and end the stalemate.


## Tragedy of the Commons
URL: https://bluenhs.org/traps/tragedy-of-the-commons/

_Nine programmes competing for shared delivery capacity. No programme gets enough to succeed. Each draws from the same pool of overloaded staff._

## What is the tragedy-of-the-commons trap at DHCW?

The Tragedy of the Commons occurs when multiple users draw from a shared resource, each acting rationally in their own interest, but collectively depleting the resource until it fails everyone. No individual user has an incentive to restrain their consumption.

## How It Manifests at DHCW

Nine major programmes compete simultaneously for the same shared delivery capacity — the same pool of staff, the same architecture teams, the same integration specialists, the same testing environments. Each programme director fights for resources. No programme gets enough to succeed. All nine are under Level 3 review.

The current set is at minimum: CANISC, LIMS, RISP, WICIS, EPMA, the NHS Wales App, GP system migration, Cloud Transition, Eye Care (OpenEyes), and Digital Maternity. Two of these — LIMS and RISP — were running concurrently in 2025 with explicit acknowledgement at the Performance and Delivery Committee that the load was producing "significant pressure on resources across NHS Wales." LIMS itself had been compressed from a four-year programme into two; Michelle Sell, PDC February 2025: "we compressed what would have been a four-year program into two years. And I think we're seeing some of the consequences of that now." The compression was a deliberate decision; the consequence was a foreseeable failure.

The shared-resource problem was diagnosed early. At month seven of DHCW's existence, the architecture team explicitly asked the board for support and prioritisation — the prescription that would have addressed the commons before it deepened. The board agreed "enthusiastically" and did nothing. Doyle, the same year, asked the board to identify "what to stop doing." No stop list followed. The prescription was offered repeatedly and rejected each time. The commons continued to be drawn from until each programme failed.

The tragedy is compounded by [L1: The Hiring Trap](/diagnosis/l1-hiring-trap/). The response to insufficient capacity is to hire more people, but the new hires draw on the same experienced staff for onboarding, further depleting the shared resource.

The tragedy is further compounded by [L7: The Competence Void](/diagnosis/l7-competence-void/). Even if you concentrated all resources on a single programme, leaders without technology delivery experience would still make poor decisions about that programme. The commons is not just depleted — it is poorly managed.

## Produced By

[L1: The Hiring Trap](/diagnosis/l1-hiring-trap/) — each programme demands more people, but adding people to the shared pool makes it less productive, not more.

## How is the tragedy-of-the-commons trap broken?

Ruthless portfolio focus. Pause six programmes. Deliver three. The three must be selected by an independent panel including health board clinicians — not by DHCW leadership, who will prioritise programmes that justify their empire over programmes that deliver clinical value. One-page delivery contracts: scope, deadline, target, consequences.

The blueprint's [Intervention 3: Portfolio Ruthlessness](/interventions/portfolio-ruthlessness/) targets this trap directly -- cutting nine competing programmes to three gives each survivor enough of the shared resource to actually deliver.


## Success to the Successful
URL: https://bluenhs.org/traps/success-to-the-successful/

_Two tiers. Loyalists accumulate power. Technical staff accumulate exits. Every promotion of a loyalist strengthens one tier and hollows the other._

## What is the 'success to the successful' trap at DHCW?

Success to the Successful occurs when initial advantages compound: winners get more resources, which produce more wins, which attract more resources. Meanwhile, the losing side is progressively starved, making recovery increasingly unlikely. The gap widens with each cycle until it becomes irreversible.

## How It Manifests at DHCW

The [loyalty selection loop (L8)](/diagnosis/l8-loyalty-selection/) has created a two-tier organisation:

**Tier 1 — Loyalists:** Promoted regardless of delivery. Protected from accountability. Credential-accumulation supported. Growing influence, headcount, and budget. Access to the CEO's patronage network.

**Tier 2 — Technical and delivery staff:** Expected to deliver despite poor leadership decisions. Concerns ignored or punished. Skills undervalued. Leaving or being pushed out. Those who remain learn not to challenge.

Every promotion of a loyalist strengthens Tier 1's control and weakens Tier 2's capacity and morale. The dismissed senior technical leader's role was replaced with a downgraded position — lower band, less authority. The oversight function was structurally eliminated. The loyalist network became more secure. Delivery capability became weaker.

The asymmetric advantage predates DHCW itself. Three successive NHS Wales CEOs were drawn from a single health board (ABUHB), via a patronage pipeline controlled by Andrew Goodall and Judith Paget. In December 2020 — four months before DHCW's founding board meeting — Helen Thomas, Rhidian Hurle, and Ifan Evans received UWTSD Professor of Practice titles via the Wales Institute of Digital Information. Tier 1 was credentialed before Tier 2 had a chance to compete. The "success" that compounds was awarded, not earned in delivery.

The compounding continues into the next strategy. Ifan Evans authored *A Healthier Wales* — the strategy DHCW was created to implement, then failed to deliver. He is now designing the next ten-year "Blueprint." The failed strategist of cycle one is the appointed strategist of cycle two. The advantage does not just persist; it produces the very mechanism that compounds it.

Eventually, Tier 2 either leaves (brain drain) or capitulates (learned helplessness), and Tier 1 has complete control of a hollow organisation.

## Produced By

[L7: The Competence Void](/diagnosis/l7-competence-void/) — technical competence flows out while incompetence flows in. [L8: The Loyalty Selection Loop](/diagnosis/l8-loyalty-selection/) — the mechanism that directs resources to the loyalist tier and away from the delivery tier.

## How is the 'success to the successful' trap broken?

Level the playing field. Impose external, independent technical assessment of all senior appointments — skills audits conducted by people outside the loyalty network, against verifiable competency frameworks used by comparable organisations.

The blueprint's [Intervention 3: Portfolio Ruthlessness](/interventions/portfolio-ruthlessness/) targets this trap directly -- when resource allocation is governed by independent clinical priority rather than internal patronage, the loyalist tier loses its mechanism for accumulating advantage.


## Escalation
URL: https://bluenhs.org/traps/escalation/

_External scrutiny escalates. Leadership escalates information control. Each move generates new evidence. They are escalating their way into formal accountability proceedings._

## What is the escalation trap at DHCW?

Escalation occurs when two parties are locked in an arms race, each escalating in response to the other's moves. Neither can back down without losing. The dynamic is self-reinforcing: each escalation provokes a counter-escalation, consuming increasing resources while the underlying problem goes unaddressed.

## How It Manifests at DHCW

An escalation dynamic has emerged between external scrutiny and leadership's information-control apparatus:

External scrutiny escalates — Senedd investigation, Level 3 enhanced monitoring, the carenhs.org campaign site. Leadership escalates information control in response — block the website, dismiss the whistleblower during the investigation, drive out the WG GDS director. External scrutiny escalates further — Employment Tribunal filing, media coverage, additional Senedd sessions. Leadership escalates further — accelerate the whistleblower removal during active investigation, tighten the information fortress.

Each side responds to the other's latest move. The cycle accelerates.

## Why This Trap Is Dangerous Now

In Meadows' framework, escalation traps resolve in one of two ways: one party exhausts its capacity to escalate and capitulates, or an external intervention imposes a ceasefire.

Leadership's escalation capacity is bounded by the legal system. They cannot dismiss every whistleblower, suppress every data request, and obstruct every oversight process without eventually triggering proceedings that bypass their information control entirely. Employment Tribunal judgments are public. Criminal investigations have compulsory powers. Audit Wales has statutory access.

But every escalation by leadership during the arms race — particularly dismissing the whistleblower during an active investigation, and obstructing Welsh Government oversight — generates additional evidence that strengthens the case against them. Each escalation step was taken with full knowledge of the consequences: the CEO and directors were informed of the harm and chose to proceed.

DHCW leadership are escalating their way into formal accountability proceedings.

By March 2026, the arms race had moved past the original framework. DHCW had spent twelve months at Level 3 enhanced monitoring with no de-escalation. Welsh Government escalated further — to Level 4 Targeted Intervention. Phase One of Level 4 found two of forty-seven milestones missed (LIMS and WRISTS). The IQPD oversight meetings — the formal vehicle for monitoring conduct under Level 3 — were being wound down, replaced by a new escalation meeting chaired by the Director General from April. In February, Cabinet Secretary Miles publicly described "a pattern of late notification that undermines system confidence." A month later, he described the entire framework as "complex, data-heavy, burdensome, lacks transparency and does not drive improvement." The minister responsible for overseeing DHCW was publicly disowning the oversight framework while continuing to operate it. Each round of external escalation has now produced both a counter-escalation by leadership and a structural admission that the existing tools are insufficient. The arms race is now operating above the framework that was designed to contain it.

## Produced By

[L9: The Whistleblower Suppression Loop](/diagnosis/l9-whistleblower-suppression/) — the dismissal is itself an escalation that triggers counter-escalation. [L10: The Information Fortress](/diagnosis/l10-information-fortress/) — each barrier raised provokes attempts to breach it. [L11: The Oversight Obstruction Loop](/diagnosis/l11-oversight-obstruction/) — degrading oversight is the most dangerous escalation, because it directly challenges the institutions with enforcement powers.

## How is the escalation trap broken?

The external parties — Senedd, Welsh Government, Audit Wales, Employment Tribunal — must win the escalation by deploying powers that leadership cannot counter: compulsory disclosure, testimony under oath, statutory audit powers. The alternative — leadership wins the escalation by exhausting external scrutiny's capacity — leads to complete institutional capture and unchecked accumulation of patient safety risk.

The blueprint's [Intervention 1: Competent Leadership](/interventions/competent-leadership/) ends this arms race by replacing the leadership that escalates information control, while [Intervention 2: Radical Transparency](/interventions/radical-transparency/) removes the information asymmetry that makes escalation possible.


# Six Targeted Interventions
URL: https://bluenhs.org/interventions/

_Six structural interventions — five sequenced at DHCW, one parallel at Welsh Government level. Each targets a different leverage point. Together they dismantle the architecture of failure and build the conditions for delivery._

The [blueprint](/blueprint/) defines the destination — what good looks like, what comparable countries built, why the "once for Wales" monopoly is the problem. These interventions are the operational plan to get there: five sequenced at DHCW with leadership reform as the prerequisite, plus one parallel intervention at Welsh Government level.

The five DHCW interventions are sequenced by dependency. Intervention 1 must happen first because every subsequent intervention depends on leadership that is competent, accountable, and committed to delivery over self-preservation. Without it, the transparency dashboard (Intervention 2) will be designed by the same leadership that blocked FOI requests. The portfolio review (Intervention 3) will be conducted by the same executives who allowed nine simultaneous failing programmes. Skip the prerequisite and the remaining four will be captured by the governance dynamics that currently protect failure from correction.

A sixth intervention runs in parallel. Welsh Government materially contributed to the conditions that produced DHCW's failure — softening RAG ratings, cutting DPIF mid-cycle, refusing capital while retaining milestones, compressing programmes, freezing recruitment via remit letter. Replacing DHCW's leadership without reforming the funder reproduces those conditions for the successor body. [Intervention 6: Reform the Funder](/interventions/reform-the-funder/) is owned by Welsh Government and the Senedd Public Accounts Committee; it begins in Months 0-3, alongside the forensic review under Intervention 1.

## Where the Leverage Is

Most reform effort targets the shallow end of a system — parameters, budgets, headcount — where the organisation absorbs the change without altering its behaviour. This is why adding 600 staff to DHCW changed nothing.

Donella Meadows ranked twelve places to intervene in a system, from shallow (adjusting numbers) to deep (changing the system's goals and operating paradigm). DHCW has operated exclusively at the bottom three levels. These interventions target the deep end — information flows, system rules, self-organisation, system goals — where structural change becomes possible. (The full [twelve-level framework](/methodology/) explains the hierarchy.)

Every deep intervention requires one prerequisite: the governance dynamics that currently protect failure — [Cluster B](/diagnosis/) — must not be in a position to sabotage them.

## The Sequence

## The 36-Month Commitment

Reform at this depth cannot be compressed into a political cycle. Trust rebuilds on a 2-5 year lag behind demonstrated delivery. Embedded teams need 6-12 months to ship. The sequencing matters as much as the components. In the first six months: an independent forensic review delivers findings, mandatory data publication begins, and leadership decisions are tied to delivery milestones.

[Read the full 36-month timeline →](/blueprint/timeline/)

## The Economics

DHCW's annual budget reached approximately £200M in 2025-26 with quantified delivered value across five years of £0.5M. Direct annual DHCW waste runs at **£100-150M per year**; cumulative five-year status-quo direct waste is **£500M-£1B**. Planned reform requires a one-off £5-15M investment that breaks even **in weeks**. Downstream impact across NHS Wales is **5-15× the direct figure** — total five-year cost of status quo to NHS Wales: **£3-10 billion**. Patient harm continues to accrue on a separate ledger that pounds cannot quantify.

[Read the full cost analysis →](/blueprint/cost-of-inaction/)

## Who Guards the Guardians?

Oversight of DHCW has been weak, intermittent, and partly captured. The interventions require competent oversight of the transition. If the oversight body is part of the problem, who oversees the reform?

[Read the governance design →](/blueprint/governance/)

## Monitoring

Monthly leading indicators across four quadrants — delivery, trust, transparency, safety. A decision tree that distinguishes premature abandonment from genuine failure. The single critical signal: one health board saying "DHCW delivered what they promised."

[Read the monitoring framework →](/blueprint/monitoring/)


## Intervention 1: Competent Leadership
URL: https://bluenhs.org/interventions/competent-leadership/

_Before anything else, the leadership producing the failure must be replaced by leadership capable of delivery. Eight parallel actions across Months 0–6._

This is the intervention that makes every other intervention possible.

1. **Commission an independent forensic review** of DHCW's recruitment practices, vendor procurement decisions, and programme delivery — conducted by reviewers from outside the NHS Wales ecosystem, with statutory Audit Wales powers.
2. **Mandate publication** of all whistleblowing data, disciplinary proceedings, staff leavers analysis, and contract values — remove DHCW's discretion to suppress this information.
3. **Require independent skills audits** of all executive and director-level roles — measured against verifiable competency frameworks used by comparable organisations.
4. **Establish a protected reporting channel** for DHCW staff to raise concerns directly to Audit Wales or Welsh Government, bypassing DHCW management entirely.
5. **Set hard, externally-verified delivery conditions** tied to leadership tenure: if specific programmes do not reach specific adoption milestones by specific dates (not self-reported), leadership is replaced. Not reshuffled. Replaced.
6. **Conduct patient safety triage** — independent clinical safety review of every live national system with documented incident history: WPAS (linked to at least one patient death), eMPI (patient record mixups), WCCG (running on unsupported technology for 8+ years against repeated technical staff warnings), WICIS ("effectively still on pause"), RISP (global worklist issue threatening cross-Wales image viewing), and PSBA-class shared infrastructure (a single failure took O365, EPMA, RISP, and radiology offline simultaneously across NHS Wales in March 2026). Pause systems where the safety case is unsupported. Publish findings.
7. **Reset the non-executive board** — replace the executive *and* audit the non-executive board against published competency criteria. Sub-committees that produced zero corrective actions across eighteen consecutive months — the Performance and Delivery Committee from May 2024 to May 2025 — must be reconstituted with technical NEDs. The departures of Rowan Gardner, Simon Jones, and Ruth Glazzard between 2025 and 2026 removed the sharpest governance voices on the board; the reset must replace them with people capable of asking the questions Gardner asked.
8. **Disclose the patronage pipeline** — full audit of declarations of interest across DHCW history, with specific examination of the 33 instances where directors declared "nil" while holding undisclosed Professor of Practice titles (Thomas 10, Evans 10, Hurle 8, Hall 5) and the 45 further instances of undeclared interests at specific meetings. The December 2020 UWTSD professorships awarded to three executive directors four months before DHCW's founding board are examined as a structural finding — the patronage architecture that pre-credentialed the executive cohort — not as a compliance lapse. The ABUHB CEO succession (three NHS Wales CEOs from one health board, via Goodall and Paget) is examined for its bearing on appointments at DHCW.

**Why first:** Every subsequent intervention depends on leadership that is competent, accountable, and committed to delivery over self-preservation. Implementing transparency dashboards under leadership that blocks websites and suppresses data is futile. Embedding clinical teams under leadership that overrides technical experts is futile. Reviewing patient safety under leadership that has presided over WPAS-linked patient harm without consequence is, on the evidence, performative. The eight actions above run in parallel from Month 0 — none can wait.


## Intervention 2: Radical Transparency
URL: https://bluenhs.org/interventions/radical-transparency/

_Live delivery dashboard, outcome-only KPIs, published vendor contracts and staff-reported confidence. Break the information fortress by statute._

Live delivery dashboard showing programme status in real time. Outcome-only KPIs replacing activity metrics. Published programme costs alongside adoption metrics — the public should see: "OpenEyes: £8.5M spent, live in 2 of 7 health boards after 6 years."

Published vendor contract values and delivery performance. No more unnamed suppliers on £47-56M contracts. Staff-reported delivery confidence published alongside management-reported programme status — if there's a gap, the staff number is more reliable.

That is the headline. The operational specification below makes it stick.

## What gets published, when

The diagnosis identifies seven categories DHCW has actively suppressed. Each requires a named publication cadence:

1. **Whistleblowing data** — monthly count by category and outcome status; 30-day maximum publication lag; year-on-year trend visible.
2. **Disciplinary proceedings** — monthly count by category and band, anonymised; outcomes tracked separately from initiations, so pretextual use becomes visible as a pattern.
3. **Staff leavers** — monthly leaver analysis by reason, role, and destination; 12-month rolling trend. The L9 finding that "the absence of data is itself data" closes by default.
4. **Contract values** — every contract above £100K published within 30 days of signature; sole-bidder contracts above £100K published within 14 days with a public justification. The RISP (£47-56M), Channel 3/Aire Logic, and Promptly Health (£11M, no published business case) undisclosed-value gap closes by default.
5. **Programme delivery status** — live dashboard, milestone-level, weekly update; current state, slippage against the previous baseline, named accountable individual.
6. **Confidence-of-delivery scores** — every milestone owner attaches a confidence score (1-5) at the start of each reporting period; both the score and the ground-truth outcome are published, so optimism bias becomes measurable. This is the direct counter to Evans's admission that "confidence of delivery is not currently measured" across 550 tracked milestones.
7. **Board and committee minutes** — full publication within 14 days of meeting; original transcript form, with personal information redacted only as required by data protection law; no editorial sanitisation between transcript and published record.

## Statutory backing

Each of the seven categories above is published under a statutory duty, not a discretionary policy. This is the highest-leverage element of the intervention: removing leadership discretion converts publication from a governance negotiation into a legal compliance matter. Non-publication becomes a matter for the courts and the Information Commissioner, not for internal escalation. This single design choice neutralises the most reliable strategy of [L10: The Information Fortress](/diagnosis/l10-information-fortress/) — it removes the option of withholding.

## Anti-sanitisation protocol

The diagnosis documents 107 sanitisation instances, 237 hiding-intent passages, and a curation ratio that fell to 10.7% in one published transcript — 3,680 published words from 34,257 spoken, with sixteen of nineteen speakers removed entirely. The transparency intervention must directly counter the curation mechanism, not merely publish more data on top of it. Three rules:

- **Minimum curation ratio of 60% across published meeting minutes.** Below that threshold, an explanation accompanies the published version naming who curated and why.
- **Named-individual challenges preserved verbatim.** Where a board or committee member raised a question or objection, it is published as spoken, not paraphrased. Chair Jones's "heap misery on misery" warning, Lloyd-Jones's "anodyne phrase" candour about technical debt, Evans's "biggest disappointment" admission about milestone owners lying — none of these would have been deletable under this protocol.
- **Audit trail of edits.** Every change between transcript and published minute is logged with editor, timestamp, and reason. Audit Wales receives the full audit trail quarterly.

## FOI and the cost of refusal

Every refused FOI request is published with the reason for refusal. ICO referrals are tracked publicly. Cabinet Office FOI compliance scores are published quarterly. Refusal itself becomes public data, which makes refusal a reputational cost rather than a costless default. This addresses the documented pattern of FOI requests being declined or delayed without consequence.

## Dashboard design integrity

The dashboard is itself a target for capture. Three rules prevent it:

- **Published technical schemas.** What is on the dashboard, where the data comes from, how it is computed — all open. External developers can replicate the dashboard from raw data and verify the result matches.
- **External audit verification.** A named external body — Audit Wales or its commissioned digital expertise — verifies the dashboard against the underlying data on a quarterly cycle.
- **No "context" columns.** If "context" or "narrative" columns are added to soften or qualify the headline numbers after launch, the dashboard has been captured. The trigger condition for governance escalation is automatic.

## Why this intervention without I1 fails

A transparency dashboard launched under leadership that blocked carenhs.org on the NHS Wales network and published zero whistleblowing statistics for five years will be designed to satisfy the letter of the rule and not the spirit. Implementation under the existing leadership is not an option — it is a guaranteed capture point. This is why [Competent Leadership](/interventions/competent-leadership/) is the prerequisite. The dashboard must be designed and operated by leadership accountable for what it shows.

This intervention dismantles [L6: The Manufactured Narrative](/diagnosis/l6-manufactured-narrative/) and [L10: The Information Fortress](/diagnosis/l10-information-fortress/) by removing the discretion the manufacture and the fortress depend on.


## Intervention 3: Portfolio Ruthlessness
URL: https://bluenhs.org/interventions/portfolio-ruthlessness/

_Pause six programmes. Focus on three selected by an independent panel. Publish one-page delivery contracts. Conduct genuine post-mortems on WCCIS and OpenEyes._

Pause six programmes. Focus on three selected by an independent panel including health board clinicians and the external digital expert — not by DHCW leadership, who will prioritise programmes that justify their empire over programmes that deliver clinical value.

For each focused programme, publish a one-page delivery contract: scope, deadline, adoption target, consequences for missing. Signed by the responsible executive. No "revised timelines." No "lessons learned exercises." Deliver or be replaced.

Conduct a genuine post-mortem on WCCIS (£42M+) and OpenEyes (£8.5M) with external facilitation. Not to blame, but to structurally understand why they failed. The [rebranding escape](/diagnosis/l4-rebranding-escape/) must be broken.

## Selection criteria for the priority three

Programme selection is published, weighted, and externally validated. Three published criteria:

1. **Patient-safety risk weight.** Programmes whose failure is causing or risking patient harm rank highest. WPAS (linked to a patient death), WCCG (8+ years on unsupported technology), and PSBA-class shared infrastructure (the March 2026 cross-NHS-Wales outage) outweigh programmes with administrative impact only.
2. **Clinical adoption potential.** Will clinicians actually use the system if it ships? An adoption forecast from named clinicians at named health boards, not from DHCW programme management.
3. **Reversibility of harm if delayed.** A programme whose delay causes accumulating clinical harm (diagnostic image flow, prescribing safety) ranks above a programme whose delay is operationally inconvenient but clinically tolerable.

## Panel composition

The selection panel is at least 50% clinical, drawn from health boards. At least one international peer from Denmark, Estonia, or NHS Digital England participates as a voting member — the comparator jurisdictions that have done what Wales is being asked to do. The Welsh Government observer is non-voting. Panel transcripts are published; selection rationale is published; dissents are recorded, not erased.

## Disposition of paused programmes

Paused does not mean cancelled. The six paused programmes retain protected status: staff are retained, vendor contracts are re-stated rather than torn up (with the new conditions of [Flip the Model](/interventions/flip-the-model/) applied to their content), and restart conditions are named in the panel's published decision. The diagnosis warns specifically against the rebranding escape; this protocol prevents the failure mode where pausing becomes covert cancellation followed by a relaunched "Connecting Care 2."

## Post-mortem methodology

Post-mortems are externally facilitated — by the National Audit Office digital function, the Cabinet Office digital function, or an equivalent independent body. Findings are structural, not individual: the question is "what conditions produced this outcome?", not "who do we blame?" The list extends beyond WCCIS and OpenEyes:

- **LIMS** — the four-year-to-two-year compression and the resulting collapse (Sell, PDC February 2025).
- **RISP** — the global worklist issue threatening cross-Wales image viewing; the legacy supplier forced to provide development for an end-of-life system.
- **WICIS** — described as "effectively still on pause" with a Welsh Government commissioned independent patient safety review.
- **WCCG** — eight years on unsupported technology against repeated technical-staff warnings.

Findings are published. The 14-day publication rule under [Radical Transparency](/interventions/radical-transparency/) applies to the post-mortem report itself.

## The stop list

Doyle requested in DHCW's founding year that the board understand "what to stop doing." It was never implemented. Under this intervention the stop list is a published deliverable from the panel, binding for 24 months: programmes are explicitly stopped, not just deferred. The 14-15 sub-strategies an 800-person organisation cannot operationalise are reduced to a deliverable count, against published criteria, with a stop list of equal weight to the do list.

This intervention dismantles [L1: The Hiring Trap](/diagnosis/l1-hiring-trap/) and addresses the [Tragedy of the Commons](/traps/tragedy-of-the-commons/) by reducing demand on the shared resource pool to a level that allows each surviving programme to succeed.


## Intervention 4: Flip the Model
URL: https://bluenhs.org/interventions/flip-the-model/

_Embed DHCW teams in health boards under clinical leadership. Recruit externally. Give embedded teams authority to bypass national mandates when clinicians need something faster._

Embed DHCW teams in health boards, working under clinical leadership. Teams report to health board clinical leadership, not to DHCW programme management. If the loyalty network controls the embedded teams, they become DHCW outposts, not clinical collaborators.

Recruit embedded team leads externally — from the commercial health tech sector, from NHS Digital England, from international equivalents. Not from within the DHCW loyalty network. Pay market rates. The Band 8c (£71-82k) salary for Head of Software Engineering signals that DHCW is not serious about external talent.

Give embedded teams authority to bypass DHCW's national architecture mandates if clinicians need something faster. Let local success prove what works before standardising.

## The architectural endpoint

DHCW becomes a standards-and-interoperability body, modelled on TEHIK (Estonia, ~200 staff for 1.3M people) and MedCom (Denmark, jointly owned by central government, regions, and local government). Tightly scoped:

- **National data standards.** Every health-board procured system must comply.
- **National patient index.** Single source of truth for "who is this patient?", maintained centrally and queried by health-board clinical systems.
- **National interoperability backbone.** A Wales equivalent of X-Road (Estonia, Finland, and seven further jurisdictions) — a federated data exchange layer, open source, operated by the standards body.
- **National cybersecurity for shared infrastructure.** Including the PSBA layer where a single failure cascaded across all NHS Wales in March 2026.

Health boards procure clinical applications within those standards. The standards body does not deliver clinical applications. The "once for Wales" delivery monopoly is dismantled — and replaced with a federated architecture that has worked in every comparable jurisdiction.

## Vendor and contract portfolio recovery

DHCW manages a contract portfolio valued at roughly £1.25 billion. The diagnosis documents 51 instances of board approval without scrutiny, multiple sole-bidder contracts, and several contract values entirely undisclosed (the RISP radiology supplier on a £47-56M contract, the Channel 3 / Aire Logic NTA contract, Promptly Health on £11M with no published business case). The PSBA outage in March 2026 demonstrated the operational consequence: a single shared dependency took O365, EPMA, RISP, and radiology offline across every NHS Wales organisation simultaneously. The portfolio is not just expensive. It is fragile.

Recovery has four steps:

1. **Audit the full portfolio.** Every contract above £100K, with value, term, sole-bidder status, performance-against-milestones data, and concentration risk. Published on the transparency dashboard under [Radical Transparency](/interventions/radical-transparency/).
2. **Re-tender or terminate sole-bidder contracts above £1M** unless a public justification for sole-bidder status is filed and accepted by the new governance body. The 51 approved-without-scrutiny instances are the precedent: the default is renewed scrutiny.
3. **Remediate single-point-of-failure dependencies.** PSBA-class shared infrastructure receives an explicit redundancy and exit plan. No single supplier holds infrastructure that, if it fails, takes the whole NHS Wales digital estate offline.
4. **Renegotiate against the new architectural endpoint.** Where the standards-body model means DHCW will no longer deliver an application, contracts transfer to (or are re-tendered by) the relevant health board, with knowledge transfer included.

## Embedded team governance

Embedded teams report to health board clinical leadership. Each team publishes a monthly delivery report — what shipped, what slipped, what changed in clinical use. Exit criteria are named: once an embedded team has shipped a stable system used by clinicians, the team disbands and the application is owned by the health board, with the standards body retaining only standards-compliance oversight.

## Recruitment partner specified

Recruitment for embedded team leads is conducted by a commercial recruiter from the health-tech sector — not by DHCW HR. Salary bands align with NHS Digital England's commercial-recruitment ranges and Scottish Government Digital. Selection criteria are published in advance. External panels conduct selection. The pattern documented at L8 — sham recruitment with predetermined outcomes — is structurally precluded by externalising the recruitment apparatus for these roles.

This intervention dismantles [L2: The Credibility Death Spiral](/diagnosis/l2-credibility-death-spiral/) and [L5: The Vendor Dependency Spiral](/diagnosis/l5-vendor-dependency-spiral/).


## Intervention 5: Break the Annual Trap
URL: https://bluenhs.org/interventions/break-annual-trap/

_Multi-year funding tied to the programme, not the organisation. Prevents leadership from using multi-year funding as a shield while giving programmes the continuity they need._

Multi-year funding tied to the *programme*, not the *organisation*. Don't give DHCW a 3-year budget — give the Electronic Prescribing programme a 3-year budget with external delivery milestones. If DHCW leadership is replaced, the programme continues with new leadership and the same funding.

This prevents leadership from using multi-year funding as a shield while giving programmes the continuity they need.

## Programme funding envelopes

Each priority programme — selected under [Portfolio Ruthlessness](/interventions/portfolio-ruthlessness/) — receives a three-year funding envelope from the start of its delivery phase. Release is milestone-based, against externally-validated milestones, not against self-reported "green" status. The 550-milestones-with-zero-confidence-scoring pattern documented at L2 is the precedent the new structure replaces: every milestone in the envelope carries a confidence score, an independent reviewer, and a release condition.

## Capital and revenue, separated

Capital allocations are granted at programme inception, not annually re-confirmed. The November 2025 refusal of capital funding for e-referrals and the integration hub — while milestones requiring those systems remained on the books — is the structural failure mode this addresses. Capital and revenue follow different cycles for a reason; aligning them under a single annual commissioning rhythm pretends they are the same thing, and the pretence breaks every year.

## Continuity protection

Funding is owned by the programme, not by the executive personalities running it. Leadership change at DHCW does not interrupt programme funding. The programme's accountable senior responsible owner — typically a clinical director embedded in a health board under [Flip the Model](/interventions/flip-the-model/) — has signing authority over draw-down within the envelope. DHCW's successor body has visibility but not veto.

## Health-board cycle alignment

Welsh Government planning cycles align with health-board Integrated Medium-Term Plan (IMTP) cycles. Currently they do not — which means a health board commits to operational support for a national programme based on funding visibility DHCW does not yet have. Funding letters not confirmed until 25% through the financial year (July 2022) is the documented precedent. The new design: WG indicative settlements are issued before health-board IMTP submission deadlines, not after them.

## What this intervention is NOT

This intervention is not a three-year envelope to *the organisation*. That becomes a shield against accountability — the kind of shield Glazzard implicitly identified, five years in: "We've always struggled because it's one-year funding. We should have worked it out by now, surely." Multi-year funding to the organisation produces multi-year insulation. Multi-year funding to programmes, with externally-validated milestones, produces multi-year continuity for delivery.

The test of the design: if DHCW leadership change disrupts programme funding, the design has failed. Funding belongs to the work, not to the leadership.

## How this relates to Reform the Funder

This intervention solves the cadence — annual cycles to multi-year envelopes, organisation to programme. [Reform the Funder](/interventions/reform-the-funder/) addresses the broader funder behaviour: capital and revenue coherence, RAG honesty, milestone realism, remit-letter discipline, co-author accountability. The two are complementary. Multi-year programme envelopes without RAG honesty produce well-funded programmes that are still being pressured to mark slipping milestones green. Both interventions are needed.

This intervention dismantles [L3: The Funding Uncertainty Trap](/diagnosis/l3-funding-uncertainty-trap/).


## Intervention 6: Reform the Funder
URL: https://bluenhs.org/interventions/reform-the-funder/

_Capital and revenue coherence, RAG honesty, milestone realism, remit-letter discipline. Without reforming Welsh Government's role as funder and overseer, replacing DHCW's leadership reproduces the same conditions for any successor body._

Welsh Government did not merely fail to oversee DHCW. It materially contributed to the conditions that produced the failure: red ratings were pressured down (Evans, Performance and Delivery Committee, November 2024); DPIF was cut from £33M planned to £28M allocated for 2024-25; capital funding for e-referrals and the integration hub was refused in November 2025 while milestones requiring those very systems were retained; LIMS was compressed from a four-year programme into two; recruitment was frozen by remit letter while delivery was tightened. Replacing DHCW's leadership without reforming the funder leaves these conditions in place for any successor body.

This intervention is parallel to I1-I5, owned by Welsh Government and the Senedd Public Accounts Committee. It has five elements.

1. **Capital and revenue coherence.** Capital allocations cannot lag the programmes that require them. Settlement before fiscal year start. The November 2025 refusal of capital funding for e-referrals and the integration hub, while milestones requiring those systems remained on the books, is the structural failure mode this addresses. The new procedural rule: no programme milestone is binding until the capital envelope underneath it is confirmed.

2. **RAG honesty.** Welsh Government ceases to pressure DHCW to soften red RAG ratings. A published RAG criteria document. A RAG audit trail visible to the Senedd PAC: every change of rating logged with author, date, and rationale. Evans's Performance and Delivery Committee admission that "Government was asking... why then are those programs rated red amber? If you've got a reasonable confidence that you will hold the implementation dates" is the precedent: this rule prevents its repetition.

3. **Milestone realism.** Externally-validated milestone setting at programme inception. No compressed timelines without published risk acknowledgement. The LIMS compression from four years to two was a deliberate decision with predictable consequences (Michelle Sell, PDC February 2025). The new rule: any compression beyond initial scoping requires a named accountable individual at Welsh Government, a published risk register, and milestone deferral if the conditions underpinning the compression change.

4. **Remit-letter discipline.** Recruitment freezes paired with milestone deferral, not concurrent with milestone tightening. The 2026 remit-letter freeze on recruitment while delivery was being escalated created a structural impossibility — Glazzard, six years in: "We've always struggled because it's one-year funding. We should have worked it out by now, surely." The new rule: any reduction in input authority (recruitment, capital, programme staffing) is paired with proportionate output relief — milestone deferral, scope reduction, or formal recognition that delivery is contingent on the input being restored.

5. **Co-author accountability.** When Welsh Government sets a condition that contributes to programme failure, Welsh Government is named in the post-mortem alongside DHCW. The entity issuing the corrective signal cannot be invisible in the post-mortem. The Performance and Delivery Committee, the Cabinet Secretary, and the responsible Director General each carry named accountability for conditions they imposed. Audit Wales is empowered to examine Welsh Government decisions as causally upstream of DHCW outcomes — not just DHCW decisions in isolation.

**Why this intervention runs in parallel, not in sequence.** Unlike Interventions 2-5, which depend on Competent Leadership being in place at DHCW, this intervention is owned by a different actor (Welsh Government and the Senedd) and addresses a different governance level. It can — and must — begin in Months 0-3, alongside the forensic review under Intervention 1. Without it, the same RAG pressure, capital refusals, milestone compressions, and remit-letter tightening will be applied to whichever leadership team replaces the current one. The conditions reproduce. Reform of DHCW alone is not reform.

This is the structural precondition. Interventions 1-5 reform DHCW. Intervention 6 reforms the funder, the overseer, and the political environment in which DHCW operates. Both are required.


# Named NHS Wales National Systems
URL: https://bluenhs.org/systems/

_One-page references for the eight national NHS Wales digital systems named across the Blueprint analysis: WPAS, RISP, LIMS, EPMA, eMPI, WCCG, WICIS, and PSBA. Each entry gives what the system is, its current status, the documented issues, and links to where it is discussed in the diagnosis._

The Blueprint analysis names eight national NHS Wales digital systems by acronym. Each has a one-page reference below — what it is, what it does, its current status, the patient-safety record where one exists, and where it is discussed across the diagnosis.

These pages are reference, not advocacy. They compile what is documented in DHCW board minutes, programme reports, Welsh Government correspondence, and the Blueprint's diagnosis. Citations link back to the relevant feedback-loop or trap analyses.


## Welsh Patient Administration System (WPAS)
URL: https://bluenhs.org/systems/wpas/

_WPAS — the Welsh Patient Administration System — is the primary national clinical administration system for NHS Wales. It has been linked to at least one patient death and carries documented patient-safety risk that has not been resolved._

## What WPAS does

WPAS is the patient-administration backbone of NHS Wales — recording inpatient admissions, outpatient appointments, and waiting-list status across the seven Welsh health boards. It is the system that determines, in clinical workflow terms, whether a patient is on a list, when they are seen, and how their journey is recorded.

## Status

In operational use across NHS Wales. No documented programme to replace it within a clinical-safety-driven timeframe.

## Patient-safety record

WPAS has been linked to at least one patient death — a fact named in DHCW board papers and not contested. The clinical-safety case has been the subject of repeated internal challenge from technical staff. Resolution has not followed.

The pattern is the one the Blueprint's [Drift to Low Performance trap](/traps/drift-to-low-performance/) describes: safety incidents are recorded, the safety case is renewed, and the system continues in operational use because there is no alternative supplier and no plausible replacement timeline. The structural condition is the [once-for-Wales monopoly](/blueprint/once-for-wales/) — health boards cannot procure an alternative; they must use WPAS or have nothing.

## Where this is discussed in the diagnosis

- [Drift to Low Performance](/traps/drift-to-low-performance/) — the trap that produces the WPAS pattern.
- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 (patient-safety triage) names WPAS explicitly.
- [Once for Wales](/blueprint/once-for-wales/) — the monopoly framing that prevents health boards from replacing WPAS.


## Regional Imaging Solution / RISP (RISP)
URL: https://bluenhs.org/systems/risp/

_RISP — the Welsh radiology and imaging programme — has a documented 'global worklist' issue that threatens cross-Wales image viewing. The programme has been a regular feature of DHCW board escalation discussions._

## What RISP does

RISP is the national radiology and imaging programme for NHS Wales — designed to provide unified access to imaging studies across the seven health boards. The clinical case is straightforward: a consultant in Swansea should be able to view a Betsi Cadwaladr patient's prior imaging instantly. The technical case is harder.

## Status

In delivery, with documented issues. RISP has been a recurring feature of DHCW board escalation discussions.

## Documented issues

- A "global worklist" issue threatens cross-Wales image viewing — flagged as a clinical risk in DHCW papers.
- In March 2026, RISP was one of four national systems taken offline simultaneously by a single failure in PSBA-class shared infrastructure (alongside O365, EPMA, and radiology). See [PSBA](/systems/psba/) for the infrastructure dimension of that outage.
- The full business case for RISP was moved into private session under "commercial sensitivity" in DHCW board discussions — one of the 51 documented instances of board approvals without scrutiny.

## Where this is discussed in the diagnosis

- [Captured Governance](/diagnosis/l11-oversight-obstruction/) — RISP appears in the discussion of business cases moved into private session.
- [Tragedy of the Commons](/traps/tragedy-of-the-commons/) — RISP is one of the nine programmes competing for shared delivery capacity.
- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 names RISP for clinical-safety triage.


## Laboratory Information Management System (LIMS)
URL: https://bluenhs.org/systems/lims/

_LIMS — the national NHS Wales Laboratory Information Management System programme — was compressed by Welsh Government from a four-year programme into two years. It was one of two Phase One Level 4 milestones documented as missed in March 2026._

## What LIMS does

LIMS is the national NHS Wales Laboratory Information Management System programme — the unified pathology and laboratory IT layer intended to support all seven health boards. Clinical-laboratory workflows (specimen tracking, result reporting, pathology storage and retrieval) depend on it.

## Status

Delayed. Documented as a missed milestone in Phase One of DHCW's Level 4 Targeted Intervention assessment in March 2026 (alongside WRISTS).

## Documented issues

- Welsh Government compressed LIMS from a four-year programme into a two-year programme. The technical complexity did not reduce; only the timeline did.
- The compressed timeline ran in parallel with a WG-imposed recruitment freeze on DHCW via the remit letter. Delivery was accelerated while hiring was frozen.
- The LIMS full business case was moved into private session under "commercial sensitivity" in DHCW board discussions.
- LIMS was one of two of forty-seven Phase One milestones DHCW missed under Level 4 monitoring in March 2026.

The pattern is the one [L3: The Funding Uncertainty Trap](/diagnosis/l3-funding-uncertainty-trap/) describes — funding signals and remit-letter demands arriving on incompatible timelines, with the delivery body absorbing the gap.

## Where this is discussed in the diagnosis

- [L3: The Funding Uncertainty Trap](/diagnosis/l3-funding-uncertainty-trap/) — the structural mechanism.
- [Captured Governance](/diagnosis/l11-oversight-obstruction/) — LIMS appears as one of the missed Phase One milestones.
- [Intervention 5: Break the Annual Trap](/interventions/break-annual-trap/) — multi-year programme funding that LIMS specifically required.
- [Intervention 6: Reform the Funder](/interventions/reform-the-funder/) — addresses the milestone-realism failure that produced the compressed timeline.


## Electronic Prescribing and Medicines Administration (EPMA)
URL: https://bluenhs.org/systems/epma/

_EPMA — the national NHS Wales Electronic Prescribing and Medicines Administration programme — illustrates the accountability gap the Blueprint diagnoses. DHCW is measured against milestones for EPMA delivery, but the accountable delivery body is the health boards, not DHCW._

## What EPMA does

EPMA is the national NHS Wales programme for Electronic Prescribing and Medicines Administration — the clinical IT layer for prescribing, dispensing, and medicines management. It is one of the highest-clinical-impact digital programmes in NHS Wales: incorrect prescribing or missed administration produces direct patient harm.

## Status

In rollout. Local organisations (the seven health boards) are the accountable delivery body; DHCW provides the platform.

## The accountability paradox

EPMA is the clearest example of the accountability gap the Blueprint diagnoses. Ifan Evans named it in November 2024:

> "This persistent ambiguity, let's call it, which we usually phrase as 'we need to clarify roles and responsibilities'… PDC is an oversight and assurance function, not a programme management function."

DHCW is being measured against EPMA programme milestones, but the accountable delivery body is the health boards. DHCW has no mechanism to hold the health boards to account; the health boards have no contractual relationship that gives DHCW that authority. The result is a programme where every party can attribute slippage to a different party, and no party owns the recovery plan.

## Documented issues

- The accountability gap was named at the Performance and Delivery Committee 46 months after DHCW's founding and was never closed.
- In March 2026, EPMA was one of four national systems taken offline simultaneously by a single PSBA-class infrastructure failure (alongside O365, RISP, and radiology).
- EPMA delivery slippage is regularly cited in DHCW board papers without a corresponding accountability conversation about who is supposed to deliver what.

## Where this is discussed in the diagnosis

- [L11: Captured Governance](/diagnosis/l11-oversight-obstruction/) — Evans's quote about the EPMA accountability gap appears here.
- [Policy Resistance](/traps/policy-resistance/) — EPMA is the canonical example of multiple actors pulling in incompatible directions.
- [Intervention 4: Flip the Model](/interventions/flip-the-model/) — embedded delivery teams in health boards close the accountability gap by design.


## Welsh Master Patient Index (EMPI)
URL: https://bluenhs.org/systems/empi/

_eMPI — the Welsh master patient index — is the national patient-identity reconciliation system for NHS Wales. It has been linked to patient record mixups, with documented incidents in DHCW board papers._

## What eMPI does

eMPI — the electronic Master Patient Index — is the national patient-identity reconciliation system for NHS Wales. It links a patient's records across multiple systems (WPAS, RISP, EPMA, WCCG, primary care) using a canonical patient identifier. Patient identity is the foundational data quality dependency for every clinical IT system downstream of it.

## Status

In operational use. No documented replacement programme on a clinical-safety-driven timeframe.

## Documented issues

- eMPI has been linked to patient record mixups — cases where one patient's data has been associated with another's record. The clinical-safety implications are direct: a clinician acting on the wrong record can deliver the wrong intervention.
- Patient-identity errors in master-index systems are particularly dangerous because they are silent — the system does not flag uncertainty; the wrong data simply arrives at the clinical workflow.
- eMPI's record-mixup incidents are documented in DHCW board papers but the safety case has not produced structural correction.

## Where this is discussed in the diagnosis

- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 names eMPI specifically for patient-safety triage ("patient record mixups").
- [Drift to Low Performance](/traps/drift-to-low-performance/) — the trap that allows a system with known patient-safety incidents to remain in operational use without remediation.


## Welsh Clinical Communications Gateway (WCCG)
URL: https://bluenhs.org/systems/wccg/

_WCCG — the Welsh Clinical Communications Gateway — has run on technology unsupported by its vendor for 8+ years, against repeated warnings from DHCW's own technical staff. It is the canonical example of the drift-to-low-performance trap._

## What WCCG does

WCCG — the Welsh Clinical Communications Gateway — is the national electronic referral and clinical communications platform across NHS Wales. It carries the electronic-referral traffic between primary care and secondary care, alongside other clinical communications. It is one of the highest-volume clinical messaging systems in NHS Wales.

## Status

In operational use, running on technology unsupported by its vendor for 8+ years.

## Documented issues

- WCCG has run on technology that is no longer supported by its vendor for more than eight years. Vendor end-of-support means no security patches, no compatibility updates, no upgrade path.
- DHCW's own technical staff have repeatedly warned about the WCCG technology stack. The warnings are documented in technical risk registers and have been escalated through DHCW governance channels.
- The warnings have not produced replacement or remediation. Each annual technical-risk review has accepted the same risk that the previous review accepted.

## Why this matters structurally

WCCG is the canonical example of the [Drift to Low Performance trap](/traps/drift-to-low-performance/). Standards have drifted past "inefficient" through "wasteful" to "dangerous", and each new low has become the new baseline. The technology being unsupported is not an emergent risk — it has been a *known* risk for 8+ years, accepted at every governance level.

The structural mechanism is what [L11: Captured Governance](/diagnosis/l11-oversight-obstruction/) describes: technical staff raise the concern → the concern enters governance channels → the governance channels accept the risk → the risk is renewed next year. The signal does not produce action because the loop that translates technical warning into structural change has been neutralised.

## Where this is discussed in the diagnosis

- [Drift to Low Performance](/traps/drift-to-low-performance/) — WCCG named explicitly.
- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 names WCCG for clinical-safety triage.
- [L11: Captured Governance](/diagnosis/l11-oversight-obstruction/) — the governance mechanism that has let the 8+-year risk persist.


## Welsh Integrated Cancer Information System (WICIS)
URL: https://bluenhs.org/systems/wicis/

_WICIS — the Welsh Integrated Cancer Information System — is the national cancer IT layer for NHS Wales. As of recent DHCW board papers, the programme is 'effectively still on pause'._

## What WICIS does

WICIS is the national NHS Wales Integrated Cancer Information System — the cancer-pathway IT layer intended to support the clinical management of cancer patients across the seven health boards. The clinical use case is direct: cancer pathways are time-sensitive, multi-specialty, and frequently cross health-board boundaries. A unified cancer-IT layer is the digital precondition for sustained cancer-waiting-time recovery.

## Status

"Effectively still on pause", per recent DHCW board papers. The programme is neither closed nor advancing.

## Documented issues

- WICIS has been in a pause state for an extended period. The pause has not been formally declared as a closure; the programme remains nominally active in the portfolio.
- Cancer waiting times in NHS Wales remain a sustained area of public and political concern. The digital system intended to support cancer-pathway management is, structurally, not available to support that concern.
- WICIS is one of nine programmes competing for shared DHCW delivery capacity. It is among the programmes that has been deprioritised in practice while remaining open on paper.

The pattern is the [Tragedy of the Commons](/traps/tragedy-of-the-commons/) — nine programmes competing for capacity that cannot satisfy all of them. WICIS is the casualty of that dynamic; cancer pathways are the population it would have served.

## Where this is discussed in the diagnosis

- [Tragedy of the Commons](/traps/tragedy-of-the-commons/) — WICIS is one of the nine in-flight programmes.
- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 names WICIS for clinical-safety triage ("effectively still on pause").
- [Intervention 3: Portfolio Ruthlessness](/interventions/portfolio-ruthlessness/) — the discipline of formally closing or actively resuming paused programmes rather than leaving them indefinitely.


## Public Sector Broadband Aggregation (PSBA)
URL: https://bluenhs.org/systems/psba/

_PSBA — Public Sector Broadband Aggregation — provides shared network infrastructure across the Welsh public sector including NHS Wales. In March 2026 a single PSBA-class failure took O365, EPMA, RISP, and radiology offline simultaneously across NHS Wales._

## What PSBA does

PSBA — Public Sector Broadband Aggregation — provides shared network infrastructure for the Welsh public sector, including NHS Wales. PSBA-class shared infrastructure is the network underlay on which national NHS Wales clinical systems run. When PSBA-class infrastructure is unavailable, the systems that depend on it are unavailable.

## Status

In operational use. The March 2026 incident exposed structural concentration risk that has not been remediated.

## The March 2026 incident

In March 2026, a single failure in PSBA-class shared infrastructure took **four national systems offline simultaneously** across NHS Wales:

- **O365** — productivity / communications platform across the seven health boards.
- **EPMA** — Electronic Prescribing and Medicines Administration.
- **RISP** — Regional Imaging Solution.
- **Radiology** — additional radiology services.

This is the operational manifestation of concentrated infrastructure dependency. A single failure point cascaded across four clinically-significant systems at once. The incident was recorded in DHCW board papers. Structural remediation — separating shared-infrastructure concerns so that a single failure cannot take out four clinical layers simultaneously — has not followed.

## Why this matters structurally

PSBA illustrates a specific risk that the Blueprint analyses across multiple loops: when delivery is concentrated in one body using one set of shared infrastructure components, the failure modes are correlated rather than independent. Independent failures degrade gracefully; correlated failures cascade. NHS Wales digital infrastructure exhibits correlated-failure behaviour because it is structurally organised that way.

The remedy is architectural separation: shared standards-setting with distributed delivery and distributed infrastructure dependencies. This is the structural argument of [Intervention 4: Flip the Model](/interventions/flip-the-model/) — the once-for-Wales monopoly produces single-point-of-failure dynamics even at the infrastructure layer.

## Where this is discussed in the diagnosis

- [Drift to Low Performance](/traps/drift-to-low-performance/) — PSBA-class outage cited as one of the dangerous-state indicators.
- [Intervention 1: Competent Leadership](/interventions/competent-leadership/) — Action 6 names PSBA for clinical-safety triage.
- [Intervention 4: Flip the Model](/interventions/flip-the-model/) — the structural separation that would prevent correlated-failure dynamics.
- [Once for Wales](/blueprint/once-for-wales/) — the monopoly framing under which PSBA-class concentration has been allowed to deepen.


# About Systems Dynamics
URL: https://bluenhs.org/methodology/

_Why systems dynamics. Stocks, flows, feedback loops, delays, and Meadows's seven traps applied to DHCW. Forty-year-old MIT discipline; falsifiable diagnosis._

Most organisational analysis looks at problems in isolation. A programme is late — hire more people. A vendor underperformed — change vendor. Trust is low — publish a new strategy. DHCW has tried all three. None changed the outcome.

When repeated fixes fail, the problem is not the fix. It is the structure regenerating the problem. Systems dynamics, developed by Jay Forrester at MIT and formalised by Donella Meadows in _Thinking in Systems_, names that structure. It replaces "what broke?" with "what pattern keeps producing breakage?" — and gives a disciplined vocabulary: stocks, flows, feedback loops, delays, and traps.

Root-cause analysis looks for a single upstream failure and stops. Consulting frameworks describe what an organisation should look like. Systems dynamics describes why it looks the way it does.

## Stocks — What Accumulates

A stock is anything that accumulates or depletes over time. Money and headcount are stocks. So are trust, institutional knowledge, delivery capability, and political capital — the things that actually determine whether an organisation functions.

Stocks are what a competent board should be watching. They are also what captured governance hides. At DHCW the visible stocks (~£200M annual budget reaching that level by 2025-26, ~£600M cumulative Welsh Government funding over five years, ~1,263 staff, 99.984% availability) look healthy. The invisible stocks (health-board trust, clinician confidence, delivery capability, institutional knowledge) are all depleting. The published record of the visible stocks is itself distorted — 107 sanitisation instances and 237 hiding-intent passages in the deep evidence base — so even the stocks that *can* be measured arrive softened. Management attention focuses on flows — this quarter's hiring rate, this quarter's spend — which is why headcount can grow 80% while delivery capability falls. The flow looks healthy. The stock tells the real story.

The full 18-stock inventory — organised by visibility — lives on the [diagnosis page](/diagnosis/). Each of the eleven feedback loops depletes a specific stock.

## Flows — What Changes The Stocks

Flows are the rates at which stocks accumulate or drain. A hiring flow raises headcount. A resignation flow lowers it. A knowledge-transfer flow builds institutional knowledge. A knowledge-loss flow — through attrition, silencing, or loyalty-based promotion — erodes it.

At DHCW the flows are visible and the stocks are not. The hiring flow has run at record rates for three years. The delivery flow — programmes actually shipped to health boards — has barely moved. The knowledge-loss flow accelerates every time an experienced engineer leaves. Measuring flows without measuring the stocks they feed is how a governance system mistakes motion for progress.

## Feedback Loops — Reinforcing vs Balancing

A feedback loop is a circular chain of cause and effect in which a change in one element comes back to affect that same element. A _reinforcing_ loop amplifies whatever is happening — when the dynamic is harmful, it becomes a vicious cycle. A _balancing_ loop corrects toward a target, like a thermostat. In a healthy organisation, balancing loops keep reinforcing loops in check.

At DHCW every balancing loop has been captured or neutralised. The diagnosis identifies eleven reinforcing loops in two clusters.

**Cluster A — five reinforcing loops of delivery failure.** Hiring without delivery, credibility death spiral, annual-funding paralysis, the rebranding escape, vendor-dependency spiral. Any national health IT body under similar constraints would face some version of these.

**Cluster B — six reinforcing loops of self-preservation.** Manufactured narrative, competence void, loyalty selection, whistleblower suppression, information fortress, oversight obstruction. Cluster B is not structural accident. It is an active engine that intercepts every corrective mechanism before it can reach Cluster A.

Cluster B has an upstream source. Three successive NHS Wales CEOs were drawn from a single health board (ABUHB), via a patronage pipeline that pre-credentialled three DHCW executive directors with UWTSD Professor of Practice titles in December 2020 — four months before DHCW's founding board meeting. The loops did not emerge gradually. They were imported.

All eleven are documented on the [diagnosis page](/diagnosis/).

## Delays — Why Intervention Feels Unrewarding At First

A delay is the time gap between action and consequence. Delays are why sensible interventions look like failures in the short term, and why destructive interventions look like successes. Hire 600 people and expect productivity to rise next quarter; it does not — new hires take 6 to 18 months to become productive. The leader hires 300 more to compensate. That is the classic systems-dynamics overshoot.

Trust is the slowest stock of all. Rebuilding credibility with a health board after a failed programme takes two to five years of consistent delivery — not two to five quarterly updates. Any measurement regime shorter than those delays will declare the right interventions failed and the wrong ones successful. Multi-year measurement is not a preference. It is a precondition for seeing anything at all.


## The 12 Leverage Points
URL: https://bluenhs.org/methodology/leverage-points/

_Meadows' hierarchy of places to intervene in a system — from shallow parameter tweaks to deep paradigm change. DHCW has operated almost exclusively at the shallow end._

Not all interventions are equal. Changing a budget number and changing an organisation's goal are both "interventions," but they operate at entirely different depths and produce entirely different results. Meadows' 1999 essay _[Leverage Points: Places to Intervene in a System](https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/)_ formalises this intuition into a twelve-level hierarchy, ordered from shallowest (least structural change per unit effort) to deepest.

The hierarchy explains a counter-intuitive pattern that recurs across decades of systems-dynamics case studies: the interventions that feel most concrete and actionable — hire more people, increase the budget, reorganise the org chart — are precisely the ones with the least power to change system behaviour. They operate on parameters and structures that the system's feedback loops will simply compensate for. The interventions that actually change behaviour target the feedback loops themselves, the information flows that feed them, the rules that constrain them, or the goals and paradigms that orient them.

## The Hierarchy

| Level | Intervenes on | Power | DHCW example |
|---|---|---|---|
| 12 | Parameters (numbers, budgets, quotas) | Lowest | "Hire 233 more people in one year" (22.6% headcount growth in 2024-25) |
| 11 | Buffer sizes (stabilising reserves) | Low | Cash reserves, bench staff |
| 10 | Stock-and-flow structures (org charts, infrastructure) | Low | "Building our Future" reorganisation |
| 9 | Delays (time gaps between action and consequence) | Medium | Shorten feedback cycles |
| 8 | Balancing feedback strength (corrective mechanisms) | Medium | Strengthen accountability |
| 7 | Reinforcing feedback gain (growth rate of amplifiers) | Medium | Accelerate virtuous cycles |
| 6 | Information flows (who sees what, when) | High | **Radical Transparency (Intervention 2)** |
| 5 | System rules (incentives, constraints, rewards) | High | **Break the Annual Trap (Intervention 5)** and **Reform the Funder (Intervention 6)**; legal protections |
| 4 | Self-organisation (power to restructure) | High | **Flip the Model (Intervention 4)** — health-board led |
| 3 | System goals (what the system optimises for) | Very high | Redefine success as clinician adoption and patient-safety outcomes, not headcount or programme count |
| 2 | Mindset / paradigm (shared assumptions) | Highest practical | **Competent Leadership (Intervention 1)** — stewardship, not entitlement |
| 1 | Transcending paradigms | Theoretical | Question whether a monopoly delivery body should exist |

## Why Organisations Target the Shallow End

The shallow levels (10–12) are where organisations instinctively reach because they are visible, measurable, and politically manageable. Hiring more people is a concrete action with a press release. Changing the information architecture of an organisation — who sees what, when, and what they can do about it — is abstract, disruptive, and threatens the people who currently control information flow. The deeper the level, the more resistance it provokes from the existing system, which is precisely why it is more effective.

This pattern is not a DHCW-specific observation. It is a fifty-year finding of the discipline. Forrester's _Urban Dynamics_ showed that housing construction programmes (Level 12) made urban decline worse. Senge's _The Fifth Discipline_ documents corporate turnarounds that only succeeded when intervention moved from parameters to paradigms. The hierarchy is predictive, not just descriptive.

## What This Means for the Blueprint

DHCW has operated almost exclusively at Levels 10–12. Every reform attempt — hiring surges, reorganisations, new programme labels — has targeted parameters and structures. The [eleven feedback loops](/diagnosis/feedback-loops/) have absorbed each intervention without changing behaviour. This is not bad luck. It is the predicted outcome of shallow intervention in a system with deep structural dysfunction.

The [blueprint](/blueprint/) targets Levels 2–6. That is why it is called a *structural* intervention — the level number directly maps the depth of change being sought. Competent leadership (Level 2) changes the paradigm. Radical transparency (Level 6) changes information flows. Flipping to health-board-led delivery (Level 4) changes self-organisation. Reforming the funder (Level 5) changes the rules under which DHCW operates. These are the levels at which the evidence base says structural change actually occurs.


## System Traps
URL: https://bluenhs.org/methodology/system-traps/

_Donella Meadows catalogued seven recurring patterns of dysfunction. Most failing organisations exhibit two or three. DHCW has all seven active simultaneously — no documented precedent._

A single feedback loop can cause harm. But the most dangerous dynamics in any organisation come not from individual loops but from the way multiple loops interact. When two or more loops lock together they produce a recognisable, repeating pattern of dysfunction that Donella Meadows called a **system trap**. Traps are dangerous precisely because they are self-reinforcing: each failed attempt to fix the problem strengthens the pattern that produced it. Recognising which trap you are in is the first analytical step, because it eliminates entire classes of solutions that look plausible but structurally cannot work.

## Why Traps Matter More Than Individual Loops

An individual loop has a single mechanism and, in principle, a single fix. A trap has a _structure_ — multiple loops feeding each other in a configuration that resists correction. The trap persists even if you break one of its constituent loops, because the remaining loops compensate. This is why organisations that "fix" individual problems — hire more staff, change vendors, publish a new strategy — find the same dysfunction reappearing in a different form. They are treating loops when the problem is the trap.

Meadows identified the escape route for each trap. The escape is never "try harder at what you are already doing." It always requires a qualitative shift: changing the information structure, changing the goal, changing the rules, or changing who holds power. That is why the [leverage points hierarchy](/methodology/leverage-points/) matters — shallow interventions (Levels 10–12) cannot escape traps that operate at deeper levels.

## The Seven System Traps

| Trap | Signature | What it looks like at DHCW |
|---|---|---|
| **Shifting the Burden** | A symptomatic fix weakens the capacity for a fundamental fix | Hiring contractors instead of building capability; Evans authoring the next ten-year strategy after failing to deliver the current one |
| **Drift to Low Performance** | The performance standard erodes to match actual performance | WPAS linked to a patient death; eight years of unsupported WCCG; the March 2026 PSBA outage across all NHS Wales |
| **Seeking the Wrong Goal** | The system optimises for a measurable proxy instead of the real goal | UWTSD professorships and FedIP credentials pursued while measured harm continued; 14-15 sub-strategies for an 800-person organisation |
| **Policy Resistance** | Multiple actors pull in different directions; effort cancels out | Health boards work around DHCW; Welsh Government escalates DHCW for conditions WG materially co-authored |
| **Tragedy of the Commons** | Each actor's rational self-interest depletes a shared resource | LIMS and RISP running concurrently against the same architecture and integration capacity ("significant pressure on resources across NHS Wales") |
| **Success to the Successful** | Winners accumulate advantage; losers lose capacity to compete | Tier 1 loyalists pre-credentialled in December 2020; technical leaders dismissed for raising concerns |
| **Escalation** | Each side responds to the other's response, ratcheting upward | Level 3 enhanced monitoring → Level 4 Targeted Intervention; information control intensifying alongside Employment Tribunal proceedings |

## Seven Simultaneous Traps

Most failing organisations exhibit two or three traps. DHCW has all seven active simultaneously. The systems-dynamics literature contains no documented precedent for this density. Seven traps do not activate in parallel by accident. They require a coordinating mechanism — a set of loops that actively maintains the conditions each trap needs. At DHCW, the *operational* mechanism is Cluster B: six reinforcing loops of self-preservation that intercept every corrective signal before it can reach the delivery failures in Cluster A. The *upstream* mechanism is older: three successive NHS Wales CEOs drawn from a single health board (ABUHB), and three DHCW directors pre-credentialled in December 2020 with UWTSD Professor of Practice titles — four months before the founding board met. The patronage architecture is what Cluster B operationalises. The traps were already wired together when DHCW opened for business; the loops then ran them.

Every governance deficit pattern observed at Level 3 escalation 34 months in was already operational at the very first board meeting. The architecture was complete on day one — exactly what systems-dynamics theory predicts of a system built with this combination of feedback loops and self-preservation mechanisms.

This is why the [blueprint](/blueprint/) sequences its interventions through the trap structure. You cannot escape _shifting the burden_ until the real problem is visible. You cannot escape _policy resistance_ until the actors resisting change have been removed. You cannot escape _seeking the wrong goal_ until the goal itself has been redefined. The traps dictate the order.

Each trap is walked through in detail — archetype, DHCW manifestation, constituent loops, and escape route — on the [system traps page](/traps/).


## Further Reading
URL: https://bluenhs.org/methodology/further-reading/

_The systems thinking canon — Meadows, Forrester, Senge, and fifty years of case studies. The methodology applied here is not bespoke; it is a forty-year-old discipline with a substantial body of practice._

The methodology applied here is not bespoke to this analysis. It is a forty-year-old discipline with a substantial body of practice. The sources below are not background reading — they are the primary literature from which the [core concepts](/methodology/), [leverage hierarchy](/methodology/leverage-points/), and [system traps](/methodology/system-traps/) used throughout this site are drawn. A reader who checks these sources will find that the analytical framework predicts exactly the patterns observed at DHCW.

**The author of the framework.** [Donella H. Meadows](https://donellameadows.org/) (1941-2001) — environmental scientist, MIT-trained systems analyst, and lead author of the 1972 Club of Rome report. Her posthumously published _[Thinking in Systems: A Primer](https://www.chelseagreen.com/product/thinking-in-systems/)_ (Chelsea Green, 2008) is the canonical short introduction and the source of the seven system traps used on this site. Her 1999 essay _[Leverage Points: Places to Intervene in a System](https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/)_ is the source of the twelve-level leverage hierarchy that structures the [blueprint](/blueprint/). The [Donella Meadows Project](https://donellameadows.org/) at the Academy for Systems Change maintains her archive.

**The founder of the discipline.** [Jay W. Forrester](https://mitsloan.mit.edu/ideas-made-to-matter/system-dynamics-a-soft-science-real-world-impact) (1918-2016) — MIT engineer, inventor of magnetic-core memory, founder of the System Dynamics Group at MIT Sloan. His 1961 _Industrial Dynamics_, 1969 _Urban Dynamics_, and 1971 _World Dynamics_ established the modelling tradition from which Meadows' work descends. The [MIT System Dynamics Group](https://mitsloan.mit.edu/faculty/academic-groups/system-dynamics) continues the research programme today.

**Case studies demonstrating the method's potency.** Forrester's _Urban Dynamics_ explained why housing-construction programmes in 1960s US cities reliably made urban decline worse — counter-intuitive at the time, confirmed by decades of evidence. The _Limits to Growth_ team applied system dynamics to global population, resources, pollution, and capital; the 1972 projections have tracked actual world trends for fifty years, with updates in _Limits to Growth: The 30-Year Update_ (2004). Peter Senge's _The Fifth Discipline_ (1990; 2006 revised) brought the same tools into organisational practice and documents dozens of corporate turnarounds driven by loop analysis. In public health, the [NASPAA systems-thinking network](https://www.naspaa.org/) and NHS England's [Improvement Academy](https://improvementacademy.org/) both treat Meadows' framework as foundational for analysis of service-delivery failure.

**Why this matters for DHCW.** Every corrective action DHCW has taken operates at Meadows' "shallow end" — Levels 10 through 12 (parameters, buffers, structural cosmetics). Fifty years of systems-dynamics practice show those levels do not change system behaviour. They absorb effort without altering outcome. That prediction — made by the discipline, not by this author — is exactly what has happened at DHCW: every governance deficit pattern observed at Level 3 escalation 34 months in was already operational at DHCW's very first board meeting. The discipline forecast the failure mode; the evidence shows it unfolding on the predicted timeline. The [blueprint](/blueprint/) targets Levels 2 through 6 because that is where the evidence base says structural change actually occurs.

**What is not in these sources.** The systems-dynamics canon provides the analytical framework. It does not provide the evidence about DHCW. The evidence — Senedd proceedings, Audit Wales reports, Employment Tribunal filings, FOI material, and witness testimony — is documented at [carenhs.org](https://carenhs.org). The [methodology page on applying this to DHCW](/methodology/applying-to-dhcw/) describes how the framework and the evidence connect.


## System Boundary
URL: https://bluenhs.org/methodology/system-boundary/

_What this analysis scopes inside, outside, and on-the-boundary. Welsh Government, the patronage pipeline that pre-credentialed the executive cohort, and the oversight function are all on the boundary — partly external, but materially shaping internal dynamics._

Defining the system boundary is the first step in any systems-dynamics analysis. It is also the step most often skipped — and skipping it is how analyses produce confident recommendations aimed at things no one can actually change, or ignore the things they can.

Everything **inside** the boundary is something the analysis treats as subject to change. Everything **outside** is treated as context — real but not the focus of intervention. Elements **on the boundary** are the uncertain ones: partly within reach, partly not. Drawing the boundary wrong in either direction produces characteristic failures. Too narrow, and the analysis misses the feedback loops that regenerate the problem from outside the frame. Too wide, and it recommends interventions that require actors or authority the reformers do not have. Both errors look like rigorous analysis. Both waste effort.

## Inside the System (DHCW can observe and influence)

- The ~1,263 staff and their allocation across programmes
- 9+ major programme portfolios and their delivery pipelines
- Vendor / contractor relationships (~£1.25B contract portfolio)
- Internal culture, hiring, retention, knowledge management
- Programme management methodology and delivery practices
- Stakeholder engagement with health boards and clinicians
- Board governance and executive decision-making
- **Leadership selection, promotion, and accountability mechanisms**
- **Internal information control and the sanitisation machine** — what gets reported in real time, what gets curated out before publication. 107 documented sanitisation instances and 237 hiding-intent passages across published board and committee minutes; in one transcript the curation ratio fell to 10.7%
- **Whistleblowing infrastructure** — how concerns are actually handled
- **The 23 off-payroll workers** operating as a shadow workforce

## Outside the System (environment, treated as fixed constraints)

- UK-wide digital talent market and salary competition
- NHS Wales structure (7 health boards, 3 trusts as sovereign bodies)
- Patient population needs and demographics
- Technology vendor market dynamics (Microsoft, InterSystems, CGI, etc.)
- Welsh political and electoral cycles
- Employment Tribunal legal proceedings
- **Patient-safety outcomes across NHS Wales** — the downstream effect that motivates the entire analysis. WPAS-linked harm, eMPI patient-record mixups, the cross-NHS-Wales PSBA outage all originate inside DHCW; the patient-experienced consequences accrue outside DHCW's direct control, on a ledger that money cannot reverse

## On the Boundary (uncertain influence)

The boundary elements are the most analytically important category, because they are where the system's actual reach diverges from its formal reach.

- **Welsh Government as funder and overseer** — formally an external control, but the deep evidence base establishes a stronger finding than partial capture: Welsh Government materially shaped the failure conditions. DPIF cut from £33M planned to £28M allocated (2024-25). Capital funding for e-referrals and the integration hub refused in November 2025 while milestones requiring those systems remained on the books. LIMS compressed from a four-year programme to two (Sell, PDC February 2025). Recruitment frozen via remit letter while delivery was tightened. Red RAG ratings pressured down (Evans, PDC November 2024). When the funder is materially shaping the failure conditions, the funder cannot be treated as a fixed external constraint. Reformed by [Intervention 6: Reform the Funder](/interventions/reform-the-funder/).

- **The patronage pipeline** — a pre-DHCW phenomenon with entirely internal effects. Three successive NHS Wales CEOs emerged from a single health board (ABUHB), via a pipeline controlled by Andrew Goodall (Welsh Government Permanent Secretary, formerly ABUHB CEO) and Judith Paget (Director General Health & Social Services, formerly ABUHB CEO). In December 2020 — four months before DHCW's founding board meeting — Helen Thomas, Rhidian Hurle, and Ifan Evans were awarded UWTSD Professor of Practice titles via the Wales Institute of Digital Information. The origin sits outside DHCW; the effect is the executive cohort that runs it.

- **Senedd Public Accounts Committee** — formally an external scrutiny body, but with explicit reform agency under [Intervention 6](/interventions/reform-the-funder/): mandating hearings, requiring testimony, commissioning Audit Wales investigations, scoring remit-letter content against milestone load. PAC cannot day-to-day manage transition, but it can install the structural conditions under which any successor leadership operates.

- **Cabinet Secretary for Health and Social Care** — formally external. But the minister's February 2026 ("a pattern of late notification that undermines system confidence") and March 2026 ("complex, data-heavy, burdensome, lacks transparency and does not drive improvement") public critiques of the WG-DHCW escalation framework materially shaped what reform looks like. When the minister chooses to act on the framework itself, the minister is on the boundary.

- **The carenhs.org evidence campaign** — external to DHCW, but the existence and accessibility of its material materially shapes DHCW behaviour. It functions as an alternative information flow that bypasses the [information fortress](/diagnosis/l10-information-fortress/) — an external element that alters internal dynamics.

The boundary itself has moved over the five-year window. Three of the sharpest internal counterweights — Chair Simon Jones, NED Rowan Gardner, Interim Chair Ruth Glazzard — were inside the system and have departed by 2025-26. Their loss is a structural change to what the system contains, not a personnel matter; the reform must rebuild that counterweight inside the boundary, not assume it survived the failure.

## Why This Matters

The [blueprint](/blueprint/) targets what is inside the boundary. Interventions aimed at "outside" elements (changing the Welsh Government budget cycle, or Wales' share of the UK talent market) require different tools and different actors. Interventions aimed at boundary elements (restoring the oversight function) must be designed with the knowledge that the boundary itself is contested.

Many previous analyses of DHCW have failed at this step. Internal reviews draw the boundary too narrowly — they scope out leadership behaviour and governance capture because those are uncomfortable to name. External reviews draw it too widely — they recommend changes to funding structures or cross-government coordination that no single reform can deliver. And reviews that treat Welsh Government as fixed external context miss that the funder materially shapes the failure conditions: DHCW is co-authored, not merely supervised.

This analysis draws the boundary where the evidence places it. Leadership conduct, information control, the sanitisation machine, and accountability mechanisms are inside. The political environment, the talent market, and the patient-safety ledger are outside. Welsh Government as funder, the patronage pipeline that pre-credentialled the executive cohort, the Senedd Public Accounts Committee, the Cabinet Secretary, and the carenhs.org evidence campaign are all on the boundary — partly external, but materially shaping internal dynamics. The blueprint addresses them through [Intervention 6: Reform the Funder](/interventions/reform-the-funder/) alongside the five DHCW-internal interventions.

Systems-dynamics honesty demands this kind of explicit scope statement. Analysis that does not name its boundary cannot be held to it.


## Applying This To DHCW
URL: https://bluenhs.org/methodology/applying-to-dhcw/

_How the raw material — Senedd proceedings, Audit Wales reports, Employment Tribunal filings, FOI refusals — becomes eleven loops and seven traps through a single analytical move._

## The Analytical Move

The raw material — Senedd proceedings, Audit Wales reports, DHCW board papers, Employment Tribunal filings, witness testimony, FOI refusals, the record at carenhs.org — becomes eleven loops and seven traps through a single analytical move: **group every observation by the stock it affects and the feedback it produces**.

Take a concrete example. A Senedd committee hears that health boards are building their own systems rather than using DHCW's. That observation affects a stock (health-board trust in DHCW — declining). It also affects a flow (adoption rate — falling). The declining trust stock feeds back into DHCW's credibility, which further reduces adoption. When you trace this chain far enough, it closes into a circle: a reinforcing feedback loop. That loop is [L2 — Credibility Death Spiral](/diagnosis/l2-credibility-death-spiral/).

Loops emerge when observations form a closed cycle. Traps emerge when multiple loops interact in Meadows' [recognised patterns](/methodology/system-traps/). This is not interpretation — it is classification. The discipline provides the taxonomy; the evidence provides the instances.

## The Evidence Aggregation

The discipline gives the taxonomy. What turns the taxonomy into a diagnosis is the evidence aggregation. Five years of DHCW board and committee meeting transcripts — 61 meetings in total — were processed into a knowledge graph of **1,779 nodes and 3,427 edges**, supplemented by Welsh Government performance-escalation correspondence, Audit Wales structured assessments, FOI disclosures, and the ABUHB dossier (123 sources). Each node is an entity, a claim, a finding, or an evidence point. Each edge is a typed relationship between them.

Some edge types carry quantitative weight:

- **51 edges of type "approved without scrutiny"** — board approvals that the published record shows received no substantive discussion before being passed.
- **45 edges of type "undeclared interest at meeting"** — instances where a director participated in a discussion in which they had an undisclosed interest.
- **33 edges of type "declared nil despite"** — instances where a director declared "nil" while holding a known undisclosed Professor of Practice title (Thomas 10, Evans 10, Hurle 8, Hall 5).
- **63 edges of type "failed to act on"** — instances where a flagged risk was logged and no corrective action followed.
- **107 sanitisation findings** plus **237 hiding-intent passages** — passages where what was spoken in the room was substantively altered, softened, or deleted before publication.

These counts are the bridge between the discipline (which predicts that captured governance produces certain edge-types) and the evidence (which shows the predicted edge-types appearing at scale). They are not the only evidence — much of the diagnosis rests on qualitative passages, named quotes, and structural observations — but they are the part that quantifies what the qualitative narrative describes.

## Two Clusters, One Architecture

The eleven loops fall into two clusters with fundamentally different characters.

**Cluster A — five reinforcing loops of delivery failure.** [Hiring without delivery](/diagnosis/l1-hiring-trap/), [credibility death spiral](/diagnosis/l2-credibility-death-spiral/), [annual-funding paralysis](/diagnosis/l3-funding-uncertainty-trap/), [the rebranding escape](/diagnosis/l4-rebranding-escape/), [vendor-dependency spiral](/diagnosis/l5-vendor-dependency-spiral/). These are structural. Any national health IT body operating under similar constraints — monopoly mandate, annual funding cycle, sovereign health-board customers — would face some version of these dynamics. They are the starting point of the diagnosis, not the whole story.

**Cluster B — six reinforcing loops of self-preservation.** [Manufactured narrative](/diagnosis/l6-manufactured-narrative/), [competence void](/diagnosis/l7-competence-void/), [loyalty selection](/diagnosis/l8-loyalty-selection/), [whistleblower suppression](/diagnosis/l9-whistleblower-suppression/), [information fortress](/diagnosis/l10-information-fortress/), [oversight obstruction](/diagnosis/l11-oversight-obstruction/). Cluster B is not structural accident. It is an active engine that intercepts every corrective mechanism — board scrutiny, ministerial oversight, staff feedback, external audit — before it can reach Cluster A. This is why repeated reform attempts have failed: the self-preservation engine absorbs them.

The [blueprint](/blueprint/) must address both clusters. Fixing delivery structure without removing the self-preservation engine means the engine will capture the new structure. Removing leadership without fixing the structure means new leadership will face the same dynamics.

## How to Read the Diagnosis

Each **loop page** has four sections: the pattern (the generic systems-dynamics mechanism), how it manifests at DHCW (specific evidence), the healthy alternative (what a functioning version looks like), and the escape route (which intervention in the [blueprint](/blueprint/) breaks the loop). Every loop identifies a specific stock being depleted, the feedback mechanism that depletes it, and the characteristic delay that makes the damage invisible until it is severe.

Each **trap page** has four sections: the archetype (Meadows' definition), the DHCW manifestation, which loops produce it, and the intervention that breaks it. Traps are how the [blueprint](/blueprint/) sequences its interventions — the trap structure dictates which changes must come first.

The 18 stocks tracked across the analysis are organised by visibility on the [stocks page](/diagnosis/stocks/): eight that are visible and measured, six that are visible but unmeasured, and four that are invisible but influential. The gap between what governance measures and what actually determines outcomes is itself a diagnostic finding.

## What the Methodology Predicted

Systems-dynamics analyses make a strong claim: when the structural conditions are present at the start, the failure patterns will be visible from the start. This is testable. The deep evidence base tests it.

Every governance deficit pattern observed at the point of Level 3 escalation thirty-four months in was already operational at DHCW's very first board meeting. Pre-DHCW UWTSD Professor of Practice titles awarded to three executive directors in December 2020 — four months before the founding board — were never declared at any subsequent board meeting through to escalation. The first meeting in April 2021 documents a director declaring "nil" while ten months into holding that title. The architecture was not gradually corrupted by stress. It was complete on day one.

This is the methodological pay-off. The discipline predicts that systems shaped by certain feedback loops and protected by certain self-preservation mechanisms will fail in characteristic ways. The evidence shows the predicted failure unfolding exactly as the discipline forecasts — and visible from inception, not in hindsight.


# About
URL: https://bluenhs.org/about/

_Who built this blueprint, why it exists, and why the author — DHCW's former Chief Technology Officer — left to build the analysis no internal review could._

This blueprint was not commissioned. It was not funded. It was built by someone who joined DHCW to help, discovered what was underneath the surface, and left to build the structural analysis that no internal review could produce.


## About the Author
URL: https://bluenhs.org/about/about-the-author/

_Who I am, what authority I speak from, and why I believed I could help DHCW before I discovered why it could not be helped from inside._

I am the kind of technology leader this blueprint identifies as missing from DHCW's executive team. I know this because I was there — serving as DHCW's Chief Technology Officer before leaving to build the structural analysis no internal review could produce.

## The profile

I am [Dr Rafal Bergman](https://ctozen.com) — PhD Computer Science, PMP, Oxford Saïd Executive Leadership Programme. I have spent twenty-five years building and delivering complex technology systems at scale. I have founded four technology companies and exited all four. I have led engineering at IBM. I have delivered national-scale systems for governments, defence ministries, and Fortune 500 enterprises across more than thirty countries.

The specifics matter, because the diagnosis on this site rests on them.

**SRT Marine Systems (Chief Technology Officer, hands-on).** I invented, designed, and built the first version of a national-scale real-time analytics and predictive-intelligence platform, then led a team of fifty-plus engineers to productionise it. The platform deployed across more than thirty countries — including national-scale programmes in the Kingdoms of Bahrain and Kuwait, Canada, Korea, Norway, Australia, Ecuador, and Azerbaijan — and grossed over four hundred million pounds in sales. The underlying architecture is patented in the United States, the United Kingdom, and the European Union.

**Improbable (Software Engineering Manager, Defence).** I built and led high-performance computing teams delivering platforms for the Ministry of Defence, the Royal Navy, and NATO — including the foundational technology platform for NATO's £10B Collective Training Transformation Programme.

**Reed Smith LLP (High Court Expert Witness).** In a criminal case in the High Court of Hong Kong involving thirty-nine casualties, I built bespoke software to analyse raw radar and AIS data, reconstructed events that a two-year Commission of Inquiry had missed, gave six days of testimony under cross-examination, and helped vindicate an innocent captain who would otherwise have served a substantial jail sentence. This was forensic reconstruction against an institutional narrative that had already hardened.

**IBM (Software R&D Manager, Warsaw).** I architected and delivered the Digital Public Services Platform (ePUAP) for the Polish Government, alongside large-scale systems for ING Bank and other financial institutions.

**AdBrain AI (Head of AI Engineering, Co-Founder).** I co-founded AdBrain and built an agentic AI platform that resolved sixty-seven per cent of customer cases without human intervention at one of Europe's largest insurance brokerages.

**GeoVS Limited (Founder, exit).** I founded GeoVS to commercialise my PhD research into a global data, logistics, and security platform used by BP, Statoil, the Port of Milford Haven, and the Panama Canal. I exited via sale to SRT Marine Systems.

I hold three patents, have published research in real-time data systems and AI for logistics, and have spent twenty-five years in hands-on engineering and engineering leadership. I name this not to impress but to anchor what follows.

> This is the background the diagnosis on this site says DHCW's executive needed and did not have.

## Why I joined DHCW

I joined DHCW as Chief Technology Officer in February 2024. I understood the scale of the challenge before I accepted. DHCW sits at the heart of digital delivery across NHS Wales. Every patient referral, every clinical system, every piece of health-data infrastructure flows through it. That is 3.16 million people. Over 611,000 of them are currently on waiting lists. The organisation was struggling, and I knew from experience that technology transformation in complex public institutions is among the hardest work there is.

I took a significant pay cut to accept the role. I closed my consulting practice. I did this because I wanted to give the NHS mission my full commitment, not a fractional one.

> This was not a career move. It was a decision to serve.

Alongside my role at DHCW, I served as Technical Lead of the Welsh Government Commission for AI in Health and Social Care, where I authored the national AI adoption and technical innovation strategy for the Welsh NHS.

## What I built in fifteen months

Within my first two months I discovered and stopped wastage of over one million pounds. Within six months I had started programmes that would save a further five million and fix some of the most broken and dangerous parts of the digital portfolio — including the systems responsible for all integrations and the flow of patient referrals across Wales. I proposed changes to re-procurement and buy-versus-build decisions that would stop wastage of twenty to thirty million pounds per year. I transformed a 300-strong technical organisation responsible for the strategic portfolio of national products underpinning healthcare delivery in Wales. I established a software-engineering culture grounded in TOGAF, agile, domain-driven design, test-driven development, CI/CD, infrastructure-as-code, platform engineering, observability, FinOps, and DevSecOps. I proposed new principles for product management to the Technical Decisions Authority.

What I encountered nearly broke me. Crippling incompetence and institutional resistance at every step, making change virtually impossible.

Almost none of my reforms survived after my departure.

> Every programme I built to save money, fix dangerous systems, or improve delivery capability was quietly deprioritised, defunded, or handed to people who lacked the skills to continue it. This was not neglect. It was strategy.

The next page explains why.


## Why This Blueprint
URL: https://bluenhs.org/about/why-this-blueprint/

_What the author discovered inside DHCW, why no internal review could surface it, and why this analysis was built pro bono using the methodology of a forensic investigation._

When I joined DHCW as the Chief DevOps Officer (at the top of Band 9), it was not a career move. It was a decision to serve. But NHS Wales and DHCW are organised to resist change, and focus on many things — stated mission and delivery not being amongst them.

## What I saw

Directors constantly commissioned expensive analyses from consulting firms. When the findings pointed to structural problems rather than simple fixes, leadership rejected them. Reports were buried. Anything that questioned competence was suppressed.

The longer I stayed, the clearer the pattern became. Many of the systemic problems were not merely tolerated.

> **They were maintained on purpose.** The goal was never catching the rabbit. It was chasing it. Actual delivery would have exposed the lack of capability to deliver.

The systemic problems served as an excuse for poor delivery and enabled a perpetual performance theatre of new initiatives, re-starts, new strategies, and new programmes.

I left because I could not fix this from the inside. What I have learned since leaving is worse. Every improvement initiative I had built — programmes to save money, fix dangerous systems, or improve delivery capability — was actively undermined: quietly deprioritised, defunded, or handed to people who lacked the skills to continue it. That was not neglect. It was strategy.

## What no internal review could surface

DHCW is not an organisation that is failing. It is a system configured not to succeed. Six structural findings, each evidenced across five years of board and committee minutes, FOI disclosures, and the ABHB dossier that documents how this leadership group operated before DHCW even existed.

### 1. DHCW is a thinly-repainted NWIS

NWIS — the NHS Wales Informatics Service — was a nationally discredited organisation. The Wales Audit Office report on its operation left no room for doubt that the body was dysfunctional, with broken governance and a culture the Senedd later described as "the antithesis of open." The political response was to wind NWIS up.

What followed was a rebrand, not a reform. DHCW was stood up to replace NWIS, adopting a new name, a new logo, a new statutory footing — and substantially the same leadership team. Helen Thomas, who had held senior executive roles in NWIS through its decline, became DHCW's Chief Executive. Rhidian Hurle, drawn from the same cohort, became DHCW's Medical Director. The patronage pipeline that had staffed NWIS was extended into its successor body.

The diagnostic significance is structural, not biographical: every reform mechanism that had failed against NWIS — internal challenge, audit, oversight, programme intervention — was facing the same people, in the same roles, with the same incentives, the day DHCW was founded. The label changed. The system did not.

### 2. No technical leadership

DHCW is a national-scale technology delivery body. None of its executive directors has any real understanding of technology, or any experience leading a technical delivery organisation. They are NHS lifers — they have never seen what good looks like, never worked inside an organisation that ships software at scale, never recruited technical staff against external criteria, never managed an engineering function, never been accountable to delivery rather than to process.

What they know is NHS bureaucracy and the hypocrisies that go with it — how to write a strategy paper, sit on a committee, prepare a board report, attend a Senedd hearing. They do not know how to deliver. So what they produce is performance theatre — new strategies, rebrands, recruitment drives, presentations, pledges, pilot programmes, awards — every artefact of an organisation that is busy, none of the substance of an organisation that delivers.

### 3. The opposite of compassion

Posts were left deliberately unfilled to meet annual financial targets — year after year. The Chair warned about it from year two: "you just heap misery on misery every year when you do that." The pattern continued through years three, four, and five.

The consequences were predictable and measurable. By 2024, the staff survey recorded a 65% burnout rate — a finding buried from the public record. In a single year of my time at DHCW, the organisation lost four senior technical leaders. Annual-leave buyback schemes were introduced because remaining staff could not take the leave they were owed. Sickness rose, driven specifically by stress and depression.

While this was happening, the board formally adopted a "Compassionate Leadership Pledge", approved in under fifteen seconds with no questions. The Pledge was a label, not a policy. The behaviour underneath it produced the burnout, the resignations, and ultimately the delivery failure that escalated DHCW to Level 4.

### 4. Empire building over delivery

Decisions that should have been driven by delivery urgency were driven by control. For three months I watched the CEO block the recruitment of leadership for DHCW's critical integration-services function. Not because the roles were unnecessary; not because budget was unavailable. The pause existed so that her protégé, the Chief Data Officer, could absorb the team into his own portfolio without a parallel leader being installed first.

The three-month delay cost an estimated £1M inside DHCW alone, and a multiple of that across the health boards that were waiting for the integration work the team would otherwise have been doing.

When I raised this through the appropriate channels — that critical delivery was being held up to engineer an internal empire — the executive response was to label me "unprofessional". The integration backlog was real. The harm to health boards was real. The label was the answer.

### 5. Whistleblower suppression

The mechanisms that should have surfaced this failure from inside the organisation were closed one by one. People who raised concerns about technical capability, programme realism, or the integrity of the published record were managed out. Pretextual charges were brought. Disciplinary processes followed paths the organisation's own policies did not permit. Roles were replaced with downgraded versions on lower bands and with less authority — the position survived, the oversight function did not.

In 2018 the Senedd found the predecessor body's culture was "the antithesis of open." Eight years later, with documented retaliation against technical dissent and a downgraded successor post in place, the same culture is operating. By 2026, DHCW publishes zero whistleblowing statistics, zero disciplinary data, and zero staff-leavers analysis.

### 6. The manufactured narrative

A counter-narrative is produced to fill the silence. DHCW simultaneously reports 80% staff satisfaction and a 65% burnout rate in the same staff survey, while operating under Level 4 Targeted Intervention. All three are factual statements about the same organisation in the same year, reconcilable only if one accepts the information-control mechanism producing the first number.

The pattern was not occasional. When a stakeholder survey returned only 13% of respondents speaking highly of DHCW, the meeting that received the result was celebratory — pats on the back, congratulations, no acknowledgement that thirteen per cent positive external sentiment is not a finding to celebrate. The same style of reporting carried through to the Welsh Government's formal intervention.
### 7. The sanitisation machine

The published board record is not a record. Entire reports — CEO reports, finance reports, performance reports, Board Assurance Framework presentations, risk-register discussions — were deleted wholesale from published minutes. In one published transcript, sixteen of nineteen speakers were entirely absent. At the extreme, the curation ratio was 10.7 per cent: 3,680 published words from 34,257 actually spoken.

The sanitisation is not random. It systematically removes admissions of failure by executives, structural financial warnings, independent member challenges, and any context that would reveal a pattern of deterioration. It produces a curated public record designed to appear functional while concealing systemic governance failure.

The Chair's own warning that the vacancy-savings strategy "heaps misery on misery every year" was stripped from the published minutes. The Director of Strategy's admission that "milestone owners said yes on track when they were not" was sanitised. The Chief Operating Officer's candour that technical debt is "an anodyne phrase" was deleted. The pattern is consistent across five years.

### 8. Predictable and predicted

Almost every element of the March 2025 escalation to Targeted Intervention was visible years earlier, named by people inside the room, and ignored.

The vacancy-savings death spiral was warned about from 2022 — "you just heap misery on misery every year when you do that." Funding instability was flagged from founding. Programme overload was identified at month seven. Technical debt was named as a structural problem in 2023. Cross-programme resource risk was raised repeatedly by the sharpest governance voice on the board, who departed in March 2026 having identified "the lack of governance and assurance mechanisms across the organisational gaps" as the biggest impediment to progress. Optimism bias — "people for understandable behavioural reasons are optimistic, the style is we're gonna make it, we're gonna make it, no we're not" — was admitted by the Director of Strategy himself.

> All governance deficit patterns observed at escalation thirty-four months later were already operational at meeting one.

### 9. The pipeline continues

The Director of Strategy who authored *A Healthier Wales*, was seconded into DHCW to implement it, then admitted forty-six months in that nobody knew who was accountable, is now designing the next ten-year digital strategy for NHS Wales. The same patronage pipeline that produced the current leadership remains intact, untouched by either DHCW's Level 4 escalation or the change of Chair, Chief Financial Officer, and independent members in the months around it.

This is the definition of a captured system. The people who failed at one health board were elevated to oversee its successor body, where they presided over the same failures. The blueprint they did not deliver is being authored again by the same hand.

## Why pro bono

For the last year I have been working pro bono on the creation of this blueprint. I have done this as a volunteer, on my own time, because I joined DHCW to make a difference and I intend to make that difference whether the organisation cooperates or not.

The work applies the systems-dynamics framework developed at MIT and articulated for institutional analysis by Donella Meadows — stocks, flows, feedback loops, delays — to map exactly why DHCW fails, how the failure is structurally protected, and what would fix it. The [methodology](/methodology/) is published in full. The analysis specifies the conditions under which its core thesis would be falsified.

> This work is designed to be tested, not believed.

**This blueprint is offered freely for political, NHS, and public use, with attribution.** It belongs to the people of Wales, not to the leadership that created the problems it describes.