Intended for healthcare professionals


Promoting patient safety in primary care

BMJ 2001; 323 doi: (Published 15 September 2001) Cite this as: BMJ 2001;323:583

Research, action, and leadership are required

  1. Tim Wilson, director, quality unit (twilson{at},
  2. Mike Pringle, chair,
  3. Aziz Sheikh, NHS R&D national primary care training fellow
  1. Royal Collage of General Practitioners, London SW7 1PU
  2. Department of General Practice and Primary Health Care, Imperial College School of Medicine, London W6 8RP

    Medicine has always put patients at risk. Modern medicine raises the stakes as its power to do good is accompanied by increasing potential for harm. Only now is the medical world waking up to the importance of ensuring patient safety. 1 2 The shift in the debate from individual mistakes to understanding the systemic factors that predispose to harm is welcome and offers the prospect for making important and sustained improvements in patient safety. 3 4 However, there is a dearth of understanding of patient safety in primary care—where the vast majority of patient-clinician encounters take place—posing a particular challenge to the nascent National Patient Safety Agency.5

    Patient care in the community is becoming increasingly complex. Early discharge from hospital, the prescribing and monitoring of potentially dangerous drugs such as methotrexate for rheumatoid arthritis, the pressure of short consultations, and the increasingly fragmented nature of primary care services all increase the risk of unintentional patient harm.

    There are, however, two advantages enjoyed by primary care. Firstly, practices are small organisations with fewer layers of bureaucracy than most hospitals. Implementing systemic changes is thus likely to prove easier than in hospitals. Secondly, the strong tradition of multidisciplinary teamwork in many practices is an important component in creating the right cultural environment for safer care.

    What then are the priorities for developing a patient safety agenda for primary care? Research, action, and leadership.

    Research is urgently needed to identify the commonest forms of patient harm in primary care and their underlying causes. This will require both epidemiological studies, from which it should be possible to create a typology of harm for primary care, and qualitative research designed to understand the contributory systems failures that predispose to such problems.6 Building a safer NHS guarantees that money to fund such research should be available.

    The need for “more research” can serve as a convenient smoke screen for inaction, but this must not be allowed to happen. We already have some information on how patients are being harmed. Defence organisations hold databases of the commonest errors that come to litigation, patient complaints can offer insights, and all of us will have experienced something going wrong with a patient's care that we never acted on. Using analytical tools such as significant event audit7 or those developed by Vincent and colleagues8 will offer an understanding of how general practice systems can fail patients. Informal estimates put the number of practices using significant event audit in the UK at around 20%. The beneficial cultural effects of analysis using a safe and supportive framework should not be underestimated.

    Nevertheless, we should also try to find areas for immediate action. Given the overwhelming feeling of being swamped that pervades primary care at present, these should be simple and preferably not include a large amount of work. For instance, three actions that could be done tomorrow in every surgery are: ensuring that messages are taken in a safe manner through the use of message books; placing sharps boxes on a shelf, out of the reach of children; and identifying patients who do not attend for their warfarin checks so that they can be offered safer alternatives such as aspirin. As a result, and within months, lives may be saved. An important first step would have been made.

    Send us reports of your errors

    We encourage readers to send us accounts of errors that they have made. These accounts should be of not more than 400 words and will be published as fillers. We must know the names and addresses of authors, but we will be willing to publish some anonymously. Authors should, however, sign the pieces whenever possible. Patient consent will be needed.

    The BMJ has a long history of publishing on medical error, and we think of our Lessons of the Week as “cock ups of the week.” The original idea behind lessons of the week was that we learn so much from our own errors that it would be good if we could learn just a little from the errors of others. But we also want to publish accounts of errors in order to help encourage a culture where we can all admit to error, give some idea of the range of errors, and sometimes—as today (p 615)—provide ideas on how such errors might be avoided.

    Lastly, we need strong leadership. Leaders of professional bodies must put safety high on their agendas. Chief executives and chairs of primary care trusts and groups should be lying awake worrying about patient harm and should ensure that improving patient safety is one of their priorities. Promoting a long term shift in culture also requires a major rethink of the way in which medical education is delivered. One small but very visible step would be for educational leaders to introduce the subject of error prevention and patient safety into undergraduate and postgraduate medical curriculums and examinations.

    Leaders need to emphasise that it is not individuals who make mistakes but systems that fail. Certainly, when misconduct has occurred individuals should be admonished or punished. But when someone reports that they have made an error or reports a risk they should be supported. In the airline industry a pilot who reports an error is immune from disciplinary action.9 Most importantly, the person who reports the problem should see the system leap into action. Leaders in primary care need to ensure a mandatory reacting system, not just mandatory reporting system.


    • TW was supported by the Commonwealth Fund, a private independent foundation.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    View Abstract