Name:
Philosophy of evidence-based medicine: Victor Montori, MD, discusses the philosophy of evidence-based medicine.
Description:
Philosophy of evidence-based medicine: Victor Montori, MD, discusses the philosophy of evidence-based medicine.
Thumbnail URL:
https://cadmoremediastorage.blob.core.windows.net/8de35f42-19bb-48ae-a242-c7eb1d683032/thumbnails/8de35f42-19bb-48ae-a242-c7eb1d683032.jpg?sv=2019-02-02&sr=c&sig=rBNXHl104mVKgXpYozBQ%2BDrwNm5d2gkAfdr9VgaY0H8%3D&st=2024-12-21T13%3A51%3A57Z&se=2024-12-21T17%3A56%3A57Z&sp=r
Duration:
T00H19M35S
Embed URL:
https://stream.cadmore.media/player/8de35f42-19bb-48ae-a242-c7eb1d683032
Content URL:
https://cadmoreoriginalmedia.blob.core.windows.net/8de35f42-19bb-48ae-a242-c7eb1d683032/6830489.mp3?sv=2019-02-02&sr=c&sig=pFTgMGvLLXQxIM6p0MvjZWNu08l3VHxWHtSwX5Vh8Tk%3D&st=2024-12-21T13%3A51%3A57Z&se=2024-12-21T15%3A56%3A57Z&sp=r
Upload Date:
2022-02-28T00:00:00.0000000
Transcript:
Language: EN.
Segment:0 .
>> I'm Joan Stephenson, editor of JAMA's medical news and perspective section. Today I have the pleasure of speaking with Dr. Victor Montori about the philosophy of evidence-based medicine, a topic discussed in Chapter 2 of "User's Guides to the Medical Literature." Dr. Montori, why don't you introduce yourself to our listeners? >> Yes. Thanks. I am Victor Montori. I'm professor of medicine at Mayo Clinic, an enthusiast of evidence-based medicine, and a researcher in how to transfer evidence into practice in patient centered ways.
>> How does evidence-based medicine differ from the traditional paradigm of medical practice? >> Fundamentally evidence-based medicine differs in the recognition of two fundamental principles. The first one is that decisions are made with greater confidence when one has solid evidence, solid research, supporting that decision, supporting the estimation of benefits and harms. So the better the research, the more confident the decision maker is. That first principle of evidence-based medicine suggests that there is research that is better than other research.
That there are observations in nature that we call evidence that are more protected from error than others. And again, the better the research, the more confident the decision maker is. That recognition is different from traditional forms of medical practice in which folks were more confident. The more gray hair they had, the more patients they had seen, the more experience they had accumulated, the more observations they personally had made. And we now recognize that some of those unsystematic personal observations are not as protected from error in developing inferences about what is true than other forms of research such as large-scale experiments or randomized trials.
The other thing that is different is that traditionally people relied heavily on their biological understanding of mechanisms. So if you understand how the body works, you should be able to figure out what this treatment will do to a patient's outcomes. And I think there's been a recognition that observations in laboratory and in animals often do not reproduce or do not predict accurately what the outcome will be in populations or individual patients. And as a result, therefore, assuming that a complete understanding of mechanisms will lead to a prediction of outcomes is not as good as now we know we can do better with better forms of evidence such as large-scale observational studies or randomized trials.
So this recognition of a hierarchy of evidence corresponds to the first principle of evidence-based medicine. The second way in which evidence-based medicine is fundamentally different from traditional forms of medical practice is through its second principle, which basically states that the research by itself never fully informs any clinical decision. And that one needs to incorporate the context, values, and preferences of the patient. Again, in traditional medical practice it's been considered that the physician alone is able to make technical decisions on behalf of the patient and tell them what treatments to take and what are the best courses of action.
However this appears to be a rare circumstance nowadays given that the patient's context can severely change what's the right thing to do even if the research evidence is strong in favor of a particular treatment. You could have very strong evidence that an antibiotic is helpful for a treatment of pneumonia in a patient, but that patient may not want to have their pneumonia treated because they are at the end of their life. So changing the context, the decision also has to change even though the evidence may be strongly in favor of one particular course of action.
And similarly, recognition that patients' values and preferences and goals could be quite diverse suggests that if one fails to recognize those, one could potentially do net harm even if one is using a fundamentally safe intervention by pursuing a goal that a patient may not want to pursue. So patients, for instance, exposed to cholesterol lowering medicines may experience costs and burdens of taking treatment every day to reduce cardiovascular risk. The patient may not be as interested in reducing cardiovascular risk with drugs than they are with, say, lifestyle changes.
And a clinician expecting them to take this medicine may be introducing an intervention that's undesirable to the patient. And the patient may respond with non-adherence and we have a situation of waste in the system and misuse of resources. So at the end of the day the second principle of evidence-based medicine that states that the evidence alone never tells you what to do, that you have to incorporate patient values, preferences, and the context, both social and clinical, highlights the importance of clinicians' expertise in balancing all these factors along with the best available evidence.
And evolution within evidence-based medicine has been to recognition of that clinician's expertise such that now I think the most enlightened version of evidence-based medicine recognizes the need to develop skills and knowledge in clinicians so that they can understand what level of evidence they're dealing with to deal with the first principle, and what is the right context that they're facing, making the correct diagnosis of it, and then to have the skills to bring the patients into participatory forms of decision making so that there's hope that their values and preferences will become known and will play a role in deciding the best course of action.
All these things were absent in traditional medical care, and that's how EBM is different from that. >> It's very different. I wonder if you could describe the hierarchy of strength of evidence for prevention and treatment decisions. >> Sure. I think one has to recognize that when one talks about the hierarchy of evidence one talks about a sort of unordered list of research designs that are ordered in a particular way and that reflect a particular goal. So, for instance, if you're interested fundamentally in getting to truths about the efficacy of treatments or preventive interventions, for large populations a randomized clinical trial of large populations is, we think, if it's properly done, going to provide you with the most reliable, the most valid, set of estimates about risks and harms.
It is interesting that for an individual patient that may not be the best estimate. If, for instance, one could do a randomized trial on just that one patient and expose the patient to, for instance -- to different interventions, and do so in a way that will be protected from bias by, for instance, blinding the patient to which intervention they're getting at what time. But unfortunately this only works for conditions that are fully reversible and for interventions that do not leave long lasting sequelae.
The list of conditions where this is applicable is actually very small. And so these so-called N of 1 randomized trials have limited use, but when they're feasible they provide the highest level of evidence for the individual patient. And it will be the patient that participates in the trial. For most other circumstances, for the vast majority of other circumstances, randomized trials that are adequately protected from bias and are adequately large could provide good estimates. In fact, most trials tend not to include a broad enough range of patients, tend not to be large enough to provide inferences, and do not allow for an assessment of whether there are important variations in effect across subgroups.
And so systematic reviews and meta-analysis, which are just pooling exercises where one takes all the relevant trials published in a topic and pools them together statistically, those could potentially overcome the limitations of each individual study not in terms of the protection from bias, but in terms of issues such as precision and consistency. So systematic reviews of randomized trials are often placed at the top of a hierarchy of evidence because, again, they're based on randomized trials, which we now know can be trusted to provide valid results if they're adequately protected from error and they're pooled together so that they have adequate precision and applicability across a broad range of patients.
After that you have the individual randomized trial, which we've now discussed why that is valuable. And then you have observational studies. And the difference between observational studies and randomized trials is that in observational studies the investigator does not get to allocate randomly the intervention or the exposure to the patients, but rather patients get exposed either by choice or from not the investigator, but their clinicians or themselves or by chance. But not the chance that is controlled by the investigators in the randomized trial, but those that are likely from the context in which they get care.
Because of those forms of allocation, mostly by choice, there could be a relationship between why they chose to take a particular intervention and the ultimate outcomes that they report. And as a result, observational studies are not as protected in terms of the selection of patients and protection against other interventions as randomized trials are. And so we can get good estimates, but the confidence we have in those estimates is preliminary pending confirmation in randomized trials. Forms of observational studies that are becoming more popular these days include studies made on registries from practicing, which all sorts of outcome measures are prospectively defined and collected.
And the only thing that is not there for a full randomized trial is that the patients did not get to choose their interventions. That seems to be a minor concern, but is actually critical given, again, the common situation that the reason why you choose an intervention is because you believe that it will derive better outcomes for you. And then all these randomized trials, observational studies, improve their value to patients if they directly bear on the questions that they and their clinicians ask. In other words, they use the right set of patients, they use the right intervention, they compare that intervention with the appropriate comparison that is either no intervention or something that is commonly used in practice in the way that it's normally used in practice.
And then measure outcomes that matter to patients rather than markers that are biochemical markers or surrogate markers that people have to imagine then how the intervention will actually play out in terms of their real outcomes. Finally, as we discussed at the beginning, unsystematic clinical observations or observations in the laboratory, there the extrapolations to real life and to real patients and to real patient outcomes is very tenuous. And those need to be paired with the outcomes in observational studies or randomized trials, and so they represent very preliminary forms of evidence.
>> Dr. Montori, what knowledge and skills are necessary for optimal evidence-based practice? >> The set of skills necessary for evidence-based practice is a field of investigation and evolving exploration. The first, I think -- if one is defining whether the skills for optimal medical practice -- in general certainly one is talking about both technical expertise or clinical expertise to properly diagnose a situation the patient is in and understand carefully the comorbidities.
In other words, the other medical conditions the patient brings to the table that are present and active at the time one is assessing the patient. So that's fundamental. The second set of skills that are important have to do with being able to use the evidence from clinical research and practice. And there are multiple services now available that bring the evidence to the desktop or to the handheld in very effective ways. And so the set of skills necessary for using the evidence are evolving from the early days in which it involved being able to negotiate with databases such as Medline to identify the relevant articles, obtain the articles in full, review them carefully, and then draw inferences to now being able to deal with summaries of those articles delivered at the point of care sometimes without any specific context or request.
And so that includes noise more than signal. So clinicians nowadays have to have skills in the consumption of evidence, in the consumption of information, skills that make the clinician efficient, but at the same time able to bring the evidence to bear as needed. And so those skills, critically identifying the literature, critically reviewing it, and then drawing the appropriate inferences, are new sort of skills that are more frequently nowadays than in the past being taught in medical school, but infrequently still role modeled in further training and very infrequently found in actual practice.
Now when new information comes forward, it is very important that clinicians have enough of an understanding of medicine to be able to identify where this new information fits in what is known. And so having adequate knowledge of the underlying biology of disease and health, but also underlying socioeconomic and other determinants of health is critical to know how new treatments and new procedures and new tests fit, both in terms of the clinical care of these patients, but also in the use of resources.
Because of the second principle of evidence-based medicine, a set of skills that perhaps may not be as obvious as the ones I mentioned are also necessary, and that is the ability to present options in an evidence-based way to patients for them to participate in decision making is important. How do you distill the range of options and the pros and cons of those options in ways that reflect not only what the evidence points out, but also how strong that evidence is, how much uncertainty there is about those pros and cons.
So presentation of information skills is something that is important, and does not get stressed enough in training of clinicians. And finally, having the ability to have conversations with patients in which patients have enough of an opportunity to voice their concerns or values, their preferences, and to deliberate with the clinician about those issues. So there's quite a bit of exploration right now about deliberating with patients and how clinicians can learn those skills, and again, how to do that in an evidence-based way.
So to summarize, it's the traditional skills of a physician or clinician in terms of basic knowledge of medicine, basic knowledge of health and healthcare, ability to identify and critically appraise the evidence, and then sort of a newer set of skills, the ability to present that evidence to the patient, help the patient understand it, deliberate with the patient about the pros and cons in ways that incorporate the patient's context, values, and preferences, and then create decisions by consensus that the patient can implement with our assistance.
All those things are required for an effective evidence-based practitioner. >> Those are a broad range of skills. Are there additional challenges that clinicians face when using evidence-based medicine? >> I think there are multiple challenges that clinicians face. One is, of course, the amount of noise coming out of the medical research enterprise. Not every study, not every trial, not every piece of research that gets highlighted by the media or put forward as a top article in many of the major medical journals, not all of them have to be reviewed and acted upon.
So there is the need for very careful consideration of when to act on the evidence. That is an important challenge. There is also the difficulty of sorting out what is ready for implementation and what is evolving knowledge. So adopting new technologies and new tests, at some point those are ready for practice and should be used in your patients. So knowing when to jump in is another sort of skill for an evidence-based clinician.
That is a challenge at the moment. The enormous direct-to-consumer advertisement and direct-to-patient advertisement of products and services sometimes clouds our judgment and affects our preferences and those of our patients for perhaps more treatment rather than better treatment. And there are incentives, of course, in place for patients and directed at clinicians to act in ways that may not be the most evidence based. There is a limit of time and competing demands to practice evidence-based medicine versus other things.
And at the end of the day we're all human and we tend to be creatures of habit and if we haven't developed habits of evidence-based practice, it will take time to do it every time, and it will take time to acquire this. So we may not be allowed or afforded the opportunity to do this and learn how to do this and to apply it in a practiced way and get feedback for it. And, as a result, we would not develop mastery, and it will always take more time, and eventually we'll stop doing it. So there are multiple challenges, both structural in the way we deliver healthcare -- and we support clinicians in doing so -- as well as with the evidence itself and in the context of healthcare, most of which is going in the direction of less evidence-based practice rather than more.
>> Is there anything else you would like to tell our listeners about evidence-based medicine? >> I have found that many people know what evidence-based medicine appears to be, but few have sat down to think about these two principles. Most recognize the first one, and they shorthand it as evidence-based medicine is about randomized trials. And clearly that is not the case. Evidence-based medicine recognizes a whole range of evidence, but also recognizes that even within randomized trials each piece of evidence will be affected by a certain degree of trouble and error.
And to the extent that we can figure out what that is, we'll know to be confident about the estimates or not. And that is the first principle. But most people do not notice that there is a second principle. And that second principle is fundamentally a patient centered principle that requires us to incorporate the values and preferences and context of our patients, which is the true nature of any clinician that cares deeply about their patients. And finally, some people may think that evidence-based medicine is some sort of mechanical application of evidence in practice without necessarily engaging much brain activity.
Certainly if you have to keep in mind at the same time the clinical context of the patient, the socioeconomic context of the patient, the patient's values and preferences, and add to this the best available research evidence, and do so in a way that engages the patient in deliberation, now you're talking about a highly skilled expert clinician. So instead of being a minimum denominator, this is an aspiration. And I think -- I hope that your listeners, as they engage with evidence-based medicine and interact with the philosophy of evidence-based medicine, they will consider this a recipe for aspiring to be better clinicians for themselves, for their own personal satisfaction and professional satisfaction, and for their patients.
>> Thank you, Dr. Montori, for this useful snapshot of a philosophy of evidence-based medicine. Additional information about this topic is available in Chapter 2 of the "User's Guides to the Medical Literature." This has been Joan Stephenson of JAMA talking with Dr. Victor Montori about the philosophy of evidence-based medicine for JAMAevidence.