General practice is making a leap in the dark

BMJ 2016; 355 doi: https://doi.org/10.1136/bmj.i5698 (Published 28 October 2016) Cite this as: BMJ 2016;355:i5698
  1. Martin Marshall, professor of healthcare improvement1,
  2. Denis Pereira Gray, emeritus professor2
  1. 1Department of Primary Care and Population Health, UCL Sir Ludwig Guttman Centre, London E20 1AS, UK
  2. 2University of Exeter, Exeter, UK
  1. Correspondence to: M Marshall martin.marshall{at}ucl.ac.uk

New models of working risk throwing the baby out with the bathwater

A strong case is being made in many countries that the traditional model of general practice needs to change. Critics claim that practices are too small and too isolated, that they are increasingly unable to meet their patients’ needs and expectations, and are unfit to lead the necessary redesign of health systems.1 2 As general practice in the UK in particular struggles with a demoralised workforce and inadequate resources,3 these criticisms are being taken on board. Quietly but rapidly, and in a largely ad hoc fashion, general practice is changing; small practices are closing or merging with other practices, practice networks are forming, the primary care workforce is becoming increasingly multidisciplinary, and new integrated models of care that bring together community and hospital based services are being developed.

Many of these changes may be good for patients and for the health system, but insufficient attention is being paid to the possible unintended consequences. One substantial risk should be exercising policy makers but is not doing so: that the emerging new models may not deliver the same benefits to patients and the health system as the traditional model.

We know that health systems with a strong focus on general practice deliver better outcomes at lower cost than those that are more specialist oriented. Starfield and others have shown that effective general practice is associated with better outcomes (including life expectancy, early detection of cancer, and reduced deaths from cardiovascular disease), better system performance (including fewer hospital admissions, lower cost, and reduced health inequalities), and better patient experience (including high rates of satisfaction and trust).4 5 The evidence, although observational, seems consistent over time and across different health systems.

What are the likely mechanisms by which general practice achieves such important outcomes?6 7 Firstly, it provides accessible care to all communities, including those with the greatest need and the greatest scope for improvement. Secondly, given that a large proportion of health is socially determined,8 the “whole person” orientation of general practice care is more likely to be effective than the disease orientation of most medical specialties. Thirdly, in dealing with uncertainty and managing risk, general practice reduces the likelihood, consequences, and the costs of overmedicalisation. Fourthly, while specialists are generally better at adhering to single disease guidelines, generalists are more effective at dealing with the growing epidemic of multimorbidity. Finally, general practice care is more likely to focus on prevention and on enabling patients to look after their own health. A commitment to continuity of care and general practitioners’ sense of responsibility for individual patients underpins these mechanisms.9

Policy makers and health system leaders should be asking whether changes to the structure, governance, and working processes of general practice will enable it to continue to deliver these benefits. The jury is not just out on this question; it has not even been convened. We simply do not know whether clinicians who have not had an opportunity to build a personal trusting relationship with their patients because they work in large organisations will be as effective at preventing unnecessary investigations or referrals. We do not know what effect moving experienced medical generalists back from the front line of care will have on the effective use of resources. We do not know whether working as an employee in larger organisations will engender the same level of good will (a poorly valued asset) as does working independently in a small practice. And we do not know whether better access to diagnostic facilities and to medical specialists in community settings will result in over use and higher costs.

These are fundamental questions that researchers need to answer. To do this we must ensure that published evidence about how to organise and deliver health services is mobilised in a way that is useful to decision makers.10 For example, we already know much about the effect of practice size on patient experience and outcomes11 and about the benefits and risks of different models of multiprofessional working. This evidence currently has insufficient influence on service redesign, and as a consequence many of the changes are ineffective and wasteful.

We also need strong commitment to evaluating the new models, building on early research into the nature and impact of scale recently published by the Nuffield Trust.12 When there is uncertainty about the best way forward, action can be based on past experience and good theory. But rigorous evaluation using the full range of available theories and methods is essential. We need to describe the rationale and the nature of proposed changes, to understand the facilitators and barriers to change, and to know whether they are achieving their intended outcomes as well as the occurrence of any unintended outcomes. In the absence of this commitment, the NHS is at serious risk of throwing the baby of general practice out with the bathwater of health service reform.


  • Competing interests: We have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.

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


View Abstract

Sign in

Log in through your institution