Interviews

Evidence-Based with Dr. Gordon Guyatt

Arnav Agarwal, MD Candidate, Faculty of Medicine, Leyla Eryuzlu, Faculty of Health Sciences, McMaster University

r. Gordon Guyatt is a Distin- clinicians need to understand the principles of us- guished Professor in the De- ing materials, but the last decade has been more of Dpartment of Clinical Epidemi- a focus on getting optimal evidence summaries and ology and Biostatistics, a Joint Member recommendations to the people who use the mate- of the Department of Medicine, and rial: the clinicians. So, that work has been focused a Member of the Clinical Advances on guideline developers. Furthermore, perhaps the through Research and Information most important thing I have done involves my work Technologies (CLARITY) group at with an organization called UpToDate, which pro- McMaster University. His areas of inter- duces a worldwide electronic medical textbook. It est include dissemination of evidence- has become exceedingly popular, and while its pen- based medicine concepts to workers etration outside educational institutions is unclear, Dr. Gordon Guyatt and consumers in health care, ascer- it is the most popular tool used within educational tainment and assessment of patient institutions and with trainees. So that’s an example of values and preferences, systematic review methodology, and the most important stuff I’m doing, which is to help clinical practice guideline and medical decision-making people who produce the materials to guide clinicians methodology. Dr. Guyatt has been cited over 50,000 times and to be optimally evidence-based. published numerous times in leading medical journals such as The New England Journal of Medicine, The Lancet, Jour- UTMJ: What was the state of the practice of medicine before nal of the American Medical Association, and British Medi- your work with evidence-based medicine? cal Journal. Dr. Guyatt’s work has earned numerous awards GG: I’m not the one who started the movement. The and honours, including nomination for the BMJ Lifetime movement started with clinical , which Achievement Award, Officer of the Order of Canada in 2011, was trying to take the principles of traditional epi- and Fellow of the Royal Society of Canada in 2012. Recog- demiology and apply them to clinical practice. It nized as having coined the term “evidence-based medicine” involved using research methods to obtain evidence in 1990, he is a pioneer in the area and has been a leading ex- in which we can be confident about how to guide ponent of the concept and its application to clinical practice. clinical practice. Then, Dr. Dave Sackett decided he would bring this movement into the McMaster Medi- UTMJ: Tell us about yourself and the primary focus of your cal School. It was his vision to train clinicians to read academic work. the medical literature. He began his work in the early GG: Well, one primary focus, and an essential contribu- 1980’s and created a series of papers for clinicians - tion, has been the development of processes for es- a reader’s guide of the medical literature - which he tablishing and understanding the best evidence for and colleagues in McMaster’s Department of Clinical health interventions, and then disseminating this to Epidemiology and Biostatistics published in the Ca- networks of clinicians. Historically, it has been our nadian Medical Association Journal where it made a focus to help clinicians use the medical literature to significant impact. improve their care, and that continues to be a ma- It was a turning point for me as well when, in 1990, jor focus. The methods and education that we have I took over as the Director of the Residency program introduced have been very well-received internation- of Internal Medicine at McMaster University, with a ally. Therefore, this is one stream – to help clinicians mission to advance the skills of the trainees in inter- to use the medical literature. nal medicine in using the medical literature to opti- In addition, as evidence-based medicine has evo- mize their clinical practice. As Program Director, I lved, it has become very evident that as much as we needed to advertise what we were doing. I thought would like to see clinicians be able to read and use what we wanted to teach was a different way of prac- the original articles, doing it well requires more than ticing medicine. the skills of the typical clinician. In the last decade, This brings me to a direct response to your ques- there has been a shift towards getting optimal sec- tion: when I trained for medicine, nobody taught me ondary material to clinicians. We still believe that anything about how to read a journal. When I came

UTMJ • Volume 92, Number 2, March 2015 11 Interviews

Evidence-Based Medicine with Dr. Gordon Guyatt

to McMaster, I participated in one of Dave Sackett’s UTMJ: It seems surprising that it was a new concept in the courses for residents. I completed a 12-week course, medical field. and, at the end of the 12 weeks, I was hoping I could GG: Oh yes. People can’t believe it. I have a friend I went read the medical literature. However, I found I could to high-school with, whom I hadn’t seen for 25 to 30 not; a 12-week course is not enough to do that. years. I ran into him in the elevator in the hospital, It is difficult to sell evidence-based medicine as a just around the time that I received the Officer of the major advance to the general public, who can’t be- Order of Canada. The news had just appeared in the lieve that doctors were not always able to use the med- newspaper and my friend said, “That’s pretty good, ical literature. They really could not at all! In the Officer of the Order of Canada! What did you get year 1990, when I came to the residency program, it it for?” And I told him in a very much abbreviated was the farthest thing from the standard part of train- form, and he said, “Oh boy, if that’s what you got it ing to be able to read and analyze the literature. for, they should give me the Order of Canada!” In What had evolved from 1980 to 1990 was a different other words, he could not believe that teaching clini- way of practice and of thinking about things. All the cians to use the medical literature to guide their prac- physiological rationale and previous educational ap- tice could actually be something acknowledged and proaches were very uncritical in terms of using one’s something that someone would think is important, own experience and opinions to guide practice. The because it should just be obvious! approaches that Dave Sackett and his colleagues were suggesting provided a much higher level of scepti- UTMJ: One of the things that I had interpreted from the cism: an acknowledgement of what we didn’t know, concept was its relation to the medical field and how and reading of the literature to fill in gaps. Essen- EBM impacts medicine itself. What is its relevance to tially, it was a different way of practicing medicine. not only medical practice, but modern-day medical I wanted to teach this different approach of prac- issues? ticing medicine and my idea of what to call this was ‘evidence-based medicine’. At that time, we had a GG: One way to highlight this is when people have gone vision of a new series of users’ guides to the medical wrong as a result of not using a proper method. literature that would teach clinicians how to use the One big story that some people can relate to is what literature in their daily practice, not just read as an people were doing about recommendations regard- academic exercise. We ended up publishing over 30 ing post-menopausal therapy. Years ago, there were papers in JAMA; the series became hugely popular. observations that women who received hormone re- Meanwhile, the term ‘evidence-based medicine’ placement therapy had a significantly reduced car- turned out to be a total hit in terms of being very diovascular risk. In response, this led to a big push rapidly and widely-disseminated. During the ‘90s, across North America for women to receive hormone people became increasingly supportive of its impor- replacement therapy. However, when the random- tance within undergraduate medical education and ized trials were done, they found an increase in breast post-graduate medical education. I didn’t follow it cancer rates and no reduction in cardiovascular risk closely, but my impression is that between the years associated with the therapy. As a result, women had 2000 and 2010, almost every institution in North been receiving the wrong information. Had people America came to view the ability to read and apply applied what we know – the standards that we now the literature as something that is a requisite skill for suggest be used – what they would have said back both undergraduate and postgraduate students. then is: “Maybe it decreases your cardiovascular risk, And it is still evolving: only very recently did the but maybe not. And, by the way, maybe it doesn’t individuals who design the American Licensing Exam increase your breast cancer risk, but maybe it does.” decide EBM should be an important part of the exam, And that is a very different way of putting the infor- and I’m now chairing a committee that is formulating mation. questions for the exam. It is a long process, but here, Another example is with regards to when patients 23 years after the term ‘evidence-based medicine’ have heart attacks and their being at a higher risk of was created, it is now featuring in the U.S. Licens- sudden death partly because of ventricular arrhyth- ing Exam. I would say, in North America now, every mia, which is when the heart starts beating irregularly medical school and every post-graduate program, as and eventually stops. There were drugs that elimi- part of their official curricula, has a component of nated asymptomatic arrhythmias and cardiologists evidence-based medicine. It doesn’t mean they do it thought they could not use these drugs to lower the well. It simply means that it is something that EBM is rate of sudden death. Fortunately, they decided to present and is acknowledged as something students test the drugs in randomized trials, and in doing so, should learn as part of their medical training. they found that the drugs actually increased risk of

12 UTMJ • Volume 92, Number 2, March 2015  Interviews

Evidence-Based Medicine with Dr. Gordon Guyatt

death. It is estimated that more Americans died from making with the patient. EBM is essential in order taking these drugs than died in the Vietnam War! to be able to understand the likely magnitude of the This last story highlights how, in evidence-based true effects and the balance between desired or un- medicine principles, we have the idea of surrogate desirable effect, and to evaluate one’s confidence in endpoints, where physiological rationale would say, estimates of those effects. This allows one to make the for example, if you get rid of asymptomatic arrhyth- most rational decisions, particularly with your patient mias, then you’ll reduce risk of sudden death. We in shared decision-making, where you are able to en- suggest deep scepticism around surrogate endpoints, sure that together you are choosing what is right for and through randomized trials, investigators have the individual. repeatedly observed discrepant results in surrogates versus patient-important outcomes. It is for this rea- UTMJ: You mentioned that evidence-based medicine has son that we try to be more skeptical and ensure that seen peaks in terms of its popularity, growing and everything is well-proven before it is introduced. widely spreading its reaches. What are some barriers and limitations of this theory and of the movement in UTMJ: You mentioned patient-reported versus surrogate both the clinical and educational domains? outcomes. Is there a certain focus in evidence-based GG: Well, we have come a long way in processing infor- medicine on one or the other? mation. We know that people need access to pre-pro- GG: Yes. We encourage a very high level of scepticism on cessed evidence. We have come a great way in creating surrogate outcomes, and we keep hammering the these collective forms of evidence, UpToDate being point that we should look carefully at outcomes that the most advanced tool we currently have. However, are important to patients. it still can go further and we need to do much better in terms of developing tools for shared decision-mak- UTMJ: If you were to speak to medical students now, what ing. We have all sorts of decision aids to help patients would your final message be towards them regarding but they are hugely underused. We need to find bet- EBM? ter ways of getting the relevant evidence integrated into the interaction between physician and patient. GG: That unless you are able to understand the evidence, Furthermore, the quality of the education of EBM it is very difficult to achieve optimal medical practice, is uncertain. Looking back, there was tremendous and particularly difficult to do what we think is im- resistance at the beginning, and there continues to portant in practice today, which is shared decision- be some resistance remaining. EBM undermines au- making. Often there are choices that are not clear thority, and as a result, it doesn’t sit well with some on whether you should be doing A or B, and indeed, authority figures. depending on the uncertainty and the situation, the correct choice for an individual might be A or might We are still improving the methods and there are be B. In order to engage in shared decision-making still major areas for improvement. There continues to with your patient, you’ve got to understand what the be work done to improve these methods and to edu- pros and cons are. If you, yourself don’t understand, cate both the people developing the guidelines and it is really not possible to engage in shared decision- the people in the front-lines who make the decisions.

UTMJ • Volume 92, Number 2, March 2015 13