Thursday, 1 January 2015

Is Evidence-Based Psychiatry Ethical?

Is Evidence-Based Psychiatry Ethical?

This post is by Mona Gupta, psychiatrist and bioethics researcher at the University of Montreal in Canada.

For the last 15 years, I’ve been working on ethical issues relating to the use of evidence-based medicine (EBM) in psychiatric practice. EBM has been enormously influential in clinical medicine, particularly in psychiatry. This intersection between ethics, EBM and psychiatry is the theme I develop in my new book, Is Evidence-Based Psychiatry Ethical?, published by Oxford University Press in 2014.

EBM is a phrase that first appeared in the medical literature in the 1990s. It promotes a seemingly irrefutable principle: that clinical decision-making should be based, as much as possible, on the most up-to-date research findings. Nowhere has this idea been more welcome than in psychiatry, a field that continues to be dogged by a legacy of controversial clinical interventions. Many mental health experts believe that following the rules of EBM is the best way of safeguarding patients from unproven fads or dangerous interventions. If something is effective or ineffective, EBM will tell us.

According to this viewpoint, EBM not only offers a more solid scientific basis to clinical practice than what came before it, it is also ethically necessary because it will enable psychiatrists to offer the best care to their patients.

It turns out that ensuring medical practice is based on good evidence is not as straightforward as it sounds. After all, evidence does not come from thin air. There are finite resources for research, which means that there is always someone deciding what topics should be researched, whose studies merit funding, and which results will be published. These kinds of decisions reflect the beliefs and values of policymakers, funders, researchers, and journal editors about what is important. And determining what is important depends on one’s goals: improving clinical practice to be sure, but also reaping profits, promoting one’s preferred hypotheses, and advancing one’s career. In other words, what counts as evidence is partly determined by values and interests.

In this book, I argue that the idea that values affect what counts as evidence is a particularly challenging problem for psychiatry. Since its inception as a distinct medical discipline, it has suffered from the criticism that it is not sufficiently scientific. To laypeople and doctors alike, a broken leg is viewed as a fact, but whether someone is mentally normal or abnormal is perceived as a value judgement. There is a hope amongst proponents of evidence-based psychiatry that EBM can take this normative component out of psychiatry, but it cannot. First, EBM is itself imbued with the values involved in the process of knowledge production. Second, we must still make value judgements that certain states of mind are abnormal. Furthermore, addressing such problems in a way that reduces suffering also requires the consideration of ethical values. The consideration of these values extends well beyond what is offered by EBM’s model of practice.

Increasingly, good practice is becoming synonymous with evidence-based practice. The book concludes by considering what ethical psychiatric practice looks like and what, if anything, EBM contributes to that vision.


  1. Interesting post - thanks!

    In the paragraph starting 'It turns out that...', you seem to move from the claim (1) that values & interests will determine which topics get researched, to the claim (2) that values & interests affect the evidential status of research findings. To say that V&I influence funding decisions and research agendas is *not* to say that "what counts as evidence is partly determined by values and interests".

    Am I misunderstanding something? If not, how does the former (unobjectionable, sociological) claim support the latter (controversial, epistemological) claim?

  2. Thanks so much for this comment. You are right that I am trying to argue that values and interests can affect what counts as evidence, not just issues external to the evidence itself (such as what topics to research). An example would the selection of the appropriate level of uncertainty to accept for statistical testing. There is the convention that is used in clinical trials but this could be different if the researchers decided that a different level of uncertainty was appropriate. What goes into that decision includes how much uncertainty they will be willing to tolerate in a given study. If a clinical trial was looking at a drug for a grave illness with few treatment options, the researchers might agree that they should accept a higher level of uncertainty than the convention in determining whether the medication is effective. Part of what informs that judgement are their own values about the seriousness of the condition in the lives of the people affected by it.


Comments are moderated.