Intended for healthcare professionals

Education And Debate

What do we mean by partnership in making decisions about treatment?

BMJ 1999; 319 doi: (Published 18 September 1999) Cite this as: BMJ 1999;319:780
  1. Cathy Charles, associate professor (charlesc{at},
  2. Tim Whelan, associate professorb,
  3. Amiram Gafni, professora
  1. a Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada L8N 3Z5
  2. b Department of Medicine, McMaster University
  1. Correspondence to: C Charles
  • Accepted 19 August 1999

For many decades, the dominant approach to making decisions about treatment in the medical encounter has been one of paternalism.17 In recent years this model has been challenged by doctors, patients, medical ethicists, and researchers who advocate more of a partnership relation between doctors and patients.2 813 The reasons for this challenge have been described in detail elsewhere and include the rise of consumerism and the notion of consumer sovereignty in healthcare decision making; the women's movement with its emphasis on challenging medical authority; the passage of legislation focusing on patients' rights in health care; and small area variations in doctors' practice patterns that seem unrelated to differences in health status.7 Though the first three factors are seen as either contributors to or facilitators of patients' participation in making decisions about health care, the fourth highlights the imprecision or the “art” of medical care14: patients with similar clinical problems may receive different treatments from different doctors, due in part to systematic variations in practice patterns across geographical areas.14 15

The call for doctor-patient partnerships opens up options beyond paternalism for approaching the task of making decisions about treatment. But it also raises new complexities. Because a partnership between patient and doctor can take different forms, it is not intuitively apparent what this model would look like. The Oxford English Dictionary (1995) defines a partner as “a person who shares or takes part with another or others.” This definition leaves several important questions unanswered. For example, does the concept of a partnership imply that both doctor and patient need to share all parts of the decision making process? What is it that both parties are sharing, and to what degree? Who is responsible for determining if a partnership is possible and for initiating the first steps? Is there room for variation in patients' and doctors' roles while still retaining the concept of a partnership? What are the clinical contexts within which a doctor-patient partnership is most needed? Drawing on our earlier conceptual papers.6 7 this article identifies and describes different types of partnerships that can be developed between a doctor and patient in the context of making decisions about treatment.

Summary points

Doctor-patient partnerships in making decisions about treatment can take different forms

Three theoretical treatment decision making models are the paternalistic, the shared, and the informed

Most clinical consultations use elements of these theoretical models, and these may change as the interaction unfolds

Doctors need to be aware of and be able to identify and explain the treatment options available

If doctor-patient partnerships are to be promoted in clinical practice, current disincentives such as time and funding constraints will need to be restructured

Theoretical models

In the figure, treatment decision making is broken down into different analytical steps or stages (though in reality these may occur together in an iterative process): information exchange, deliberation or discussion of treatment preferences, and deciding on the treatment to implement. Information exchange can be further subclassified according to the flow, direction, type, and amount of information exchanged.


Models of decision making about treatment

Paternalistic model

The characteristics of the paternalistic model are well known. In the purest form of this model, the patient passively acquiesces to professional authority by agreeing to the doctor's choice of treatment. An assumption underlying this model is that the doctor will make the best treatment decision for the patient and can do so without eliciting personal information from the patient or involving him or her in the decision making process. There is no sharing of any of the decision making steps in this model so, by definition, a doctor-patient partnership does not exist.

In cases where both the patient and the doctor prefer this approach, it can be argued that they have entered a form of partnership based on agreement about how the process should be undertaken. But this would require an explicit discussion of alternative decision making models and an explicit agreement by the doctor to adopt the patient's preferred model Neither of these are likely to occur if the doctor adopts a paternalistic model at the outset of the decision making process.

Informed model

The informed model involves a partnership between doctor and patient that is based on a division of labour. In the information exchange stage, for example, the doctor leads and the communication is one way, from doctor to patient The doctor communicates to the patient information on all relevant treatment options and their benefits and risks. The amount and type of information communicated includes, at a minimum, sufficient information to enable the patient to make an informed treatment decision.

Information transfer is seen as the key responsibility and only legitimate contribution of the doctor to the decision making process. This model assumes that the deliberation and decision making steps are the sole prerogative of the patient. The doctor is accorded no legitimate claim for involvement in these phases or legitimate investment in the treatment decision that the patient makes. Eddy, a doctor who advocates this approach, states the case clearly: “The people whose preferences count are the patients, because they are the ones who will have to live (or die) with the outcomes. … Ideally, you and I are not even in the picture. What matters is what Mrs Smith thinks. … It is also quite possible that Mrs Smith's preferences will differ from Mrs Brown's preferences. If so, both are correct, because “correct” is defined separately for each woman. Assuming that both women are accurately informed regarding the outcomes, neither should be persuaded to change her mind.”8

Shared model

In contrast to the informed model, the essential characteristic of the shared model is its interactional nature in that the doctor and patient share all stages of the decision making process simultaneously.6 7 In its purest form, there is a two way exchange of information, both doctor and patient reveal treatment preferences, and both agree on the decision to implement. This approach assumes that both the patient and the doctor have a legitimate investment in the treatment decision; hence, both declare treatment preferences and their rationale while trying to build a consensus on the appropriate treatment to implement. The challenge for doctors using this model is to create an environment in which the patient feels comfortable expressing his or her treatment preferences (including doing nothing or watchful waiting).

Intermediate approaches

The figure also indicates that there are intermediate approaches. Doctors in a clinical encounter may not use a pure model, but rather a hybrid of elements of more than one model.

The reality of clinical decision making

In the real world of everyday practice, many clinical decision making interactions are likely to reflect some form of hybrid model. Human behaviour rarely conforms to ideal types. Given the often dynamic, complex, and personal nature of the doctor-patient relationship, the decision making model adopted at the outset of a clinical encounter may not unfold according to the pattern of the ideal type but instead be modified to reflect the needs of individual patients

A doctor, for example, who favours a shared decision making model (by our definition) may find, as the interaction proceeds, that the patient has gained enough confidence and gathered enough information to make the decision on his or her own. At this point, the process might shift from a shared to more of an informed model as a result of the learning that has occurred in the interaction itself. Given this possibility, assessing patients' preferences for participation in making decisions about treatment is perhaps more appropriately conceptualised as an ongoing task for the doctor, rather than an assessment made, either implicitly or explicitly, at the outset of the encounter.

Clinical contexts for partnership relationships

The importance of developing some form of partnership in making decisions about treatment depends on the specific clinical context as well as patients' and doctors' preferences. When patients bring minor everyday problems (such as a rash) to the doctor and there is a routine and simple treatment of benefit, there is no compelling clinical need to initiate a process of eliciting patients' preferences about how the decision making process should proceed, especially when the time available for each interaction is limited.

The need for some form of doctor-patient partnership is most compelling when the patient presents with a serious or life threatening illness; different treatment options exist, with different benefits and risks; and outcomes are uncertain. In this situation, the stakes are high and there is no one “right” treatment. Since the patient will bear the consequences of whatever treatment is implemented, it is important that his or her values and preferences are known and respected. Patients in this situation are likely to feel vulnerable and may not initiate such a discussion; it is the doctor's responsibility to ensure that this occurs.

In current healthcare systems, both time and funding constraints can act as disincentives for doctors to explore and respond to patients' preferences regarding the type of partnership they would prefer in the process of making decisions about treatment. Both of these potential barriers are amenable to policy change if the political will to do so exists. Whether current disincentives are restructured into incentives depends on the priority given to the goal of facilitating doctor-patient partnerships in making decisions about treatment relative to other competing clinical and health system goals.


Funding: Funding for research on shared decision making about treatment has been provided by the Canadian Breast Cancer Research Initiative of the National Cancer Institute of Canada, and the Ontario Ministry of Health, through its system linked research program. We gratefully acknowledge this support.


  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
View Abstract