Clinical reasoning now a ‘foundational basic science’

An important and emerging need for medical students is developing the ability to apply knowledge and use it to determine the correct diagnoses for individual patients. Diagnostic reasoning has only recently begun to be specifically taught in medical school and residency.


Medical students and new physicians are tasked with mastering myriad topics, including biology, anatomy, and pharmacology. Perhaps the most important aspect of medical training, however, is developing the ability to apply all that knowledge and use it to determine the correct diagnoses for individual patients. This type of diagnostic reasoning has only recently begun to be specifically taught in medical school and residency, according to Robert L. Trowbridge Jr., MD, FACP. Dr. Trowbridge, one of the editors of ACP's recently published book “Teaching Clinical Reasoning,” talked to ACP Internist about the importance of this topic in medical education and ways to incorporate it into the curriculum.

Q: Is the idea of teaching clinical reasoning as a topic fairly new?

A: Certainly, people have learned how to do clinical reasoning over the course of their medical education for a long time. The paradigm before simply was that if you saw a lot of patients, it would all work out and you would become proficient in clinical reasoning. We exposed learners to multiple clinical situations and sometimes coached them with experienced clinician educators. And that's always going to be a central part of [medical education], in terms of using a case-based approach where coaches or mentors can help [learners] through the process.

Image  American College of Physicians
Image © American College of Physicians

The thought that we can explicitly teach it and pay explicit attention to it, I think, is something that's relatively new. There is a growing trend towards putting it in the curriculum, both with a stand-alone course and with attention paid to clinical reasoning within other courses, as well as clerkships and residencies. Several medical schools now have courses that are titled “Introduction to Clinical Reasoning,” where folks go through the basics of clinical reasoning, how it works, kind of the mechanics of it, and give the students a heads-up on how it's going to work.

Q: How did this trend come about?

A: In part, it's driven by us having a better understanding of how we reason clinically, although we're really a long way from knowing exactly how we do it. A lot of it was driven by advances in other fields—in education and cognitive psychology. Those changes and advances are now being applied to medical education.

Q: How do you go about teaching clinical reasoning? Isn't it an abstract thing to teach?

A: It is an abstract thing, but it's also so central to what we do as physicians. When patients come to the physician, a lot of times they want to know 3 things: what they have, what their prognosis is, and whether there is a treatment available. All that really hinges on a diagnosis. It is a rather abstract thing, but it has a very tangible outcome to it, and so teaching folks the different ways and the best ways to get to that outcome isn't nearly as abstract.

Q: How do you take into account the fact that people tend to reason in different ways?

A: One of the important points is there's no one right way to reason. What we want to be able to do is to show [students] that the more tools you have in your toolbox, the more likely it is that you're going to be able to get to the right answer. We all reason a bit differently, but there are ways and techniques we can use to teach people how to reason so that they're more likely to get to the right answer.

Clinical reasoning is very context-specific, so it's dependent on the content of the patient encounter and the patient complaints. For example, a cardiologist may be fantastic at figuring out what's going on with a patient with cardiac complaints but not have really much ability to diagnose a rash. Content specificity is a part of context specificity, where it's also dependent on that particular patient in the environment that the clinician and the patient are operating in. There are so many different things that go into making clinical diagnostic decisions. It's not just what's going on in the doctor's head; it's what's going on with the patient and what's going on around the patient and the physician.

Q: Do you teach students processes that they can follow in order to make decisions?

A: There's nonanalytical reasoning or pattern recognition, where folks kind of just subconsciously recognize the key features of a particular presentation and come to a diagnosis. A lot of honing those skills is experience and seeing many, many patients and many, many presentations of different disorders. In some respects, learning clinical reasoning is about seeing a ton of patients and reading a lot. Honestly, that's how most of us learned it: The more patients we saw, the more we read, the more variations on presentations we saw, the better we got at it.

But then there are also specific techniques that we can use to both hone those pattern recognition skills or those nonanalytical reasoning skills, and then we can also teach analytical reasoning techniques. We can teach students to use something as simple as worst-case scenario medicine—what's the worst this could be?—or do things that are a bit more sophisticated and technique-dependent, like Bayesian analysis and causal reasoning.

Clinical reasoning is something people do every day not as physicians but as people, in terms of making decisions. Why did I make that decision? Why do I make really good decisions sometimes and really bad decisions other times? I think that's actually why [the topic] resonates with physicians, especially internists, because thinking is our procedure. This starts helping us teach that procedure.

Q: Are there aspects of clinical reasoning that are harder to impart or harder for people to grasp?

A: One of the most difficult things is a lot of times people who have expertise in clinical reasoning don't really understand how they got to the diagnosis. Unlike a lot of other things where you can crack it open and say, “Oh, this is how this happened,” we don't have the introspective skills available to really say, “This is how I got to that diagnosis.”

Sometimes, for example, an attending might walk out of the room after examining a patient and say, “That patient clearly has acute cholecystitis,” but they have a hard time explaining exactly why they think that. That can be a really difficult thing for both the learners and the teachers, because the learners look at it as this giant black box. They say, “I have no idea how he did that.” And sometimes the teachers also think, “Well, it just is! Look at it: Of course it is acute cholecystitis.” They can't explain exactly what it is that pushed them towards that diagnosis.

That's one of the challenging things about teaching clinical reasoning: The processes that experienced clinicians use to arrive at a diagnosis are fairly opaque. Even when they know a lot about the clinical reasoning process, it's not at all clear they can say, “Oh, I used this particular process to get there,” because a lot of that runs subconsciously. There are a lot of different patterns that are running in our brain at any given time, both pattern recognition and analytical reasoning, and it's really hard for somebody to parse out exactly what they did to get that diagnosis. That can be frustrating for learners and teachers alike.

Also, assessment of clinical reasoning is one of the biggest challenges that we've faced. Assessment drives learning, and we struggle with the best way to assess clinical reasoning, in part because everything is so context specific. Again, you can do really well in one situation but very poorly in a different situation even if the patient presentation is the same. What we need to do in terms of assessing clinical reasoning right now is have a broad sample in a variety of contexts and probably use a variety of assessment methods to get a reasonable estimate of somebody's clinical reasoning abilities. One thing that we've really stressed is that clinical reasoning isn't a skill, it's not a trait, it's a state or an ability.

Q: How does the book “Teaching Clinical Reasoning” approach the overall topic?

A: The book deals with it fairly comprehensively by looking at it at several different levels. There's a level of looking at it in terms of a curricular approach. How do you build a curriculum in clinical reasoning? When you're designing a curriculum, how do you do that for a preclinical student and how do you do that in the clinical setting? That's part of it, but a much bigger part of it is how clinician educators can do this at the bedside, in the classroom, on the wards, in the office, and giving folks specific teaching techniques that they can use to help their learners understand and apply the clinical reasoning process.

We try to be very specific in terms of teaching techniques while knowing that there's no one right specific way to teach this, that it's really having a lot of arrows in your quiver and knowing enough about the underlying mechanism of clinical reasoning to teach it effectively. It also gets into the assessment of clinical reasoning, how you actually assess how somebody is doing with it. And all of this is based on the underlying theory. I think people may shy away from reading the theoretical concepts chapter, but it's arguably one of the most important chapters in the book because it informs everything else and is very interesting.

Q: What advice would you give an educator who wants to start formally introducing clinical reasoning into the curriculum?

A: For individual clinicians, [my advice would be to] find the teaching techniques that work well in their particular setting and with the types of patients that they have and the environment that they're in and figure out which teaching techniques work the best for them. There are a ton of different teaching techniques, but there are some that aren't going to be applicable to all clinical settings, and there are going to be some that some clinician educators aren't comfortable with. As with most teaching, it's really trying to find the teaching behaviors that work best for your clinical situation.

In terms of developing a curriculum, try to push it as a foundational basic science. Much in the same way that we need to have an understanding of pathology and microbiology and pharmacology to be effective clinicians, we need to have a good understanding of clinical reasoning and to accept it as a foundational basic science and look at it that way rather than as an add-on to an already packed curriculum. The earlier we start this in the education of an individual learner, the better things are going to go.

The other side of this is that this is something that resonates with the learners. This is what people went to medical school and residency to do. It's about figuring out what's going on with the patient. It's a lot of fun to teach, and it's a lot of fun to learn. Emphasizing that in the curriculum can really help a lot.