Article Text


Response to Dr Gupta
  1. Peter Szatmari, MD
  1. Editor, EBMH

    Statistics from

    It is not uncommon for proponents of evidence-based practice (EBP) to state that once a treatment has been found to be ineffective in a randomised control trial it is “unethical” to continue to practice it. Dr Gupta challenges this statement. She feels there is no justification for stating “we should practice evidence-based mental health because it is ethical”. She buttresses her argument by stating that since there are so many concerns about the “truth” of the evidence provided in scientific studies that one should not base the practice of EBP on that foundation.

    It is an interesting point but I think it may be dangerous to caricature EBM. I do not recollect any statement by an advocate of EBP saying that the evidence in a randomised controlled trial is the same as “truth”. Indeed, I think that what we learn by practicing EBP is rather “the error of our ways”. I would admit however, that many advocates of EBP would agree that as the evidence accumulates and insofar as it is consistent (as demonstrated in a meta-analysis), empirical evidence becomes a closer and closer approximation to the truth. It is the “best available evidence”, but it will never be the truth itself. Nobody who practices in this way should believe that the truth is a categorical phenomena; that is, true or untrue. Rather, it is “more” or “less” true.

    Dr Gupta challenges us not to practice EB mental health because it is ethical. I would ask then, on what basis do I make a clinical decision? If I see a young child with autism, I have a choice: I can refer him to behaviour therapy or not. I would prefer to make my clinical decision on the basis of the best available evidence. What is the alternative to using evidence as a guide? I could do nothing, but that would be clinical paralysis and that surely is unethical. I could make a decision at random by simply flipping a coin, but that does not feel like an ethical thing to do if there is existing evidence (although randomisation in an N of 1 trial may be the best solution if there is inadequate evidence). I could make my decision on the basis on how I was trained. But as I was trained many years ago, that information is now quite out of date. I could instead make my decision based on what I know about the pathophysiology of autism (just like many physicians make clinical decisions about prescribing medication for depression on the basis of their knowledge of neurotransmitters in mood disorders). But as what we know about pathophysiology of autism is so limited (and, I would argue, is equally limited for all psychiatric disorders), I think it is very difficult, if not impossible, to make clinical decisions about treatment based on our very incomplete knowledge of aetiology. Finally, I could make my clinical decision based on my values. What values do I hold about the most appropriate, the most humane, the most empowering form of treatment? In fact, I personally do not find applied behavioural analysis (ABA) very humane. The use of massed discreet trials to teach simple tasks such as matching often looks intrusive and critical. I would much prefer a more developmental sociocognitive approach and indeed such treatments are available but they do not yet have the evidence to support them. But if I were to choose a treatment based on my values over the evidence, I would have to do so by the rules of informed consent. To be ethical, I would have to make that preference known to the parents of this child with autism and they would also have to choose (or not to choose) my values over the evidence. I wonder what a reasonable adult would do in such a circumstance? I bet they would choose treatment that is supported by the evidence over and above my own values and so would choose behaviour therapy.

    Do we have an alternative? Instead of saying that it is ethical to practice EBM because the evidence leads one to the “truth”, one could say “it is useful to practice EBM”. More people tend to get better when a treatment has been shown to be effective by a randomised controlled trial than when an alternative is employed. In this circumstance, more people get better when they receive the “experimental” treatment than when they receive, say, a placebo or the standard treatment. It is not so much a matter of “truthfulness” as it is of “usefulness”. Whether the results from a particular study are generalisable to my clinical case load is an open question. But choosing a treatment based on its ability to do good for the most people is a utilitarian approach to the truth, not an absolute categorical glimpse of the truth. It is moreover a value laden statement. EBP is ethical because it produces the most good for the largest number of people (whether or not it is also based on any knowledge of the truth). Surely that is a useful way to treat those with severe mental illness.

    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.