Failure demand and the NHS: why the productivity conversation starts with fixing the system, not the staff

The Centre for Workforce Intelligence and Planning | June 2026

When Roy Lilley, one of the NHS's most widely read commentators, wrote this week about his experience as a day-case patient at Whipps Cross Hospital, he was not intending to write an academic paper on service design theory. But what he described — arriving for a procedure, not being on the printed list despite a confirmed booking, a nurse spending two hours navigating system failures just to get him registered, duplicate consent forms describing the wrong procedure entirely — is, in the language of systems thinking, a textbook illustration of failure demand (Lilley, 2026).

The nurse herself provided the most revealing data point. When Lilley suggested he would not bother to complain, she replied: "if you don't, things like this will keep happening. They don't do anything." And when she reported back from the manager: "in a meeting and too busy to come to the clinic." In a functioning system, those sentences would be unthinkable. In the NHS today, they describe something that has become, in too many places, unremarkably normal. Failure activity has been absorbed into the structure of daily working life and, in the process, has become structurally invisible.

Understanding what this means — for productivity, for workforce capacity, and for the technology agenda now driving NHS reform — is, we think, one of the most important conversations the health system needs to have.

What is failure demand?

The concept was articulated by John Seddon, management thinker and founder of Vanguard Consulting, who has spent four decades applying systems thinking to service organisations (Seddon, 2008). Seddon draws a sharp distinction between two fundamentally different types of demand on any service.

Value demand is what the service exists to provide. A patient needs a procedure; the NHS performs it. A citizen needs to file a tax return; HMRC processes it. This is the work the system was designed for, and it is what any service's resources should, in principle, be entirely directed towards.

Failure demand is demand caused by a failure to do something, or to do something right, the first time (Seddon, 2008). The patient who rings to chase a letter that was never sent. The clinician who re-enters data already held in another system that doesn't communicate with theirs. The appointment booked twice because two booking systems aren't integrated. The manager who spends an afternoon managing a complaint that a joined-up process would have prevented. None of this activity delivers care. It absorbs the capacity needed to deliver it.

It is important to understand that failure demand is not, primarily, about individual mistakes. It is a systems property — an emergent characteristic of how services are designed and how processes are connected (or not). Seddon argues that it arises almost inevitably from what he calls the "command and control" design of most public services: functional specialisation, hierarchical gatekeeping, performance management that measures activity rather than outcomes, and a fundamental failure to understand demand from the perspective of the person being served (Seddon, 2008). When a system is not designed around what people actually need, it generates work for itself in the process of failing to meet those needs.

The scale of this in the NHS is substantial. The Strategy Unit — working in partnership with the Health Foundation as part of the NHS Productivity Commission — is currently engaged in a major project to define and measure failure demand in NHS settings (Strategy Unit, 2025). Preliminary outputs suggest that failure demand encompasses repeat appointments generated by inadequate first contacts, delayed diagnoses that lead to more complex and costly interventions, and waits for treatment that are so long they reduce its effectiveness and generate further downstream demand (Health Foundation, 2025a). This work is expected to feed into the Health Foundation's broader policy recommendations, due in September 2026, on how to drive sustainable NHS productivity growth over the next decade (Health Foundation, 2025b).

The implication for how the NHS thinks about its productivity challenge is significant. If a substantial proportion of current NHS activity is failure activity — work created by the system's own processes rather than by genuine clinical need — then adding capacity, increasing targets, and intensifying performance management will not resolve it. It will, in all probability, intensify it.

This is not about rocket science

Before going further, it is worth being direct about something that is too rarely said plainly in the NHS technology debate. The public conversation around NHS technology reform sometimes creates the impression that the gap between where the NHS is and where it needs to be is a gap of sophisticated, cutting-edge innovation — that solving the problem requires artificial intelligence of the kind that lives in the realm of advanced data science and academic research. This misunderstands both the problem and the solution.

For the vast majority of the failure demand described above, the technology needed to address it already exists. It is not experimental. It is not on the horizon. It is available today, and in many cases the NHS has already paid for it. What is needed is not invention. It is adoption — and, more specifically, the consistent, system-wide deployment of technology that enables processes to be joined up, tasks to be automated, and different parts of the system to work together as if they were designed with the patient in mind.

The electronic patient record (EPR) is the clearest example. Ninety percent of NHS trusts now have an EPR system in place (NHS England, 2024). Yet the Health Foundation's analysis finds that many are not using them to meaningfully improve care quality, staff experience, or productivity (Health Foundation, 2024). The reason is not that the technology doesn't work. It is that EPR systems across the NHS frequently do not communicate with one another, leaving clinicians to re-enter information already held elsewhere, chase records from other providers, and work around gaps in their digital picture of the patient. A recent cross-sectional survey found that poor EPR interoperability negatively impacted work productivity, created additional communication barriers between clinical teams, increased clinician burnout, and in some cases compromised patient safety (Payne et al., 2025). Meanwhile, NHS trusts in England are expected to spend more than £13.5 million in 2026 alone correcting data errors that arise after EPR go-lives — the costs of systems that have been implemented without the integration needed to prevent failure demand at source (Digital Health, 2026).

This pattern — investment in technology that then generates new failure demand rather than eliminating existing failure demand — is a consequence not of the technology but of the approach: implementing systems without first understanding the demand they are meant to resolve, or without connecting them to the other systems around them. A 2025 report from NHS England warned that the NHS must "move EPR use beyond digitising paper" — that the current approach in too many trusts has been to replace paper-based processes with digital equivalents, rather than to redesign those processes around what patients actually need (Digital Health, 2025).

The same logic applies across the system: appointment booking, referral management, discharge coordination, medication reconciliation, patient communication, consent processes. In each of these areas, the technology to automate, integrate, and simplify is available. The NHS's challenge — as the new Health Secretary, James Murray, made clear in his keynote address at NHS ConfedExpo 2026 — is to make full use of what already exists, not to wait for the next wave of innovation (Murray, 2026). Murray's framing was direct: "the NHS cannot simply manage its way out of current pressures — it must transform its way out of them. The choice is not between reform and recovery; the task is to use reform to accelerate recovery" (Murray, 2026).

In this context, the conversation about "AI in the NHS" needs to be understood in more grounded terms than the discourse sometimes allows. Much of what is most immediately impactful is not AI in the sense of autonomous machine intelligence. It is workflow automation: systems that route referrals correctly without human intervention; ambient voice tools that complete clinical documentation while the clinician talks to the patient; scheduling algorithms that optimise appointment booking across available capacity; interoperability standards that allow a GP's system to share information with an acute trust's EPR. These are not science fiction. They are 21st-century tools that most modern service organisations already use as a matter of course. The NHS's task is to make them a natural part of how it runs — every day, at every level, not only in the organisations that have strong digital leadership.

What other sectors tell us

The evidence from sectors that have grappled seriously with failure demand is instructive — both for what it reveals about the scale of the problem and for what it demonstrates about what resolving it requires.

Local government and adult social care

Vanguard Consulting's application of Seddon's systems thinking to UK local authority services has produced some of the most extensively documented results. In a published adult social care case study, a full system redesign found that 74 percent of demand under the original design was failure demand (Vanguard Consulting, n.d.). After redesign — with processes rebuilt around understanding and meeting need at first contact — failure demand fell to under 10 percent of total demand. The average administrative cost per case fell from £998 to £134, a reduction of 87 percent (Vanguard Consulting, n.d.). This did not require advanced technology. It required a rigorous analysis of what people actually needed and why the existing system was failing to provide it, followed by a redesign that eliminated the structural sources of failure demand.

Studies of police services have found similar patterns. Research in Gloucestershire Constabulary found that 32 percent of incoming incidents represented failure demand — repeat reports of unresolved problems, calls to chase officers who had not attended, contacts made because crime victims had not been given direct contact details for their assigned officer (Seddon, 2008). Structural redesign, rather than technology investment, resolved the majority of this.

Utilities and telecommunications

In UK telecoms and utilities, failure demand in customer contact has been estimated at between 40 and 60 percent of total contact volume (Seddon, 2008). At Westward, a utilities provider, a two-week redesign of how incoming demand was received and handled reduced call transfer rates from 34 percent to 14 percent — with no new technology, simply a different organisational approach to resolving demand at the point of first contact (Call Centre Helper, n.d.).

HMRC

The tax authority provides perhaps the most instructive public-sector comparator — not least because it was cited by James Murray at NHS ConfedExpo as a model for NHS digital transformation (Murray, 2026). HMRC's own analysis found that 72 percent of calls to its telephone helplines were the result of failure demand — contact caused by process failures, delays, or confusion, rather than a genuine need for substantive advice (NAO, 2024). The subsequent digital transformation programme, which Murray referenced as successfully freeing resources for more complex cases, has produced mixed results (ICAEW, 2026). Where digital self-service processes have successfully met need at first contact, contact volumes have fallen. Where they have added a layer of digital complexity without resolving the underlying need — where the digital process itself becomes a source of confusion or error — they have generated new failure demand of their own (NAO, 2024).

The lesson from HMRC is important: digitisation, in isolation, does not reduce failure demand. What reduces failure demand is understanding what people need, designing processes that reliably meet that need, and then using technology to make those processes accessible, automatic, and connected. Technology is the enabler. System design is the solution.

Why it is harder in the NHS

If the concept is clear and the evidence from other sectors is substantial, why is failure demand so persistent in NHS settings? A 2017 analysis found that failure demand reduction had achieved "limited success in the public sector" due to the difficulty managers face in enacting systemic change based on their analyses of it (Radnor, 2017). Several barriers are worth examining in detail, with particular attention to the workforce dimension.

Failure demand and the erosion of professional purpose

Before examining the practical barriers to change, it is worth dwelling on what failure demand actually does to the people who carry it — because this is, in our view, the most underappreciated dimension of the problem.

Lord Darzi's independent investigation into the NHS, published in September 2024, captured it with unusual directness. Writing about the consequences of falling productivity, he stated: "it needs to be stressed that falling productivity doesn't reduce the workload for staff. Rather, it crushes their enjoyment of work. Instead of putting their time and talents into achieving better outcomes, clinicians' efforts are wasted on solving process problems, such as ringing around wards desperately trying to find available beds" (Darzi, 2024). This is not a peripheral observation. It is a description of what failure demand feels like from the inside — and it has profound implications for workforce engagement, retention, and wellbeing.

People do not enter clinical and caring professions to manage the consequences of broken systems. They enter them to help people. When a meaningful proportion of every working day is consumed by failure activity — chasing information that should be immediately available, repeating tasks that should only need doing once, managing patients through processes that generate confusion rather than care — the gap between the work that people trained for and the work they are actually doing grows. That gap is not neutral. It is one of the most significant drivers of disengagement, burnout, and the quiet departure of experienced staff that the NHS can least afford to lose.

Darzi's report documented the cumulative toll: "distressingly high levels of sickness absence — as much as one working month a year for each nurse and each midwife working in the NHS" and a "marked reduction in discretionary effort across all staff groups" in the pandemic's aftermath (Darzi, 2024). The NHS Staff Survey data confirms this picture, with staff engagement and satisfaction scores tracking downward for several consecutive years (NHS Employers, 2024). It is worth asking how much of what registers as staff shortage exhaustion is, in fact, the accumulated weight of failure demand — years of being required to battle a system that was not designed around the people using it, neither patients nor staff. As Darzi put it in a different passage: the ambition should be to be "smarter, not just faster," in a way that is "better for patients' outcomes and experiences and for staff and their enjoyment of work" (Darzi, 2024).

This is why the failure demand conversation and the workforce engagement conversation cannot be held in separate rooms. Reducing failure demand — through better-connected technology, redesigned processes, and a system that starts with the patient and works backwards — is not only a productivity intervention. It is a workforce retention and wellbeing intervention. It is about restoring the conditions in which staff can do the work they actually came to do. If the NHS is serious about the engagement crisis that Darzi documented, addressing the structural sources of failure demand is not optional. It is foundational.

Operational pressure leaves no headroom

The most immediate barrier is that relentless operational pressure leaves NHS organisations with almost no capacity to step back from failure activity and redesign the processes generating it. When a trust is running at 99 percent bed occupancy, managing a waiting list in the tens of thousands, and the Service fielding 90,000 corridor-care cases in a single month (HSJ, 2026), the prospect of a service redesign programme — however well evidenced — is experienced not as an opportunity but as yet another demand on a depleted system. The NHS Providers Digital Transformation Survey (2024) found that trust leaders and staff had "limited headroom for undertaking large-scale digital transformation programmes" as a direct consequence of operational pressure (NHS Providers, 2024). This is not resistance to change. It is rational prioritisation under extreme constraint — and it is, itself, a consequence of the failure demand the system has not yet addressed.

The digital skills gap is real and uneven

A second barrier is the uneven distribution of digital capability across the NHS workforce. The Corndel NHS Digital Skills Report (2025), drawing on research with NHS decision-makers, found significant gaps in the capacity to adopt and use new technology effectively (Corndel, 2025). This is not primarily a matter of individual willingness. The NHS employs 1.5 million people across dozens of professions, geographies, and organisational settings, with highly variable access to training, inconsistent approaches to upskilling, and a national digital workforce strategy that has not yet translated into consistent local provision (Corndel, 2025). The OECD's 2025 analysis of digital and AI skills in health occupations found that demand for these skills in health settings is growing significantly faster than the workforce's current capacity to develop them (OECD, 2025).

Critically, this skills gap is not only a technical matter. It is also a question of confidence and trust. Research on health information technology adoption by nurses found that clinicians become risk averse and resistant when they suspect new technology will negatively affect their relationship with patients, or when implementation is experienced as difficult without adequate support (Harte et al., 2021). In a workforce already under pressure from moral injury, rising harassment, and concerns about deskilling (Nursing Times, 2026a), digital transformation can feel like one more imposition rather than a route to relief.

Professional structures and organisational fragmentation

The NHS's organisational complexity creates a third set of structural barriers. With more than 40 recognised clinical specialties, a highly fragmented architecture of trusts, ICBs, primary care networks, community providers, and local authorities, and professional hierarchies that can impede cross-system working, diffusing both technology and changes to service design is extraordinarily difficult (Global Counsel, 2024). Research on technology adoption in the NHS consistently identifies "structural complexity and professional rivalries" as significant inhibitors (NCBI, 2015). Where a utilities company can redesign a contact centre process in a fortnight, the equivalent in an NHS setting involves clinical governance, patient safety frameworks, job evaluation implications, commissioner sign-off, union consultation — frequently involving a leadership team that has changed between the decision and the implementation.

The risk of technology as new failure demand

A fourth barrier, perhaps the most practically important, is that technology deployed without careful system design can generate new failure demand rather than eliminating existing failure demand. The NHS's EPR programme illustrates this risk: significant investment in systems that, in many trusts, have not yet delivered the interoperability and process improvement they were procured to achieve, and which in some cases have added administrative burden during and after implementation (Orion Health, 2024; Digital Health, 2026). The HMRC experience demonstrates the same dynamic at scale. The risk is not technology per se; it is technology deployed without first understanding the demand it is meant to resolve. Seddon's discipline — studying the system before intervening in it — is as relevant to the NHS's digital programme as it is to any other reform agenda.

The deskilling concern

A concern that deserves serious engagement, rather than dismissal, is the risk that increasing automation of clinical and administrative tasks will erode the skills and professional judgment that staff have developed through doing those tasks. Nursing Times' analysis in 2026 warned that AI adoption and workforce reforms put nurses at risk of deskilling — that as tasks are automated, the clinical reasoning and interpersonal capabilities honed through practice may atrophy over time (Nursing Times, 2026a). This is not an argument against technology adoption; it is an argument for ensuring that the design of technology deployment actively supports professional development rather than inadvertently substituting for it.

Entrenched failure activity as the new normal

Perhaps the most insidious challenge is the one Lilley's nurse captured so precisely: "it often happens." When failure demand becomes so routine that it is no longer recognised as failure — when the two-hour registration process, the duplicate consent form, the missed booking, the repeated data entry, have become simply "how things are" — it has achieved a kind of structural invisibility. Organisations adapt around it. Roles are designed to manage it. Whole informal systems of workarounds develop. Cultures form that treat failure activity as an inevitable feature of the environment rather than a correctable design flaw. Seddon (2008) identifies this as the most significant obstacle to improvement: not the failure demand itself, but the management systems and professional cultures that normalise it, and the leadership structures that fail to surface it.

What good looks like — and what it requires

The evidence from sectors that have successfully reduced failure demand points consistently to a set of enabling conditions. None of them is straightforwardly easy. All of them are within reach.

Start with understanding demand, not deploying solutions. In every case where failure demand has been substantially reduced — in adult social care, in utilities, in policing — the starting point has been a rigorous, honest analysis of what people actually need, why they are making contact, and what the system is and is not doing to meet that need. In the NHS, this means moving beyond activity metrics to ask a harder question: of all the activity we are generating, what proportion was generated by the system's own failure to get something right the first time? The Strategy Unit and Health Foundation's current measurement work is foundational in this respect (Strategy Unit, 2025), and its outputs — due in 2026 — deserve to be taken seriously at every level of system leadership.

Engage clinical staff as designers, not recipients. Research is consistent that successful technology adoption in clinical settings requires genuine involvement of frontline staff in the design of new processes, adequate and sustained training, and leadership that understands the difference between mandating change and enabling it (NCBI, 2015; NHS Providers, 2024). The NHS Providers survey found that "recruiting and retaining digital skillsets and developing digital capabilities within the workforce" were identified as critical, with respondents calling for "additional, consistent capacity to make further progress" (NHS Providers, 2024). Digital capability is not a one-off training event; it is a workforce development agenda.

Create cultures in which failure can be seen and said. Lilley's nurse knew that "if you don't [raise it], things like this will keep happening." The manager was unavailable. That dynamic — frontline staff seeing failure activity clearly, and leadership systems that do not create the conditions for that insight to reach decision-makers — is among the most significant workforce dimensions of the failure demand problem. Organisations that have successfully reduced failure demand have built cultures in which the people closest to the work are the people whose analysis shapes the design of the system. This is not a technology challenge. It is a leadership and organisational development challenge.

Prioritise consistent adoption over endless piloting. The NHS has no shortage of outstanding proof-of-concept examples: the eTriage tool at Royal Berkshire reducing unnecessary face-to-face outpatient appointments; the community wellness team in South Cumbria cutting emergency admissions by two-thirds; the growing use of ambient voice documentation to eliminate one of the most significant sources of clinical time waste (Murray, 2026). The challenge, as Murray correctly identified, is not demonstrating that these things work. It is ensuring that they are adopted consistently across the system — not only in organisations with strong digital leadership, but in the ordinary, everyday trusts and community services where failure demand continues to accumulate unremarked.

Conclusion

The failure demand conversation is, in its deepest sense, a workforce conversation. Every minute a nurse spends on a two-hour registration odyssey for a patient who should have been on the list is a minute not spent on the procedure that patient came for. Every hour a clinician spends chasing a referral, re-entering data, or managing a complaint that an integrated system would have prevented is capacity consumed by the system's own failure to work, rather than by the genuine demand it exists to serve. These are not abstractions. They are the daily texture of working life for hundreds of thousands of NHS staff — and they are directly connected to the moral injury, burnout, and attrition that make the NHS's workforce challenge so acute.

The technology to address a significant proportion of this already exists. It does not require breakthroughs. It requires the NHS to use what it has — and has spent billions acquiring — in the way it was intended: to join things up, automate what should be automated, and build a system that starts with the patient and works backwards. Seddon's observation, made nearly two decades ago, remains the right frame: "little things are the big things; doing them twice costs twice as much" (Seddon, 2008).

The question we would like the sector to sit with is this: how do we ensure that technology — including AI — is deployed consistently and appropriately across the NHS to tackle the failure activity that has become entrenched? Not only in the trusts that already have strong digital leadership. Not only in the pilots and vanguards. But across the board, in the ordinary settings where failure demand quietly consumes the capacity of a workforce that is running out of room to absorb it. That is the 21st-century standard the NHS must reach — and reaching it starts not with technology, but with the will to look honestly at what the system has been doing to itself, and to the people who carry it.

References

Call Centre Helper (n.d.) Failure demand: a technique to reduce cost and improve the customer experience. Available at: https://www.callcentrehelper.com/failure-demand-a-technique-to-reduce-cost-and-improve-the-customer-experience-91527.htm (Accessed: 16 June 2026).

Darzi, A. (2024) Independent investigation of the National Health Service in England. London: Department of Health and Social Care. Available at: https://www.gov.uk/government/publications/independent-investigation-of-the-nhs-in-england/independent-investigation-of-the-national-health-service-in-england-accessible-version (Accessed: 16 June 2026).

Corndel (2025) NHS Digital Skills Report 2025. London: Corndel. Available at: https://www.corndel.com/report/nhs-digital-skills-gap/ (Accessed: 16 June 2026).

Digital Health (2025) 'NHS must move EPR use beyond digitising paper, warns report', Digital Health, April. Available at: https://www.digitalhealth.net/2025/04/nhs-must-move-epr-use-beyond-digitising-paper-warns-report/ (Accessed: 16 June 2026).

Digital Health (2026) 'Repairing EPR data errors could cost NHS at least £13.5m in 2026', Digital Health, May. Available at: https://www.digitalhealth.net/2026/05/repairing-epr-data-errors-could-cost-nhs-at-least-13-5m-in-2026/ (Accessed: 16 June 2026).

Global Counsel (2024) Unleashing innovation in the NHS: barriers and opportunities for the adoption and uptake of healthcare technologies. London: Global Counsel. Available at: https://www.global-counsel.com/insights/report/unleashing-innovation-nhs-barriers-and-opportunities-adoption-and-uptake-healthcare (Accessed: 16 June 2026).

Harte, R. et al. (2020) 'Identification of factors influencing the adoption of health information technology by nurses who are digitally lagging: in-depth interview study', Journal of Medical Internet Research, 22(11), e18690. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7455866/ (Accessed: 16 June 2026).

Health Foundation (2024) Electronic patient record systems in England: what do NHS staff think? London: The Health Foundation. Available at: https://www.health.org.uk/reports-and-analysis/analysis/electronic-patient-record-systems-in-england-what-do-nhs-staff-think (Accessed: 16 June 2026).

Health Foundation (2025a) 'Failure demand: the hidden work holding back NHS productivity', Health Foundation Blog, June. Available at: https://www.health.org.uk/features-and-opinion/blogs/failure-demand-the-hidden-work-holding-back-nhs-productivity (Accessed: 16 June 2026).

Health Foundation (2025b) Analytical partnership with the Strategy Unit. London: The Health Foundation. Available at: https://www.health.org.uk/funding-and-partnerships/our-partnerships/analytical-partnership-with-the-strategy-unit (Accessed: 16 June 2026).

HSJ (2026) 'Corridor care cases hit 90k in May', Health Service Journal, June. Available at: https://www.hsj.co.uk/quality-and-performance/corridor-care-cases-hit-90k-in-may/8023676.article (Accessed: 16 June 2026).

ICAEW (2026) 'Why HMRC must change to succeed', ICAEW Tax Faculty, January. Available at: https://www.icaew.com/technical/tax/tax-faculty/taxline/articles/2026/why-hrmc-must-change-to-succeed (Accessed: 16 June 2026).

Lilley, R. (2026) 'Everything', nhsManagers.net eLetters, 15 June. Available at: https://myemail.constantcontact.com/Everything.html?soid=1102665899193&aid=rq8e7gc-hM8 (Accessed: 16 June 2026).

Murray, J. (2026) NHS reform, technology and the future of health and social care: speech at NHS ConfedExpo 2026. Manchester, 11 June. London: Department of Health and Social Care. Available at: https://www.gov.uk/government/speeches/health-and-social-care-secretary-speech-at-nhs-confedexpo-2026 (Accessed: 16 June 2026).

National Audit Office (NAO) (2024) HMRC customer service. London: National Audit Office. Available at: https://www.nao.org.uk/reports/hmrc-customer-service/ (Accessed: 16 June 2026).

NCBI (2015) Facilitating technology adoption in the NHS: negotiating the organisational and policy context — a qualitative study. Available at: https://www.ncbi.nlm.nih.gov/books/NBK259898/ (Accessed: 16 June 2026).

NHS England (2024) Digitising the frontline. London: NHS England. Available at: https://www.england.nhs.uk/digitaltechnology/digitising-the-frontline/ (Accessed: 16 June 2026).

NHS Employers (2024) NHS Staff Survey results 2024. London: NHS Employers. Available at: https://www.nhsemployers.org/articles/nhs-staff-survey-results-2024 (Accessed: 16 June 2026).

NHS Providers (2024) Digital transformation survey 2024: survey findings — barriers and opportunities. London: NHS Providers. Available at: https://nhsproviders.org/resources/digital-transformation-survey-2024/survey-findings-barriers-and-opportunities (Accessed: 16 June 2026).

Nursing Times (2026a) 'AI adoption and workforce reforms put nurses at risk of deskilling', Nursing Times, 4 June. Available at: https://www.nursingtimes.net/digital-and-technology/ai-adoption-and-workforce-reforms-put-nurses-at-risk-of-deskilling-04-06-2026/ (Accessed: 16 June 2026).

OECD (2025) Digital and AI skills in health occupations: what do we know about new demand? Paris: OECD Publishing. Available at: https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/05/digital-and-ai-skills-in-health-occupations_f428e5a9/5fbd42ab-en.pdf (Accessed: 16 June 2026).

Orion Health (2024) 'NHS EPR optimisation and the productivity reset', Orion Health Blog. Available at: https://orionhealth.com/us/blog/nhs-epr-optimisation-and-the-productivity-reset/ (Accessed: 16 June 2026).

Payne, T. et al. (2025) 'Physician experiences of electronic health record interoperability and its practical impact on care delivery in the English NHS: a cross-sectional survey study', BMC Health Services Research. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12161308/ (Accessed: 16 June 2026).

Radnor, Z. (2017) cited in Randell, R. et al. (2021) 'Implementing failure demand reduction as part of a demand management strategy', Public Money and Management, 42(3), pp. 187–195. Available at: https://www.tandfonline.com/doi/full/10.1080/09540962.2021.1978163 (Accessed: 16 June 2026).

Seddon, J. (2008) Systems thinking in the public sector: the failure of the reform regime… and a manifesto for a better way. Axminster: Triarchy Press.

Strategy Unit (2025) 'Making sense of failure demand in the NHS', The Strategy Unit. Available at: https://www.strategyunitwm.nhs.uk/news/making-sense-failure-demand-nhs (Accessed: 16 June 2026).

Vanguard Consulting (n.d.) 'The Vanguard Method in ASC: a case study', Vanguard Consulting. Available at: https://beyondcommandandcontrol.com/library/articles/non-academic-articles/the-vanguard-method-in-asc-a-case-study/ (Accessed: 16 June 2026).