Patients and medical power

BMJ 2001; 323 doi: http://dx.doi.org/10.1136/bmj.323.7310.414 (Published 25 August 2001) Cite this as: BMJ 2001;323:414

Shifting power in favour of the patient may not be so straightforward

  1. Richard Canter, consultant surgeon
  1. Royal United Hospital, Bath BA1 3AG

    After the announcement of the Alder Hey report into retained organs after postmortem examinations the health secretary, Alan Milburn, said he wanted the balance of power to move away from doctors and “shift decisively in favour of the patient.” He suggested that patients should be offered free choices to make up their own minds about treatment options. Many assumptions underpin such an idea, including the belief that medical “power” is a currency that can be transferred from doctor to patient, so that the patient becomes “free” to choose. It is not as simple as this.

    Power is the means by which A gets B to do something, and many models exist to explain how A might do this. The medical profession sees itself, and is seen, as expert in understanding and managing disease. However, expert power is only part of the picture. The concept of power also includes reward, coercive, legitimate, and charismatic power.1 Individual clinicians may exercise elements of each of these types of power in different proportions so that each has his or her own distinctive consultation style.

    Perhaps Mr Milburn really means that the medical profession should become more charismatic and less coercive in its approach to patients. At first sight this may seem more desirable in a consumerist society, but coercive power is easy to recognise while charismatic power is not. Charismatic doctors may do great harm to a population of patients who remain grateful to them. Perhaps the full report on the murderous general practitioner Harold Shipman will illustrate this point.

    Current thinking is that the analysis of power is far more complex than this simple model suggests. Lukes has proposed a threefold description of power that might be relevant here2: first dimensional power, in which A forces B to do something; second dimensional power, in which A controls the agenda in any interaction with B; and third dimensional power, in which A controls the world as B sees it.

    First dimensional power is akin to the coercive power in the older model. Such power might be exercised appropriately in emergencies, in certain acute psychiatric states, or when patients are extremely distressed or anxious and do not want to debate choices. Second dimensional power is exercised when the conversation is deliberately steered away from or towards certain topics that influence the outcome. The advantages of a treatment, or its complications, may be emphasised in a way that influences the patient's apparent choices. Exercising power in this way also includes the power to silence through lack of clinical time, poor ambience, poor listening skills on the part of the doctor, and a host of other ways that make it difficult for patients to ask questions in a clinic.

    Many doctors might argue that they do not operate in either of these ways but simply “present the facts to patients in an unbiased way” and allow them to make their own choice. This is how they would shift power decisively in favour of the patient. Lukes' model would suggest, however, that A has constructed for B a worldview of disease and treatment, in which B believes he or she moves autonomously but in fact B's actions are shaped by the flow of medical knowledge (supplied by A) that underpins these choices. Similarly, A may believe that B moves autonomously, because both may be the subject of a world view of another agency.

    Consider the situation faced by a patient with malignant disease of the larynx who has to make choices about the treatment options of palliation, radiotherapy, chemotherapy, surgery, or combinations of each. The encounter begins by the clinician setting the clinical problem in a conventional biomedical model. A discussion will follow looking at the effectiveness of each of these treatment options based on the physician's understanding of results reported in the medical literature together with his or her experience of previous cases. The patient might then make a “choice,” but the whole encounter is located firmly within a conventional biomedical framework where there is no place for other frameworks, such as herbal medicine, acupuncture, osteopathy, or other complementary medical alternatives. Indeed, in this example even the mention of alternative medicine may strike many readers as odd. Because third dimensional power is so pervasive and bound up with knowledge itself, it is difficult to recognise. Both A and B may be largely unaware that their worldview is shaped by the prevailing paradigm.

    The reasoning then is that all forms of knowledge, including medical knowledge, produce images of the world that then operate as if they are true.3 There is therefore a powerful argument that when the exercise of clinical power shifts from crude, but easily recognisable, coercive or first dimensional power to the more subtle and harder to recognise third dimensional power, the reality is that nothing may have changed.

    What then can be done to encourage a more “patient centred” healthcare system? Firstly, a little more honesty on the subject wouldn't go amiss. Healthcare professionals must recognise that they do not hold a privileged position from which they alone recognise all medical truths. Medical paradigms come and go, and medicine often develops new paradigms to support the continuation of certain practices when faced with conflicting data.4 Secondly, the argument seems to be about the power model appropriate to the clinical situation. Perhaps healthcare workers should be taught to keep some basic frameworks in mind to enable them to be more sensitive to power in all its manifestations. At the very least a debate that goes beyond the rather naive idea that power should be “handed over” needs to begin, for at the heart of this proposal is the very nature of medical knowledge itself.


    1. 1.
    2. 2.
    3. 3.
    4. 4.