IN 2002 the BMJ published a paper claiming that Kaiser Permanente, the long established and much respected health maintenance organisation based in California, gave better value for money than the National Health Service (NHS) in the United Kingdom (UK).1 In order to make the comparison the authors had to make a number of assumptions. Numerous letters later pointed out where many of these assumptions might have been mistaken and how they almost certainly tilted the scales against the NHS. Despite this, the paper proved influential among policy makers, who set about identifying how Kaiser manages to be so much more efficient. A second paper duly appeared a year later, confirming the original suggestion that better integration between primary and secondary care leads to much shorter hospital stays, and a more efficient, cheaper service overall.2 Two months ago, when we published a paper challenging the findings of the original article,3 the authors countered by stating that, although there were indeed problems inherent in the original comparison, the later paper confirmed that the NHS has much to learn from Kaiser in terms of better integration between inpatient and outpatient care.4 When the Talbot-Smith paper was featured in the Guardian,5 it evoked a similar response from John Reid the Secretary of State for Health: ‘Kaiser … is a model of care from which we can learn’.6
Sceptics might ask themselves first, whether there would be such interest in Kaiser's integration if the original paper had made different assumptions and concluded that, of the two, the NHS was the more efficient system. Then there is case mix; we have enough difficulty trying to allow for different case mix when comparing neighbouring hospitals in the UK. How can the authors be sure that they have successfully corrected for it when comparing two systems across huge chasms of geography, ethnicity, culture, and system of healthcare? Trevor Sheldon warned against the dangers of assuming that practices can be imported and successfully incorporated into our own system.7 Again this shouldn't surprise detached observers of the NHS, where we don't always manage to spread best practice into every corner of the NHS, both primary and secondary sectors.
Yet it is impossible to dismiss the conclusion altogether. Many UK doctors, both general practitioners (GPs) and hospital specialists, would surely agree that the way in which patient care is passed between the two sectors is often cumbersome and inefficient, and can be a source of endless frustration. Despite the changes in undergraduate and postgraduate education, much mutual misunderstanding of the content and pressures of each others' work persists between GPs and specialists. However good the service, we should always be willing to learn from others, and the similar way in which health maintenance organisations use registered lists and primary care referral to manage costs make them a suitable model for UK planners.
The system of referral, with GPs acting as gatekeepers to specialists, is a product of the complete division between the two that has existed for many years in the UK. Doctors in the UK accept it as an immutable part of the medical landscape. However, it is worth remembering that, far from being built into modern medicine, the UK was, at least until comparatively recently, unusual in having such a deep divide between the two parts of the profession. Sadly, the most acute chronicler of the divide, Frank Honigsbaum, died recently. He charted its history in a book of 1979,8 and later described the problems he felt it had created in two articles in the BMJ.9,10 The separation arose from a long sequence of developments, beginning with, and perhaps before, Lloyd George's National Health Insurance Act of 1911. Critically, when the NHS came into being in 1948, with a shortage of hospital doctors and enough GPs available to cover some of the gaps, the opportunity existed to reduce the division. But familiar difficulties conspired to prevent this happening: mutual suspicion between the two groups, exacerbated by different administrative bodies and systems of payment; the desire of hospital doctors to control and improve the standards of hospital care; as well as the need for GPs to be on call for their primary, community-based responsibilities 24 hours a day.8
When he returned to the theme in 1985, Honigsbaum was despairing of what had happened to general practice in the UK:
‘General practitioners working in the National Health Service are independent contractors who rarely recognise that a reciprocal relationship exists between them and the health service: to a large extent both depend on the survival of the other. It is hard to see how the restricted range of general practitioner care in Britain could continue without the financial security provided by the health service.’9
He saw GPs failing to use their clinical skills, indeed without access to facilities or equipment to use them (remember this was 1985, not the 1949 of the Collings report11), failing to achieve satisfactory immunisation rates, failing even to provide continuity of care since any contact with the hosp-ital meant such a dramatic loss of continuity. Honigsbaum's solution was to recommend, with uncanny prescience, widespread employment of nurse practitioners to take on much of the routine work, and longer consultations, leaving GPs to concentrate on extending their range of clinical skills.
At the same time, Honigsbaum tacitly recognised that the separation was responsible for the maintenance of a cadre of GPs covering the whole country, which guaranteed the NHS its two cardinal principles of free access and universal coverage, while still paradoxically keeping health costs ‘among the lowest in the world’.9 To that we might add some other aspects of general practice in the NHS that we tend to take for granted, but which may owe their existence, at least partially, to the historic separation. It has enabled general practice to develop as a distinct academic discipline with its own body of knowledge, making a distinct contribution to the undergraduate and postgraduate education of all doctors. The discipline retains two distinctive features. It remains generalist, itself a triumph when medicine in every other respect has become so specialised. It values personal doctoring, even if recent developments have seen the progressive dilution of this aspect of care. The resulting enhancement of its professional status has helped general practice market itself as a sufficiently attractive career option to maintain the workforce and continue to guarantee universal coverage that Honigsbaum welcomed, at least for most of the last 30 years. Finally, while integration with the hospital sector has been notably lacking, enormous efforts have been made to integrate with other primary care agencies in order to try to realise the promise of coherent health and social care in the community.
So when the policy makers and planners start exhorting us to integrate better with hospital medicine, we should be wary. In the business world, attempts to translate best practice throughout an organisation often fail.12 Recent efforts to integrate by reviving the idea of getting specialists to conduct consultations in community settings have proved expensive without encouraging much integration.13,14 Even if it could be achieved, integration might seduce GPs back into the exciting world of hospitals and threaten so much of what has been gained in the last 30 years. On the other hand, perhaps all those developments have now made general practice mature enough to be able to take on the challenge of such integration without losing any of its distinctive characteristics. Ideally, if we could extend the idea of integration so that it encompasses everyone in the NHS working together it would certainly be worth the effort. If not, it might be better to stick with the separation, and stay inside the ‘Chinese walls’ described by Aneurin Bevan in 1954.15
Acknowledgments
Thanks to Professor Mike Pringle who commented on an early draft of this editorial.
- © British Journal of General Practice, 2004.