NHS Reforms – 2 – Get rid of the market

Posted on

The biggest criticism of the NHS is that it is riddled with bureaucracy.

The criticism is based on fact. The NHS is riddled with bureaucracy. However, the question has to be asked, why is this the case?

The answer is quite straightforward. This bureaucracy was introduced when the NHS was turned into a quasi-market.

To be a market the NHS had to be broken up into large numbers of quasi-independent units all contracting with each other, all transacting with each other, and all having to spend vast amounts of time managing those contracts and transactions before then accounting for them.

There were several hundred Primary Care Trusts (or their predecessors) when the process of marketisation began. This number has been reduced to 150 or thereabouts now. But shortly they'll be replaced by 500 GP consortia.

And then there are Foundation Hospitals. And Ambulance Trusts. And Mental Health Care Trusts, and on and on and on, all of which are undertaking all those contracts, transactions and accounting.

All of this is ludicrous. There is one NHS, paid for out of one taxation fund, ultimately accountable to one minister in England (and others, I admit, in Scotland and Wales), consolidated into one final central account. Which means that all that separate accounting is at the end of the day a complete waste of time.

Don't get me wrong: systems of control are vital in the NHS. But as I've already explained, there can never be a market in the NHS: we cannot afford for NHS services to fail and if that is the case the basic pre-requisite of the market ‚ that participants can fail ‚ does not exist. So accounting using market logics is utterly irrelevant.

In that case the obvious requirement of the NHS is that it operate at minimum cost with maximum effectiveness. And no one, but no one, would do that by creating literally thousands of self accounting units within the NHS, each with its own management structure, policies, price structures, drug formularies, own procedural rules, accounting systems, and so on if that was there aim. It takes one moment to stand back and say such a policy was bound to deliver just one thing ‚ and that that was massive bureaucracy at enormous cost with no gain in efficiency, let alone medical outcomes.

So if there's one thing that's needed to improve efficiency in the NHS it's an end to the internal market.

The alternative

It's easy to say the model is wrong. But what replaces it?

My suggestions are as follows:

Integrated management

First, primary, secondary and tertiary care within the NHS has to be under one structure. To pretend that an efficient health service can be provided without all sectors working together is absurd: it is just not possible. The NHS is dependent for its success, and its cost minimisation, on all professionals working as closely together as possible. That's the exact opposite of what Andre Lansley proposes. That means integrated and not separated management is essential. That way the benefits of economies from sectors working together can be realised. Without it there's always an incentive to save at cost to someone else's budget.

Geography matters

Second, management has to be spread over wide geographic areas to ensure that:

a) there is consistency in supply
b) layers of management are kept to a minimum
c) cooperation between branches of the NHS and between hospitals is maximised
d) the advantages of creating regional centres of excellence for some specialisms where this is essential can be secured, but without management being too remote.

So, taking as an example the area where I live, Norfolk, Suffolk and Cambridge have natural affinities that suggest they should be managed together‚ covering a range of teaching, district general and specialist hospitals all of which need to coordinate services to deliver best results whilst providing a base level of service which will have at least 90% commonality from location to location.

Beneath this secondary care core, which has to be managed on an integrated basis to ensure best delivery of specialist, non-duplicated, services, there will be underlying primary care service and community heath services. Again, these will need to vary to meet local need. For example, mid-Suffolk has some vey different needs from deeply deprived Great Yarmouth, and yet there will in all likelihood still be 90% commonality that suggests the similarities are much greater than the differences between locations and NHS operating units.

Allocate budgets based on needs

Budgets have, therefore, to be allocated regionally based on a weighting of age, social need and population to ensure consistency of likely funding across the NHS as a whole.

Within regions there has to be allocation of budgets again on the basis of weighted need ‚so areas with skewed elderly populations, for example, get extra resources at both primary and secondary level.

Within this constraint there is a requirement for consistency of service that is vital to ensure the NHS is a) consistently good and b) consistently available.

This suggests that at least 90% of funds need to be directed centrally to tasks that are considered essential and core to NHS supply but with maybe 10% at most of budgets subject to local discretion by local professionals, made up of clinicians, managers and local politicians, to meet local needs. More variation than this within the NHS is likely to break down the principle of universality that is fundamental to the NHS. Recognising this will massively minimize the expenditure of management time because decisions that have to be centrally will be done centrally to ensure consistency. So, for example:

a) It is ludicrous to have local formularies throughout the NHS when a central formulary makes more sense: the postcode lottery has to go and buying efficiency is essential;
b) Having large numbers of bodies setting policies on identical policy areas wastes vast resources: one policy will do, and will be no better or worse read than one created locally;
c) Ensuring a guaranteed base level of service is core to the social contract;
d) But if there is perceived local need then it is vital that discretion to service it be provided.

Centralise buying

To maximise efficiency it makes no sense whatsoever to have individual organisations throughout the NHS undertaking buying. One central purchasing function makes compete sense. A bed pan is a bed pan and so long as there are multiple suppliers there need be only one buyer.

Efficiency in buying is important of course. But this can be secured by having buying performance benchmarked against prices in nearby European partners to avoid supplier price fixing.

If there is central buying for the NHS then supplies should then be drawn down from NHS central supplies at standard prices fixed for annual periods. Performance monitoring on price then need only be an issue centrally.

Performance monitoring locally need then only be on use so that consumption per head of population becomes the basement for assessment of performance. This needs to be population weighted, of course: a skewed elderly population (found, for example, in seaside towns) has abnormally high use of some resources, but this is relatively easy to structure; it focuses on clinical issues and not cost and encourages management by exception, which is efficient.


Monitoring is a key part of all management, but normal performance should not require monitoring. It's a reality, recognised at the core of economic theory, that most behaviour is normal and so long as behaviour is normal with regard to resource usage it should be accepted. It is exceptions that need monitoring. As example:

a) Drug use needs to be monitored down to GP level. Average use is acceptable: over and under use needs explanation, especially if the drugs in question can be traded;
b) GP referrals need to be monitored: excess can indicate good or bad practice (specialists over refer in their own areas of expertise, for example, as they know their weaknesses, paradoxically). But it can also indicate laziness. Both are worthy of investigation;
c) Aberrant lengths of stay in hospital, aberrant levels of procedure per head of population and an absence of diagnoses all indicate areas for intervention. Clinicians must accept they are accountable for their practice, and be willing to defend it, good or bad (especially good);
d) Resource use in hospitals must be subject to similar monitoring, but so too must interaction with community based facilities when they offer alternative solutions.


Accounting must be simplified by basing assessment on medical clinical criteria, not on costs that will always be arbitrary due to the enormous problem of allocating overheads in the NHS.

The basis for assessing interaction between NHS facilities and the demand they place on each other already exists: the Read code system universally used within the NHS to record diagnoses can be used for this purpose. So, for example, a hospital need not ,charge‚a GP for use of its services for a patient: the hospital merely needs to record the use of its resource by the general practice by allocating the Read code for the services supplied to the relevant GP when issuing a clinical discharge summary: the basis for accounting is then complete, and subject to immediate review by the GP if necessary as soon as the discharge is actioned.

This does not mean that every cost of every procedure within the NHS will be accounted for: it need not be. To seek to do so would result in arbitrary allocation of costs and game playing in charge coding in any event. And there would be no gain from doing so bar extensive disputes about reallocating reward from one NHS budget to another, a process that makes no overall sense within the NHS as a whole, which is what matters.

Accounting on the basis proposed will have a focus on clinical actions ‚which should be the core focus for attention in a National Health Service. Of course money may constrain action ‚that's a political reality ‚ but that constraint has to be managed at high level ‚ at least regionally, where the interface with government that imposes the constraint is possible. To do so at any lower level is meaningless and cannot improve decision outcomes and so should not be done. Nor will anything but accounts based on clinical use of services be used ‚ firstly because clinicians do not have the skills to use financial accounts and secondly because they will rapidly realise that costs in the NHS are entirely arbitrary making any accounts virtually meaningless as measures of performance ‚ whereas clinical outcomes really matter.


So, let me summarise what has become an over-long blog.

First the market does not wok in the NHS.

In that case excellent, consistent, coordinated, low cost supply is vital to the success of the organisation.

That demands geographic coordination of all NHS services in an area, with a requirement that a basic set of functions be supplied well by all similar units (hospitals, GP surgeries, ambulance services, community health services) to which tasks are delegated within that area, all working within a consistent management and policy framework.

Buying should be centralised to ensure efficiency.

Accounting between units should be kept to the minimum possible to ensure that key performance indicators ‚the most important of which will, by far, be clinically driven, can be monitored with only aberrant behaviour being investigated.

Remaining resource usage should be on a standard cost basis within each NHS district, and maybe even nationally, to recognise a) that most mangers do not and should not have resource buying powers, excepting labour, the rest being centrally managed as far as possible for cost saving purposes and b) that the extent of budget delegation to local control should be limited to 10% at most of the local budget to ensure national consistency, and then with that local budget being determined by age and social weighting to ensure real needs ‚such as those of skewed elderly populations and inner city locations - are really met.

The advantages of this proposal are:

1) Significant saving in buying costs
2) Massive savings in management time currently expended on setting polices for quasi-competing and supposedly independent organisations all setting up systems when all under common control, giving rise to multitudinous inefficiency and hopeless budget competition;
3) Elimination of spurious accounting;
4) Making clinical outcomes the focus of monitoring by making cost control a central function and resource usage in pursuit of clinical outcomes the appropriate centre of devolved responsibility.

The result would be the end of the NHS market, and some centralization ‚ but also an end to the postcode lottery and a real re-focusing on the NHS as a health service provider.

That's something that has been missing from its management for twenty years.

It's time it was reclaimed.

Disclosure: I am married to a GP