Commissioning will never work in the NHS

Hallelujah! Let us assume the improbable; that following Sir David Nicholson’s speech at the NHS Confederation conference in June politicians and policymakers accept that commissioning, as an activity separated from health provision, is never going to work; that commissioning is the dead horse that finally deserves a rest. Assume they are seriously ready to consider the alternatives. What would the new model look like?

It certainly would not be a simple piece of magic, because health and its interactions with social care and wider society are so complex. However, looking at what has worked before might help.

‘There is no evidence that GPs as a group are empowered with supernatural abilities to manage large budgets and organisations’

The NHS has been through about 10 major reorganisations (it depends how you count them) since 1948, though the first was not until 1974. So a lot of options have been tried, although in every case the next reorganisation has been bubbling under before the last one has had the chance to be tested. The current changes have much in common with the Labour reorganisation of 2000, with its clinically led, small primary care trusts.

All the reorganisations have been predicated on the assumption that politicians, think tanks and management consultants know best, and clinicians and managers in the NHS are not to be trusted. Ironically, this has even been true during the several occasions when the change was badged as putting the NHS in the hands of frontline clinicians.

The right configuration?

So what would be the main characteristics of an alternative system based on previous experience? The key features would be:

• Integration of service provision and planning around a defined population and individual patients.
• The best degree of fit possible with social care and other local government services.
• Integration of support services for the defined population, crucially finance and information, to reduce unnecessary overheads.
• Consistency of policy around the key indicators of health of populations, patient outcomes and their experience so comparisons can be made across organisations and time.

There is no right answer to the configuration of health organisations across England and the solution will always be a compromise. However, experience would suggest that London is always a special case and should not influence the best arrangements for the rest of England.

Outside London, each health organisation will generally have one main acute provider, though in urban settings this may be on more than one inter-related site. They have to be large enough to have economies of scale, but not so large they are remote from the population or frontline professionals, while strongly relating to local government arrangements.

Previous configurations suggest in the order of 100 health organisations for England gives the best compromise.

Each of these organisations should be responsible for planning and provision for the resident population and the provision of services to residents from elsewhere who choose to use its services.

Planning and the role of GPs

An increasing number of services are only effectively planned and/or provided at a regional or national level, and it is difficult to imagine how the NHS could function without at least a strong planning and coordination role at the intermediate level. The more specialised the service, the less practical any form of meaningful competition becomes.

General practice has a special role in the NHS as the gatekeeper for much of health and social care. It has the immensely important and complex job of weighing up the best interests of the individual while having regard to the overall best use of resources.

‘Why the fetish for separating commissioning from provision, despite the paucity of evidence it will ever work?’

However, there is no evidence that general practitioners as a group are empowered with supernatural abilities to manage large budgets and organisations or plan and develop complex services. Indeed, if we fall for that one we compromise their relationship with their patients both in focus and in time. Rather, GPs and their representatives must be engaged in the planning and resource allocation process without dominating it.

If we discount the full democratisation of health in the near future − either directly or through local government − then the bodies charged with overall control of their population’s health must be as widely representative as possible, including local government, social care, patient groups and the professions.

General practice does need to be closely integrated with resource use in a way that encourages the fair share of resources across the population served. However, this is technically very complex and subject to large swings, such that it is unlikely to be possible to create robust practice budgets unless practices get very large and lose their personal touch.

Unnecessary division

For the last 20 odd years, dividing the health service into commissioning (or purchasing) and provision has been the only show in town. First, NHS trusts were divided from health authorities and GP fundholders added to spice the brew. Then primary care trusts were created with practice based commissioning bolted on.

Interestingly, in both cases, GP purchasing/commissioning was run in competition to health authorities/PCTs; rather than to provide synergy. When this ran into difficulties, particularly in restraining the costs of acute trusts, the “world class commissioning” programme was created and PCTs were encouraged to buy in all the best brains in the private sector to smarten up their act. PCTs were even forced to divest themselves of direct management responsibility for community services in case this sullied the purity of their commissioning role.

Now all faith is being placed in clinical commissioning groups and GPs being the magic ingredient that will make commissioning the powerhouse of efficiency and effectiveness in the health service.

‘The NHS does not make iPhones and there are killer reasons why commissioning health is fiendishly more difficult than smartphones’

But why the fetish for separating commissioning from provision, despite the paucity of evidence it will ever work? Why is it that many highly successful private sector companies regard vertical integration as the preferred model?

Apple is one of the smartest and most successful companies on the planet: it knows the art of commissioning down to perfection. Its ultra-smart scientists, designers and marketers plan their attractive products in California. The iPhone, for example, is specified by performance, looks, physical characteristics and price, down to the very last detail. Then the product is sourced and produced in China to the exact specification. And it works. Each phone identical to the next one, and the next one. This is commissioning perfection; so perhaps we should get Apple to run NHS commissioning with a little help from GPs?

Most of you will have spotted the flaw in the argument. The NHS does not make iPhones and there are, sadly, four killer reasons why commissioning health is fiendishly more difficult than smartphones.

Health provides millions upon millions of unique packages of care and treatment for individuals. We may parcel them up for convenience and management as, say, “hip replacements”, but in reality, Mrs Jones’s hip replacement (she may have multiple comorbidities) is not the same as Mr Smith’s (he may have dreadful social circumstances).

“Commissioning” is one of those Alice In Wonderland words that can mean pretty well anything you want it to. (Liberating the NHS defined commissioning as the process of assessing the needs of a local population and putting in place services to meet those needs.) In health, what tends to happen in practice is that commissioning gets redefined in more detail at several levels, the nearer it gets to the individual patient. There is no clear boundary between commissioning and provision; it is a continuum.

The heart of the matter

So why does commissioning not work for health?

• Complexity and scale. In most health interventions, there are no agreed outcomes or quality measures or realistic ways of monitoring them. Even with the imperfect parcelling of packages of treatment and care, there are then potentially thousands of them each requiring defining and setting out in some form of contract that must then be monitored. All the time, treatment/care protocols are changing, sometimes because of scientific advances, sometimes because of differing professional preferences, other times due to the resources available at the time (no two doctors or health professionals have exactly the same skill set and competence). This mapping and monitoring of all healthcare, even if it were achievable, would be hopelessly expensive and bureaucratic.
• Fragmentation and coordination. Although some health interventions require just one hospital or other service input, most require services integrated across several organisations and revolving around the needs of an individual patient. In these cases the commissioner either needs to be omniscient and manage the boundaries between services or rely on a lead provider with all the consequent scope for buck passing and confusion.
• Technical expertise. It is difficult to imagine how commissioning specialist services from, for example, hospitals can be done when most of the technical expertise resides in the provider rather than the commissioner. Unless, of course, the whole process is cooperative rather than competitive, which rather defeats the objective of splitting commissioning from provision
• Barriers to entry. With the exception of London, almost all health economies rely on one principal provider of acute hospital services and one main provider of mental health. Having real choice then requires a new provider either cherry picking the easy and profitable services or massive start-up costs and surplus capacity. In either case, the NHS picks up the additional costs.

Of course, it maybe that the people who worked for PCTs (including many of the GPs now leading commissioning) just weren’t very good at commissioning and all we need to do is get in a private healthcare company to do it for us.

Unfortunately, this does not alter the basic problem of defining the indefinable; but worse you have to construct a contract for delivering the undefined commissioning. Undoubtedly this would be good for lawyers but not so good for taxpayers or patients.

‘Perhaps now is the time to raise our eyes from the ground and seriously consider the alternatives to commissioning’

Commissioning does have a place at the margins of health, where there is a system failure or the need to redefine the way in which a service is delivered. In these cases, some controlled competition can be very useful. There have been many good examples of service improvement achieved by PCTs, GPs and clinical commissioners. The problem comes when you try to scale these processes up to cover all healthcare required by a population and begin to drown in the bureaucracy created.

Given the number of different models of commissioning health that have been tested to destruction and found wanting, perhaps now is the time to raise our eyes from the ground and seriously consider the alternatives. Is commissioning the dead horse that finally deserves a rest?

Chris Fewtrell is senior associate at the Health Services Management Centre, Birmingham University, and a research fellow at the school of health and related research at Sheffield University