Reconfiguring health systems

Many doctors and other health professionals are engaged in discussions about reconfiguring health services, and in particular building new hospitals. Here Andy Black, a former NHS hospital chief executive, argues that in England at least discussions about reconfiguring health services are too dominated by discussions about buildings. We lack a vocabulary to discuss other ways of investing in health gain. Perhaps surprisingly, the problems of labour supply may stimulate the radical change that is needed

At a seminar on “the hospital of the future” earlier this year, the Department of Health estimated that 75% of English hospitals were involved in some form of reconfiguration debate—that is, discussions about where and how secondary and tertiary services should be provided. The Royal College Physicians claims that 61 English hospitals are isolated, provide only acute medical services, and should therefore be “reconfigured.”¹ Many English readers of this journal will be involved in these discussions. These discussions should also be familiar to health professionals worldwide—though the precise balance of problems and opportunities that I discuss here will differ from place to place.

Summary points

• Discussions about reconfiguring specialist health services are dominated by ideas of centralising services in big hospitals—yet the public persists in valuing local services

• Investment in hospitals is well understood and professionals have good arguments for specialisation and centralisation

• The NHS has no mechanisms for encouraging or assessing different sorts of investment for health gain

• Labour supply problems and the promise offered by information technology may both force and enable future reconfigurations of services that genuinely improve health

Why reconfigure?

Health systems must balance two purposes. In the short term they must respond to the demands of the population for access to existing services. At the same time they must try to improve the health of the whole population: get the waiting list down this year and improve coronary health status in the longer term.

In a logical world these reconfigurations should be helping with one or both of these aims. But more often they are provoked by a search for financial “breakeven” or to escape some perceived imminent collapse of clinical viability (often associated with emergency on-call rotas.) This year's spectre is the “working hours directive”—a phrase on everyone's lips and a reference to European legislation limiting hours of work to 48 hours a week. It means that junior and senior doctors can no longer be relied on to provide the long hours they have traditionally worked.

Reconfigurations are supposedly set in the context of clear strategies and are built on the foundations of a strong local consensus among all stakeholders: the rational consideration of options leads to the selection of the best solution in the circumstances. But often it's not like this. Sometimes they are a bloody battle through the courts and elections—as in the experience of the former Worcestershire Health Authority and its allied NHS bodies.²

Here the NHS managed to achieve its objectives—concentration of services in a rebuilt hospital in one town and the closure of an old hospital in another town. But it also achieved the first collateral total reconfiguration of the political system in Kidderminster (the town that “lost” it's hospital). A new political party was born out of the battle and triumphed in both the local and national elections (see www.healthconcern.org.uk).

As we look across these many reconfiguration debates, there are perhaps four problems that consistently present themselves and four positive angles that are newly emerging. The problems are: (a) a lack of a clear investment framework for health gain; (b) discussions entirely dominated by buildings; (c) employment patterns and pay policies underused as levers for change; and (d) a bias towards centralisation.

Problems and opportunities

Problems

Lack of investment framework for health gain

Discussions dominated by buildings

Employment and remuneration underused as engines of change

Bias towards centralisation

Opportunities

Managed clinical networks

Virtuality and knowledge technologies

Ready access to capital and increasing revenue

Recognition of labour supply problems as a design issue in health systems

Problems

Lack of investment framework for investing in health gain

Major investments in local health systems often mean that the pattern of local services has to be changed. The people putting up the money are often much clearer about what a good loan looks like than the people using it are clear about what a good health service looks like.

Commercial investors have clear models and investment criteria to assess the commercial quality of any investment. As well as the numerical calculation of return on investment over time, the rating agencies will, as needed, attach a qualitative tag to debt which is clearly understood across the international investment community. Usually the actors are experienced: this will be one in a long series of similar investments they have been involved in.

Not so on the health services side. Few people outside academe life have any awareness of the concepts of health gain or social gain. What commercial investors get back is more of the same—money. What social investors do is use money to buy “social good.” Even the rudimentary instruments for measuring the social good of health gain—for example, quality adjusted life years³—are controversial.

When the big NHS investment deals (which often determine configurations of services for generations) are being shaped up, the health gain experts are not in the room. Even if they were in that room, as things stand today, they would not be credible communicators across the languages of management, investment, and public health. The point here is that those shaping and approving health investments have weak and untested tools with which to evaluate investment returns. Unfortunately they are also often first time buyers.

NHS reconfigurations are building-centric

The rhetoric is about a “whole systems” approach, but the reality is point investment in specific projects which are often large new hospitals. Why is this? A part of the answer comes from the ease with which doctors, managers, and politicians can coalesce around such a project, each getting a good fit with their own agendas.

The smooth capture of the private finance initiative process by the UK construction industry may prove a great and unremarked tragedy for the NHS. This is the continuation of the association in the minds of many in the NHS that big investment=new buildings. The Thatcher era saw the lowering of the Berlin wall between private and public investment. Suddenly the NHS had, at last, access to investment funding. The great bulk of these early investment waves will probably be channelled into asset based project finance: buildings, new buildings, big buildings.

The construction industry likes building things. NHS Estates likes building things. The banks like to lend money secured on things that have been built. The NHS bureaucracy likes buildings. Of course buildings are necessary, but they do not themselves contribute much to “health gain.” They certainly do not comprise a balanced investment ticket.

The private finance distraction

Is the furious debate about the private finance initiative really important? The argument whether private finance initiative is better value for money than public procurement may be an argument about the relative speed of two bullets racing towards the wrong target. Those with an interest should read the parliamentary select committee report:

Although our inquiry looked at public private partnerships of all kinds in the NHS, we received the most evidence on the PFI. A wide range of organisations and interested parties gave evidence. Even a cursory examination of the material suggests how polarised the debate has become, with exaggerated claims being made on both sides of the argument in a climate not conducive to rational analysis. (para 59) ⁴

Questions never asked

Perhaps this debate should shift focus on to the question: What are big hospitals for? The Byzantine capital approval processes in the NHS⁵—which are a jobsworth sham—never pose this question. The fine rhetoric of the “strategic outline case” leads inevitably to a mountainous “outline business case” which compares five different ways of buying a new hospital with an apocalyptic “do nothing” option.

When will we see an option appraisal which compares spending £500m ($790m; €790m) on a new hospital for, say, east London compared with £500m spent on hypermodern systems for chronic disease management or £500m spent on an integrated health system for the east end of London of the sort typified by Kaiser Permanente in the United States?⁶ Better still, when will there be open competition among investors to come forward with independent investment plans that will deliver measurable health gain benefits.

Remuneration underused as an engine of change

Three quarters of all health expenditure goes on paying staff. Yet the design of employment and remuneration structures is underused as an engine of change. When local NHS leaders are discussing the configuration of health services that would best fit their local community, they quickly encounter a range of constraints imposed by fixed employment regulations and pay systems. There has been a longstanding discontinuity between the national settlement of remuneration and employment issues and the attempt to design health systems on a local basis. This is not to say that it is made impossible: it is just that one of the most important levers of local change is mostly locked in a fixed position.

One striking exception was the advent of general practitioner fundholding. Everyone seemed to agree that it was effective but disagreed about whether it was destroying or revolutionising the NHS. It certainly showed the latent power of changing reward and financial structures. The Economist recently made a passing link between fundholding and subsequent improvements in hospital productivity.⁷

An example here. It's current government policy to have an NHS led by primary care, and primary care trusts have been formed to provide primary care and commission secondary and tertiary care for their populations. Yet several UK cities have mounting vacancies for general practitioners, which look set to increase, as the recruitment market is adverse. The current configuration of primary care—of having one general practitioner per 2000 citizens—is probably unsustainable for the future. The need for reconfiguration and investment is recognised, and the local improvement finance trust (LIFT) programme to allow primary care access to private finance is being introduced. Typically these LIFT programmes are already focusing on the buildings that will be needed.

Meanwhile, the government is coordinating a new national standard contract for general practitioners, but it is not clear how this will help a northern city like Liverpool to design and introduce a “new deal” which will attract good quality young doctors to practice primary care in areas of high social deprivation. An integrated approach in a single city to redesign the pattern of services, reinvest in the buildings, and introduce completely new working practices is still a theoretical objective but a practical nightmare.

A bias to centralisation

The bias towards centralisation that is apparent throughout the NHS applies particularly in hospital provision. There are many drivers, including falling lengths of stay and the increasing overhead costs of running a modern hospital. But perhaps the most important is the nexus of medical training, accreditation, and governance issues that militate against the small local acute hospital which is often the heart's desire of the local community. Again there is a discontinuity between what the professional rational policy makers want and what the local politicians and community stakeholders will buy: Kidderminster again.

The medical royal colleges, which control specialty training, do not design ideal health services. They are more likely to design ideal working conditions for their members. Decades of increasing subspecialisation have created an undertow towards increasing centralisation of services. Narrower and narrower fields of interest need larger and larger catchment populations.

There are many advantages in centralising acute services, and these are more or less continuously advocated. A random sample of NHS consultation documents would produce a good cross section of reasons why it would be better to regroup acute service on fewer larger sites. Policy statements from the medical royal colleges and the British Medical Association would also produce many valid arguments as to why there are professional and governance advantages in practising in teams with differentiated clinical work and the right equipment.

Health professionals are constantly irritated that the British public resists this logic and insists on judging health service configurations purely from a locality point of view. Yet in large parts of Wales, Scotland, Northern Ireland, and rural England locality really matters.

The point here is that in striking a balance between the benefits of centralisation and the desirability of local access there is a bias to the former. The more difficult challenge of finding safe, effective ways of running acute services in a decentralised manner has been under addressed.

Opportunities

And so to the positive side of things.

Managed clinical networks

One of the great advances in NHS planning was the development of the concept of managed clinical networks in Sir David Carter's review of acute services for Scotland.⁸ It is still an emerging concept, theoretically attractive but practically difficult to implement. However, it is a much needed addition to the slightly more worn concepts of centralisation and “hub and spoke.” The review recognised that such networks would need much development work and could range across traditional planning boundaries:

Managed clinical networks are consonant with a renewed emphasis on the role of primary care in acute health care. Far from favouring centralisation, such distributed networks will promote the delivery of acute services in new collaborative organisations, transcending traditional boundaries between hospitals, community hospitals and primary care.⁸

Virtuality and knowledge technologies

Perhaps “virtuality” is one of the main core issues for the future of health service delivery. The question is: to what degree can virtual consultation supplant physical presence? This is a question for the future and not one that can be resolved now. A generation of patients and health professionals must have time to explore the possibilities and adapt the clinical arts and science.

Two points could be made in passing. Firstly, major hospital projects are typically bringing debt cycles of 30 years. Secondly, the first mention of the mobile telephone in the UK media was in the Goon Show in the 1950s, when it rang in someone's pocket and he answered it—it got a real laugh. Decades pass swiftly in health planning and technology development. Virtuality holds one exciting prospect that has hitherto been unobtainable—in which both the clinician and the patient can be where they want to be but still give and receive a service as effective as if they were physically present together. The clinical professionals can work together in centres and the patients can stay in their localities. It is not possible today but is it an attainable future goal?

Alongside virtuality sit the new computerised “clinical knowledge systems” as a second source of new design possibilities. These knowledge systems are complicated in design and purpose and would require a separate paper to explain. For brevity I'll define them as meta-computer systems that sit above existing health computing systems and use the myriad individual transactions being entered about individual patients, their health status, and their consumption of drugs and services to learn (and predict) how healthcare systems work. We are already seeing the introduction of early and rudimentary applications (NHS Direct) that sift information and proceed through layered algorithms to direct the health worker in real time. We can expect more and we can expect current systems to be continuously refined through use.

Consider parallels in other endeavours. In the 1970s and 1980s we chess players scoffed at the early computer chess engines that we beat with ease—now 30 years later there are perhaps two or three humans alive with an even chance of drawing with the best of them. We no longer scoff. “Medicine is not chess,” I hear you say: time will tell.

Ready access to capital and increasing revenues

The private finance initiative and public private partnerships in theory remove constraints on access to investment where a sound economic case exists. Furthermore, the British government's guarantees of an expanding funding base provide an unprecedented positive backcloth against which to plan health services and design health systems. For decades we have planned in a scrimp and scrape climate. Now more money should offer the possibility of more health gain.

Recognition of labour supply problems

It might seem odd to cite labour supply problems as a positive development, but a plentiful supply of relatively cheap labour is arguably a very conservative influence. Since the 1980s we have known that a large proportion of the time of trained clinical staff is used up by simple documentation and scheduling tasks. Lathrop measured this in the USA in his brilliant treatise Restructuring Health Care⁹ (which marked the origins of the “patient focus” line of thought for streamlining episodes of health care) (see figure).

This chart should be studied in the light of the recent Wanless review on the funding of the NHS.¹⁰ Although couched in the polite rationalism of such documents, the meaning between the lines is there for all to hear:

Changing skill mix and increasing workforce capacity cannot happen quickly; it needs to be planned and actively managed. The workforce modelling suggests that there is sufficient capacity, but only just, in the short term to deliver the three scenarios. But before the end of the decade, there needs to be considerable progress on skill mix and pay modernisation to avoid capacity constraints. (para 5.57)¹⁰

The NHS is coming from a tradition of mass employment of relatively lowly paid staff with low technology adoption but is travelling towards an era of labour supply problems and escalating labour costs. This will be a marvellous stimulant to do things differently.

Thoughts for those discussions

And so, if your health system is in the throes of a reconfiguration debate but is merely shuffling the old furniture on a sloping deck do not despair. Another reconfiguration will be along soon, and perhaps then it will be possible to emphasise some of the emerging innovations and rely less on the hackneyed old solutions.

Figure.

DAVID DAVIES/PA

The furious debate about PFI is a distraction

Figure.

How health staff time is used (adapted from Lathrop⁹)