Caring with evidence based medicine

BMJ 2016; 353 doi: https://doi.org/10.1136/bmj.i3530 (Published 28 June 2016) Cite this as: BMJ 2016;353:i3530
  1. Ian Hargraves, designer1,
  2. Marleen Kunneman, postdoctoral research fellow1 2,
  3. Juan P Brito, assistant professor of medicine1 3,
  4. Victor M Montori, professor of medicine1 3
  1. 1Knowledge and Evaluation Research Unit, Mayo Clinic, Rochester, USA
  2. 2Department of Medical Psychology, Academic Medical Centre, University of Amsterdam, the Netherlands
  3. 3Division of Endocrinology, Diabetes, Metabolism and Nutrition, Mayo Clinic, Rochester, MN 55905, USA
  1. Correspondence to: V M Montori montori.victor{at}mayo.edu

Using evidence for kind and careful care

The task of evidence based medicine is to care for each patient. The “evidence based” adjective draws attention to the need to explicitly consider clinical research findings in caring for patients.1 In the 25 years since Guyatt first used the term, many tools—such as practice guidelines—have sought to bring evidence based medicine to all.

Three recent articles in The BMJ suggest that something got lost along the way and offer solutions. Yudkin and colleagues note that guidelines and clinicians overestimate the value of healthcare, and they suggest risk estimators to calibrate expectations. 2 Elwyn and colleagues think that guidelines are not built to support how clinicians use knowledge in practice in social, intuitive, and contextual ways and suggest new evidence conduits for the clinical encounter. 3 McCartney and colleagues join them in arguing that guidelines do not support evidence informed, individualised decision making. 4All three articles advocate for shared decision making.

It is thanks to evidence based medicine, however, that we have learnt about the corruption of the evidence and about the flawed production and dissemination of guidelines. From the beginning, a core clinical principle of evidence based medicine stated that the evidence alone should never dictate care for a patient. The authors are right to critique guidelines when they fail to account for the shortcomings of evidence or forget the core clinical principle and fail to contribute to the work of caring for people. And when guidelines do fail, new or adapted evidence based medicine tools should emerge to ensure individualised care for each patient. In advancing these new tools, however, we must be careful not to fetishise them.

No simple answer

Analysing video recordings from our experience in developing and testing aids for use in clinical encounters has been sobering. We have seen sophisticated forms of persuasion in which the decision aid is used as a prop. We have seen encounters in which the technical steps for shared decision making are as evident as the lack of real connection with the patient. Like guidelines, shared decision making and the aids that support it don’t necessarily result in care. When mandated and mechanically deployed, shared decision making does not make evidence care for people. To be successful, shared decision making must be a caring response to a person and their health situation.

Both guidelines and shared decision making fail when they seek to inject certainty rather than do justice to the uncertainty of illness. When we seek certainty in the tools we risk making evidence, clinical practice, and the articulation of patient preferences definitive rather than contributory in caring for the person. By unduly advancing certainty—for instance, when guidelines make strong recommendations—we breed hubris, a sense that medicine is simply a matter of knowing what to do and that all the best clinicians and their patients should have this ability at their fingertips. Anyone who lives with or cares for people with depression, cancer, or heart disease knows that this is rarely the case. The real challenge is to deal with illness and how it fragments, disarticulates, and renders uncertain the conduct and dignity of human lives.

Maintaining a critical eye on the contribution of evidenced based tools to the care of individual patients is the way forward. Guidelines serve well to communicate our understanding of what’s best for most patients when this is clear. In less clear situations, appropriately constructed guidelines are also helpful, as some guidance is often better than none. The GRADE approach, for instance, requires risk calculation and risk presentation in absolute terms, and it is explicit about considering sources of uncertainty in the evidence and its application.5 6

Shared decision making should not be seen as the communication of expert knowledge or preference, or as the joint review of knowledge summaries designed for the consultation. Rather, it should be a thoughtful response to the threat and uncertainty of illness.

In our work, we emphasise shared decision making in conversations.7 The conversation is an appropriate environment to use evidence from clinical research to help distinguish among available options—a set of options perhaps curated in guidelines.8 Options are hypotheses for what is best for this patient now. These hypotheses are tested in conversation until a purposeful path forward becomes clear. It is in conversation that clinicians and patients draw out the importance of evidence, options, biography, purpose, goals, and needs as these contribute to an individualised treatment plan. In so doing, evidence based medicine, as originally intended, becomes careful and kind care.


  • Competing interests: We have read and understood BMJ policy on declaration of interests and have nothing to declare.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract

Sign in

Log in through your institution