Intended for healthcare professionals


Competition and integration in the English National Health Service

BMJ 2008; 336 doi: (Published 10 April 2008) Cite this as: BMJ 2008;336:805
  1. Chris Ham, professor of health policy and management
  1. 1University of Birmingham, Birmingham B15 2RT
  1. c.j.ham{at}
  • Accepted 12 March 2008

Market based reforms of health care have helped cut waiting times, but Chris Ham argues that a different approach is needed to meet government objectives on disease prevention and chronic diseases

New healthcare objectives are often announced without changing the mechanisms used to implement them, leading to delays and frustration for policy makers. Gordon Brown and the ministerial team at the Department of Health face the challenge of avoiding this problem after the prime minister’s first major speech on the NHS signalled that greater priority is to be given to disease prevention and the treatment of chronic diseases.1 What changes need to be made to the health reform programme to enable these priorities to be implemented?

Policy options

One approach would be for the government to revert to the use of targets and top down performance management to achieve its objectives. Although there are some signs that the government may be moving in this direction—most obviously the instruction to the NHS to undertake a deep cleaning programme of hospitals—it seems unlikely that top down performance management will be the main means used to implement new priorities.

Top down control is known to have important limitations. These include the disempowerment of staff, the stifling of innovation, and the risk that areas of health care not identified as national priorities will be neglected. At a time when Lord Darzi’s review is seeking to engage front line staff in the next stages of reform and to put clinical teams in the driving seat of change, it would be inconsistent to revert to command and control mechanisms.

An alternative approach would be to continue with the implementation of the quasi market reforms introduced by Tony Blair. These reforms, centred on offering patients a wider range of choices between NHS and independent sector providers and ensuring that money follows patients through payment by results, have contributed to the reductions in waiting lists and waiting times. They are likely to have a continuing role in implementation of objectives such as the maximum 18 week waiting time from referral to treatment.

The Brown government has sent out mixed messages about its attitude to the quasi market. On the one hand, it has scaled back the independent sector treatment centre programme, while on the other it has encouraged new providers to enter the primary care market. Although ministers have indicated their support for policies on patient choice and increased plurality of provision, the prime minister has argued that health care is a sector that shows the limits to markets.2

Even if the government was committed to quasi markets in health care, it is doubtful whether they would be able to ensure that priority is given to disease prevention and chronic diseases. This is because the quasi market in its present form was designed to support the implementation of other priorities, most notably improvements in access to planned care, and it will suck more resources into acute hospitals in the absence of countervailing pressures. The key question facing health policy makers is whether these countervailing pressures are in place.

Problems of practice based commissioning

In theory, primary care trusts and practice based commissioners can use their resources to support population health programmes and to strengthen care outside hospitals. In practice, commissioning was the weak link in the internal market in the 1990s,3 and current arrangements seem unlikely to be more effective.

The Audit Commission’s most recent review of practice based commissioning found that it was being led by a few enthusiastic practices working with supportive primary care trusts.4 It found little evidence of practices engaging with public health staff and local authorities on the broader population health agenda and therefore a risk that resources would not be aligned with primary care trusts’ strategic objectives. The review concluded that there were “few signs of the scale of service change envisaged by the Department of Health.”4

Ministers have spoken boldly of their ambition to develop world class commissioning, but international as well as UK experience indicates no healthcare system is commissioning consistently well.5 An explanation of this can be found in the work of economists who have analysed the transaction costs arising from negotiating contracts between buyers and sellers and monitoring compliance with these contracts. Coase and Williamson contend that where transaction costs are high because of the complex nature of the products being bought and sold, integration has advantages over market based relationships.6 7

Need for integration

The case for integration is empirical as well as theoretical. A recent example from the NHS is the experience of the Birmingham East and North Primary Care Trust in developing community health services. The trust’s chief executive has described the benefits of doing this work within the trust rather than contracting with external providers:

“If we had tried to specify and tender this service from the start we would have got it badly wrong. We would have been tied into contracts, which limited our ability to make adjustments without cost and delay . . . In-house, we have been able to make a difference in quality and service availability in days or weeks.”8

The advantage of integrated approaches is confirmed by experience in the United States, where systems such as Kaiser Permanente and the Veterans Health Administration achieve high levels of performance.9 A key feature of integrated delivery systems is that they internalise commissioning by combining the roles of insurers and providers.

Integrated delivery systems are relevant to the English NHS because they focus on the prevention and treatment of chronic diseases and deliver good outcomes for patients with these diseases. These outcomes are achieved by giving multispecialty medical groups control over capitated budgets, thereby creating strong incentives for doctors to keep patients healthy and minimise expenditure on future medical interventions. A study of Kaiser Permanente showed the benefits of aligning the incentives for clinicians with those of the healthcare organisation.10

How integration could work

The experience of integrated delivery systems indicates that the design of the English health reforms needs to be reviewed. An alternative to the quasi market in its present form would be to encourage the development of clinically integrated groups. The challenge is how to do this without causing another lurch in health reform and a destabilising restructuring of NHS organisations. One way forward would be through groups of practices reaching into hospitals and community services to build networks of care for people with chronic diseases and groups of patients who also need social care, such as frail older people. Another path to integration would be through hospital based specialists supporting primary care teams to manage demand and provide care closer to home.

Clinical integration would require practices to work closely with hospital based specialists in deciding how to use their resources, especially specialists who work mainly in community settings. General practitioners and specialists would then jointly commission and provide services. As integrated groups evolve, specialists may move out of hospitals to become equity sharing partners in the multispecialty practices.

To sharpen the incentives, multispecialty practices would take on responsibility for hard budgets covering all the services needed by the patients registered with them, with the ability to retain savings and manage deficits. While the offer of hard budgets might deter some doctors, for others it would provide the spur that is lacking under practice based commissioning.

Hospital services would still be provided by NHS foundation trusts and independent sector providers. They would be organised separately from clinically integrated groups but would develop long term relationships with those groups that provide most of their referrals. Planned care would be funded under the policies of patient choice and payment by results because these policies were designed primarily with this kind of care in mind and continue to be relevant.

Social care would be included within clinically integrated groups where there is local interest in so doing. The current division of responsibility for the funding and provision of health and social care between the NHS and local government seems anomalous since older patients with chronic diseases require care that is effectively coordinated at all stages in the pathway. Establishing integrated health and social care groups would be an important step in this direction.

Primary care trusts would have a role in overseeing the performance of integrated groups and publishing data on their performance. This would include ensuring that the role of groups as commissioners and providers did not give rise to conflicts of interest that were detrimental to patient care. Primary care trusts would also be responsible for focusing on the health of the populations they serve, including working in partnership with local authorities and other agencies.

Maintaining choice

Collaboration between practices, specialists, and their clinical teams in both the commissioning and provision of care would emulate some of the features of integrated delivery systems in the United States. However, it risks creating geographical monopolies that would limit patient choice and lack the stimulus of competition that is thought to drive improvements in performance.11 To avoid this risk, it has been argued that patients should be able to choose between clinically integrated groups.12 13 14

To facilitate choice, clinically integrated groups would need to develop around like minded (rather than simply geographically contiguous) practices. Patients would then choose between competing integrated groups, and risk adjusted person based capitation payments would follow their choices. A variety of organisational forms is likely to emerge, ranging from clinically integrated groups that commission and directly provide most forms of care other than acute inpatient services, to virtually integrated models in which a core group of commissioners and providers deliver some care themselves and develop long term contracting relationships with a range of other providers.

Whatever the model, integrated groups would decide how much care to provide themselves and how much to buy in from other providers. In rural areas, where the scope for choice is more limited, there are likely to be advantages in vertical integration, whereby a single organisation takes responsibility for the commissioning and provision of all forms of care, including hospital services. The performance of these vertically integrated systems could be compared with the performance of groups operating in a competitive environment, providing an empirical test of the argument that integration and competition are both needed to improve performance.

Next steps

The ideas set out here provide an outline of an alternative to the quasi market currently being implemented in the English NHS. The current mechanisms of reform are not aligned with the government’s policy objectives and will inevitably lead to frustration on the part of politicians, delays in the implementation of policy, and reinforcement of the argument that government is not competent to run health services.

More work is needed on the detail of how to garner the benefits of choice and payment by results for planned care, and the potential of clinically integrated groups for other areas of care. In view of the uncertainties involved, clinically integrated groups should be piloted before widespread implementation, building on experience in those parts of England which have been at the forefront of integrated care.15 As part of this programme, the scope for offering patients choice between groups needs to be explored and tested.

Summary points

  • The quasi market being implemented in the English NHS was designed to improve planned care

  • Newer priorities on disease prevention and chronic diseases are not well served by the quasi market

  • Clinical integration may be more beneficial

  • Policy makers need to review the design of the market


  • doi: 10.1136/bmj.39524.400498.AD
  • Competing interests: CH was director of the strategy unit in the Department of Health between 2000 and 2004.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract