Intended for healthcare professionals

Education And Debate

Refining and implementing the Tavistock principles for everybody in health careCommentary: Justice in health care—a response to Tavistock

BMJ 2001; 323 doi: (Published 15 September 2001) Cite this as: BMJ 2001;323:616

Refining and implementing the Tavistock principles for everybody in health care

  1. Don Berwick, presidenta,
  2. Frank Davidoff, former editor, Annals of Internal Medicineb,
  3. Howard Hiatt, professor of medicinec,
  4. Richard Smith, editor (editor{at}
  1. a Institute for Healthcare Improvement, 375 Longwood Avenue, Boston, MA 02215, USA
  2. b Wethersfield, CT 06109, USA
  3. c Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA
  4. d BMJ, BMA House, London WC1H 9JR
  5. Centre for Health Services Research, University of Tennessee, Memphis, TN38163, USA
  1. Correspondence to: R Smith

    The ethicist Will Gaylin argued that healthcare reform often fails because it attempts technical solutions to ethical problems.1 Agreeing with this position, the Tavistock Group tried to develop ethical principles that might be useful to everybody involved in health care.26 They were intended for those who are responsible for the healthcare system, those who work in it, and those who use it. This article describes the origins of the principles, discusses the thinking behind them, considers how they might be used, provides case studies, and reflects on where the venture might go now.

    Summary points

    The problems of health systems are in the last analysis ethical. Who will live, who will die, and who will decide and how?

    The ethical codes of individual professions may be divisive rather than helpful

    The Tavistock principles are intended to provide an ethical compass for all those involved in health care, including patients and owners of health systems

    The principles cover rights, balance, comprehensiveness, cooperation, improvement, safety, and openness

    Experience of using the principles is limited and not always encouraging; research is under way on how to implement and “validate” them

    Origins of the principles

    The idea that it might be useful to develop ethical principles for everybody involved in health care stemmed from the recognition that much of health care is multidisciplinary yet ethical codes usually cover only one discipline.2 The codes may thus be used as ammunition in interdisciplinary battles rather than as tools to think about deep problems. We advanced the idea of developing ethical principles for everybody in the BMJ in 19972 and then convened a group to develop some principles. The Tavistock Group, a collection of people with long experience of health care and ethical debate, developed the principles, which they published in 1999. 3 4 The principles are not evidence based and have not been validated in any scientific sense. We offer them with humility as something that might be useful, but which, like any innovation, could conceivably do more harm than good.

    We sent the first draft of the principles to many healthcare organisations in the United States and the United Kingdom, inviting a response. The principles were refined in response to the feedback. (The group restricted itself to those two countries, largely because of its predominantly Anglo-American membership, although people from other countries may want to become involved.)

    A meeting of about 150 invited people was held in Cambridge, Massachusetts, in April 2000 to debate each principle and to consider how the principles might be used. The meeting also heard from some US institutions that had tried using the principles. After the meeting the principles were distilled further and published again.5 As the debate has intensified and deepened, the principles have become shorter. Box 1 lists the seven principles.

    Box 1 : The Tavistock principles

    Rights—People have a right to health and health care

    Balance—Care of individual patients is central, but the health of populations is also our concern

    Comprehensiveness—In addition to treating illness, we have an obligation to ease suffering, minimise disability, prevent disease, and promote health

    Cooperation—Health care succeeds only if we cooperate with those we serve, each other, and those in other sectors

    Improvement—Improving health care is a serious and continuing responsibility

    Safety—Do no harm

    Openness—Being open, honest, and trustworthy is vital in health care

    The thinking behind the principles


    This principle causes more difficulty than any other, particularly in the United States. What does it mean, to say that health care is a right when 40 million people in the United States and most of the world's population do not have access to health care? And isn't it even more absurd to argue that people have a right to health?

    The argument over rights extends back to the 18th century, with Tom Paine arguing for them and Jeremy Bentham arguing that they were “nonsense on stilts.” For every rights holder, argued Bentham, there must be an obligation provider. But where is that provider in the case of health care or health? Rights, he argued, are not “in nature” but need institutions and legislation to make them real.

    Immanuel Kant distinguished between “perfect” and “imperfect” obligations. Perfect obligations impose a duty on particular people and institutions, whereas imperfect obligations do not. In many countries health care has become a perfect obligation (for instance, in Britain, where the government has accepted the duty to provide health care), although it remains an imperfect obligation in others. But imperfect obligations can move—perhaps through legislation—to become perfect obligations.

    Amartya Sen, Nobel prize winning economist and master of Trinity College, Cambridge, England, explained to the meeting in April last year why it was not absurd to make health and health care a right. By making health and health care a right, he argued, we gain people's attention: a debate begins on who might have the duty to try to achieve health and health care for everybody. There is a pressure to begin implementation. And it's important also to make health a human right because the main health determinants are not health care but sanitation, nutrition, housing, social justice, employment, and the like.

    Any institution adopting the Tavistock principles would be accepting the imperfect obligation to bring health and health care to everybody. That would create a tension in institutions that provide care only for those who can afford it, but the tension ought to be creative. Those working in such institutions would search for opportunities to make health and health care more broadly available, even in countries beyond their own. And there would have to be deep discomfort around actions that restricted access to health and health care.


    Many of those at the April meeting wanted “and” rather than “but” in this principle (box 1). They hoped to escape the tension between caring for individuals and populations. Sometimes no tension exists, but often—particularly with resources—there is tension. Resources devoted to one patient will be denied to another, or they will be denied to an enterprise that might promote public health. Antibiotics might bring benefit to individuals with mild infections while harming public health by increasing microbial resistance.

    This principle calls on institutions and individuals within them to think beyond individuals to populations. Yet many people's work in health care does not go beyond the care of individuals. Paul Farmer, director of the programme in infectious disease and social change at Harvard, cited the most extreme denial of this principle—the worldwide neglect of destitute sick people.


    This principle is important for understanding the continuum of health care, said William Foege, professor of international health from Emory University, in Atlanta, Georgia. A nurse trying to help an adolescent to stop smoking is already treating a disorder, while a surgeon treating a patient with lung cancer must be both a surgeon and a counsellor. Practitioners can easily think of themselves as providing simply a technical service. We must strive to be both specialists and generalists.


    This principle might again seem like a truism, but following the principle might lead to profound change. Even within health care there are struggles among different groups—managers and doctors, nurses and doctors—and the debate over health care, particularly in the United States, is often dominated by blame. And, despite the vogue for patient partnership, patients often feel like the recipients of care rather than partners in a process of healing.

    This principle is in many ways at the heart of the Tavistock principles. It recognises that all those who work in health care depend on each other, on patients, and on those outside health care—for example, politicians, researchers, and social workers. Pulling out the principle in the middle of a bitter dispute in a hospital might prove extremely useful.

    Jo Ivey Boufford, dean of the Robert F Wagner School of Public Service at New York University, argued that “cooperation” was too weak a word. She wanted recognition that patients are “coproducers” of health and supported the notion of “nothing about me without me”—in other words, practitioners would not make decision about patients without their direct involvement, and health authorities would not make policy decisions without public (and professional) involvement.


    This principle means that it isn't good enough to do well. We must aspire to do better, recognising the escalating rate of new knowledge, the rapid advances in technology, that patients want to be partners, and that our healthcare systems are too complex, giving too much room for error and waste.

    Being serious about improvement (rather than simply paying lip service) means learning the skills of improvement, being willing to accept and even encourage change, and recognising that improvement is never ending. Most health professionals have not mastered the improvement skills, and many resist change.

    Maureen Bisognano, executive vice president of the Institute for Healthcare Improvement, Boston, said that health care suffers simultaneously from overuse, underuse, and misuse of interventions. Problems of service and access abound. In other words, there is huge room for improvement.


    The working draft of the Tavistock principles published in 1999 had only five principles.24 This sixth one was added as a result of consultation (and the seventh was added after discussion at the April meeting). Initially there was anxiety over “do no harm” because it is so strongly associated with doctors. But it seemed important to include because there is increasing recognition of just how much harm healthcare systems produce and of how policies with benign intentions can create harm.

    “Do no harm,” however, is impossible to achieve, pointed out Uwe Reinhardt, professor of political economy at Princeton University, and John Eisenberg, director of the Agency for Healthcare Research and Quality in the US Department of Health and Human Services. All effective interventions may harm, but the intention behind the principle is not that practitioners should never make an intervention; it is that they should struggle to maximise benefit, minimise harm, and reduce error.

    Box 2 : Denying patients a new treatment

    A doctor working in an NHS trust thinks it wrong that his patients will be denied a new treatment for cancer (the hospital formulary committee had decided that it should not be prescribed). Should he contact the local media? Should the trust punish him if he does?

    The “balance” principle recognises that a tension exists between what is good for individuals and for populations. It was probably on these grounds that the committee decided that the new drug would not be made available. The “cooperation” principle suggests that the doctor should cooperate with his colleagues and implies that contacting the media would not be helpful. But the “openness” principle means that the committee should be open with patients, doctors, and the community (through the media perhaps) on why it is denying patients a drug. The doctor might decide that the hospital is not living by the openness principle and so contact the media himself. If he does that, he should abide by the openness principle and give the whole story, not just his version. If the trust has lived by the principles and the doctor has not, then it might be legitimate to punish him. It clearly would not be legitimate if the doctor lived by the principles but the trust did not.



    This last principle might be both the most banal and the most profound. Nobody could argue against being open, honest, and trustworthy, and yet every day in every healthcare system people fail on all three counts. It's difficult to be open and honest about deficiencies in your hospital or practice. There's always a way to “soften the blow” or “be economical with the truth.” You worry that you might lose the trust of patients or the public, yet nothing destroys trust faster than being found to have deceived.

    Experiences of using the principles

    The April meeting heard from three US institutions that had tried to use the principles. Tom Hale from the Unity Medical Group in St Louis described how they had tried using them in their outpatient setting, which covered 125 000 patients in 80 locations. They had planned to gain acceptance of the principles, integrate them into their operations, communicate them to patients, quantify and measure outcomes, benchmark their practice against the principles, and then use feedback to assess the process. In fact they became stuck with the first principle (rights): the doctors thought that profit was important and saw this as being in conflict with health care being a right. They were also worried by the abstractness of the principles and the fact that they were not evidence based. The most important change, concluded Hale, was that they were thinking about the principles.

    Box 3 : Are the costs of improvement excessive?

    A health maintenance organisation in the United States considers investing in improvements in its system for caring for patients with AIDS. The vice president for marketing warns that such improvements may lead to selective enrolment of unprofitable members—namely, those with HIV infection. Is the organisation ethically bound to improve its HIV care, even if that may reduce its financial viability?

    The “improvement” principle states that improvement is a serious and continuing responsibility. The “balance” principle recognises the tension that may exist between the needs of individual patients and those of the population, and this principle should be considered if the investment might threaten services to other patients. The “safety” principle suggests that it would be wrong to retain a deficient system because avoidable harm could result. The “rights” principle means that it would be poor behaviour to seek to deny the right to health care by avoiding changes that might attract more patients. According to the principles, it would be wrong not to make the investment.


    Nancy Boucher described how the management of the Crozer-Keystone health system in the United States had found that the principles had helped them with some difficult decisions, including whether they should disenfranchise some patients to avoid a financial loss. The management worried, however, that if it adopted the principles and competitors did not, then “the playing field would not be level.” Would competitive advantage come from adopting or not adopting the principles? A survey of staff found that they were mostly positive about the principles but worried about having the resources to implement them.

    The board of Avera Health had debated the principles twice and found them “helpful but too broad and too vague,” reported Jean Reed. They did help different groups to talk together, and the board thought that they might be particularly useful when it came to decisions on forming joint ventures (you would feel more comfortable forming a joint venture with another organisation that had adopted the principles).

    Validation of the principles

    How might we validate the Tavistock principles? We find this a difficult question. Have the Hippocratic oath or the Ten Commandments been validated? They have perhaps been validated in that they have been widely adopted and in some countries incorporated into law. But they have not been validated in the way that new treatments or policies might be validated—through the use of randomised trials or similar methods. It might be conceivable to validate the Tavistock principles by randomly assigning them to be used in different institutions and then measure performance outcomes, including staff motivation. But such an experiment would be difficult and expensive. It seems more sensible to validate the principles by seeing if any institutions adopt them and find them useful.


    Making strategies and principles change anything is difficult. David Garvin, professor of business administration at Harvard, thought that it would be important to build up a series of cases in which the principles were applied. This would make them more concrete and operational. Boxes 2, 3, and 4 describe some such cases, and several more cases are available on the BMJ's website ( Some of these cases were described in the original editorial about developing principles.2 (We invite readers to submit further cases, preferably including an analysis using the principles.) Garvin also suggested the creation of a user's guide: a first draft is available on

    Box 4 : A nurse with HIV infection

    Managers of a health provider discover that one of their nurses is infected with HIV but has told nobody. Should they release the nurse's name to the media? Should they notify all those who may have been treated by the nurse even though the chances of anybody being infected are vanishingly small?

    Four principles (balance, comprehensiveness, safety, and openness) suggest that the media and patients should be fully informed. The “cooperation” principle may be taken to mean that the nurse's name should not be released without her consent. If the nurse did not consent, judgment would have to be made about releasing the name, but the weight of the principles suggests it should be disclosed. The principles favour disclosure.


    Kenneth Roth, executive director of Human Rights Watch, New York, considered how the Tavistock Group might learn from the successes and failures of other codes. One way to implement a code is to incorporate it into law. This is unlikely to happen with the Tavistock principles, and the American experience of trying to create a legal bill of rights for patients is not encouraging. An alternative code is an aspirational code, which assumes good faith on the part of those trying to live by it. Such codes are often formally ratified. Those who adopt the principles are expected to live by them, but there are usually no teeth. The Tavistock principles are essentially an aspirational code.

    The meeting struggled with implementation, but proposals emerged and are described on The Tavistock Group has three broad strategies: to publicise the principles and let people do with them what they will; to do more work to encourage the adoption of the principles, mainly through being opportunistic; and to try to raise money, employ some staff, and be energetic in implementing the principles. For now, the group is following the first two strategies. But we would be delighted if anybody wanted to take the lead in pursuing the third.


    • Funding Kellog Foundation, Robert Wood Johnson Foundation, and the American Academy of Sciences.

    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.

    Commentary: Justice in health care—a response to Tavistock

    1. David M Mirvis, director (dmirvis{at},
    2. James E Bailey, associate professor
    1. a Institute for Healthcare Improvement, 375 Longwood Avenue, Boston, MA 02215, USA
    2. b Wethersfield, CT 06109, USA
    3. c Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA
    4. d BMJ, BMA House, London WC1H 9JR
    5. Centre for Health Services Research, University of Tennessee, Memphis, TN38163, USA

      In January 1999 the Tavistock Group proposed five “ethical principles for those who shape and give health care” to better govern and manage the healthcare system. 1 2 These principles unite diverse groups behind a publicly articulated philosophy that emphasises what they have in common and provides an inviolate set of standards for measuring our individual and group commitment and performance.3

      The Justice in Health Care Foundation has also produced basic principles for health care.4 Our group, composed of healthcare consumers, practitioners, academicians, and health system leaders, seeks to emphasise the primary importance of the consumer—the patient—in the healthcare system. Our premise is that health systems will change only if educated and empowered consumers actively and responsibly assert their needs and assertively and strategically insist that providers and systems meet these needs.

      We derived our principles from a different perspective from that of the Tavistock Group. The Tavistock Group began from the perspective of the provider and evaluated the behaviours that members of the healthcare delivery system should exhibit. We started from the perspective of the consumer and assessed the characteristics of a system that would best meet consumer needs.

      This difference is fundamental. Healthcare systems may and often do have purposes and driving forces that differ from those of patients. We believe that consumers should be included directly in the functions of healthcare systems rather than be viewed as external objects of such functions. This empowerment leads not only to rightful demands but also to potent obligations on the part of consumers.

      Both sets of principles, however, are remarkably similar (box). We agree that the primary goal of a healthcare system is to improve health and that care for sick people should be provided according to need as a societal obligation.

      Principles of the Justice in Health Care Foundation4

      Health—Health systems should pursue health as their primary goal

      Access—Health systems should provide care primarily according to need rather than ability to pay

      Accountability—Consumers, providers, and healthcare institutions must take responsibility for health and healthcare resources with which they are entrusted

      Choice—Consumers must have the real ability to choose their healthcare systems, providers, and treatments in order to seek the best value in health care for themselves

      Education—Education of consumers, providers, and institutions regarding value and quality in health care is necessary for responsible and informed health choices

      We emphasise two additional concepts—choice and accountability—that reflect our consumer focus. By choice, we mean choice of health plan, of provider, and of care—as a right of autonomy and self determination. For healthcare decisions to be effective, they must be personal decisions that reflect personal values as well as medical knowledge.5 Accountability for the consequences of healthcare decisions is a cornerstone principle. It includes responsibilities of consumers as well as of payers and providers. Taking personal responsibility for our own health, as consumers, means paying according to our means; prudently and appropriately using limited healthcare resources; adopting health promoting behaviours; continually learning about important health issues; and actively participating in decision making about our own health as well as that of the community. This draws consumers into the healthcare system as partners, not just as payers or subjects of care.

      By including and preferably emphasising the legitimate needs and obligations of the consumer, we can devise and implement what the Tavistock Group referred to as “a clear, strong, and reasonable set of principles for conduct that all [authors' emphasis] stakeholders who give or shape health care can recognise and accept as guides to correct action.”1 Who has more at stake than the patient?


      DMM is a member of the board of directors and JEB is the president of the Justice in Health Care Foundation, in Memphis, Tennessee.


      • Competing interests None declared.

        Embedded Image Case studies, a users' guide, and a report on implementation are available on the BMJ's website


      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.
      View Abstract