PELC: “Creating a Culture of Clinical Reasoning: Tips for Educators” by Thilan Wijesekera, MD, MHS

November 06, 2023

Information

“Creating a Culture of Clinical Reasoning: Tips for Educators” by Thilan Wijesekera, MD, MHS

ID10947

To CiteDCA Citation Guide

00:00We've got, let's see, we are recording,
00:04I see Catherine on here Dungen, great.
00:07And so again we tend to
00:09kind of filter in slowly.
00:11So wanted to introduce everybody to
00:15one of the members of our education
00:20leadership in the school. You know,
00:26Felon and I I practice this very often.
00:32Wish a second wish a Sacre, am I?
00:36Hopefully there.
00:38Benita, you're putting so
00:40much pressure on yourself.
00:41I'm happy to finish my introduction
00:42as well and get started.
00:44I'm excited to be here,
00:46but I'm actually going to give a little
00:48bit more than introductions here.
00:50And so Dylan got his MD from the University
00:53University of Rochester School of Medicine.
00:57He did residency training at our at
01:00Yale's primary care residency program.
01:03And then Phelan did a general internal
01:06medicine fellowship in medical
01:08education at Yale and then also got his
01:13MEHSMED in our institution.
01:182018, he joined the academic
01:22hospitalist program in Yale's
01:25General Internal Medicine program.
01:28He's very active in medical education.
01:31He's the director of Clinical Reasoning
01:33at the Teaching and Learning Center.
01:36He's an associate for Clinical
01:38Reasoning Educator Development,
01:40And so he provides consultations,
01:43workshops, and scholarship related
01:46to teaching clinical reasoning.
01:48And his research interests
01:51include clinical reasoning,
01:52diagnostic error with publications
01:55and academic medicine,
01:57medical Teacher,
01:58and the Journal of General Internal Medicine.
02:01And so with that introduction,
02:06all yours. Awesome.
02:07Thank you so much Pernina,
02:09and it's great having all of y'all here.
02:12I know that you are in between
02:15things often times with lunch,
02:18so feel free to eat and chat.
02:19But I will say part of my one of
02:22my favorite parts of these sessions
02:24is just the conversations we can
02:26have and learning from each other.
02:28And especially in your experience
02:30'cause I give these workshops
02:32to different departments where
02:34obviously it's not my expertise,
02:36but yeah, we're going to talk
02:39about strategies for evaluating
02:41and teaching clinical reasoning.
02:43I. So for the CME credit,
02:47make sure to text
02:5042047 to the Yale CME number and that'll
02:56be put in the chat again. Again, that's
03:0042047. Just a quick disclosure for me.
03:04So I am a consultant for the National
03:06Board of Medical Examiners around like
03:09developing assessments for management,
03:11reasoning, an area of my research and
03:13something we'll touch on a little bit.
03:15But just for your note, So what are
03:19we going to be talking about today?
03:213 broad aims for our workshops
03:23that I usually talk about, One,
03:25why clinical reasoning matters and
03:28some broad information terms around it,
03:302 how do we assess clinical
03:33reason in the clinical setting.
03:35You know, it's interesting.
03:36I've been giving a lot of workshops with
03:38colleagues across the country for this.
03:40I always like to pair assessment
03:42before teaching just because,
03:44like diagnosing the learner,
03:47no pun intended,
03:48can be so helpful for deciding what to teach.
03:50Because I know a lot of people don't
03:52necessarily have a lot of time when you
03:54are in your clinical settings with all
03:55the patients that you have to see as well.
03:58And we'll practice with a couple role plays
04:01for both assessing and teaching strategies.
04:04And I'd love to hear and encourage
04:07you to share what you're thinking
04:09about for these hypothetical learners
04:11in these situations.
04:13So let's start talking about
04:15clinical reasoning, epidemiology,
04:16terminology, and theory.
04:18So why does clinical reasoning matter?
04:21So we've been talking about
04:23clinical reasoning for decades.
04:24The literature becomes like pretty
04:27robust as of like 40 years ago.
04:30But what really started to pick up steam
04:32and interest around the topic is the
04:35National Academy of Medicine report
04:36Improving Diagnosis in healthcare,
04:38which came out in 2015 and 16.
04:41Then in this report it summarized
04:43the literature.
04:43The diagnostic errors happen a lot,
04:46anywhere from 10 to 15% of the counters
04:49in the more cognitive specialties.
04:50So medicine, internal medicine,
04:53Pediatrics, emergency medicine,
04:54maybe less so in some of the
04:57perceptual specialities,
04:58something like a radiology and pathology.
05:00And these have been consistently
05:03shown across various forms of study.
05:06Chart review being the most common
05:08types of study and they can have
05:10a significant impact.
05:11Various back in the envelope mass have
05:14used numbers estimating that it could.
05:17The diagnostic errors happened
05:19to literally everyone.
05:21The mortality,
05:21all the numbers depending on how
05:23you want to define diagnostic
05:24errors and the harm they cause,
05:25can go up and down from that 40 to 80,000
05:29number yearly in the US that I mentioned.
05:32And what makes clinical reasoning so
05:34important in this process is that
05:37when we have looked at diagnostic
05:38errors and I was actually a part of
05:41one of these chart review studies
05:43affiliated with UCSF here and about
05:4770% of them have been affiliated to
05:50be cognitive in some form or another,
05:53which is why we like to teach
05:55about clinical reasoning.
05:56Now,
05:56clinical reasoning can be
05:58defined in many different ways,
06:00ranging from broadly problem
06:02solving to critical thinking,
06:05sometimes more gestalt,
06:07like this learner's judgement
06:09that I'm observing right now,
06:12but the way I like to define it,
06:15and it's not particularly the new definition.
06:17This is from Kevin Eva,
06:18the editor of Medical Education,
06:20defined it as the ability to sort
06:23through a cluster of features presented
06:26by a patient and the sign a diagnostic
06:29label with the development of
06:31appropriate treatment as our end goal.
06:35Now when it comes to clinical
06:37reasoning theories,
06:37there are so many frameworks
06:39that inform what we teach,
06:40often drawn from other sciences,
06:43particularly the cognitive science
06:46realm and cognitive psychology,
06:49to maybe even a little bit
06:50from business as well.
06:52I don't know if any of
06:53these are familiar to you,
06:54but it's just so I can give
06:57a broad strokes on them.
06:58The deliberate practice,
07:00just the importance of doing effortful,
07:03meaningful practice with getting
07:05reps around your around clinical
07:09reasoning in different settings,
07:11knowledge organization,
07:12both frameworks and approaches
07:14to your cases and also how you
07:18store that information yourself.
07:19The term illness script has become
07:22really popular and we've recently
07:24developed a framework called the
07:26management script of how you organize
07:28information about diagnosis and
07:30management options in that given patient.
07:33In addition to that,
07:36sorry about quick Trigger on
07:38my mouse dual process theory,
07:40the idea of thinking fast and slow
07:43might be familiar to y'all so thinking fast,
07:46so system.
07:47So this is popularized by Daniel Kahneman
07:49who won actually a Nobel Prize for this work.
07:51He was actually getting
07:52economics I want to say,
07:54but system one being pattern recognition,
07:57which is what our more seasoned experts use,
08:00versus system 2 which is our
08:02more analytical thinking.
08:03But it's much more time consuming and
08:06eventually over time learners move
08:08more from System 2 to system one.
08:10And then finally,
08:12probably the biggest theory that's
08:14gaining traction in the clinical
08:16reasoning literature is around situativity.
08:18So what's going on in the environment
08:20knowing that clinical reasoning
08:21isn't just practice inside the head,
08:24that it's practice in an environment
08:26that changes,
08:27that involves many different partners,
08:29including the patients involved
08:31in their care.
08:32And they're even subcategories
08:33of this that I won't believe.
08:35But just the idea of context is
08:38important in clinical reason.
08:40So if I had to break down a cognitive
08:42model and it's really started to
08:44become more solidified over the past,
08:47I want to say like 7 to 10 years.
08:50There are 4 main steps in the
08:53clinical reasoning process that can
08:54be like ordered in different ways,
08:56but those steps are data collection.
09:00So how your learners get information,
09:02all of these will derive from the
09:04the definition that I gave a clinical
09:07reasoning problem representation.
09:08So how you synthesize that information,
09:11how do you determine signal from
09:13noise sick or not sick.
09:15We usually see that in the one liner
09:18that assessment that we think about,
09:20you know the part where we as
09:22attending is actually like perk
09:23up and listen a little bit more
09:25closely to our learner.
09:26Next prioritizing A differential diagnosis,
09:29obviously what we think of
09:31most with clinical reasoning.
09:32If I could break it down just into two steps,
09:34a little bit more, I would say one,
09:36just coming up with a list of diagnosis.
09:39This is what we usually see our more
09:42junior learner struggle with the
09:45right like your medical PA students.
09:47Whereas when you get to more senior learners,
09:50it's more about how do they decide
09:52which of those diagnosis files,
09:54those illness scripts, should they choose,
09:56which one matches more to this diagnosis.
10:00And that's where,
10:01you know they can decide on what
10:02eventually they need to do,
10:04which is the next step of management
10:06reasoning and break this down into
10:08two components, management scripts,
10:10how do you decide what your options are?
10:14And then finally,
10:15testing and treatment thresholds,
10:16whether or not to do something at all.
10:19So with those 4 broad steps,
10:21I want to think about how we assess
10:23our learners and teach our learners.
10:25But before I do,
10:27did anybody have any questions
10:29about this so far?
10:33All right. Last thing I'll just
10:36say about this is that this model,
10:38so to speak, of clinical
10:40reasoning is constantly happening.
10:42The process of clinical
10:44reasoning is iterative.
10:45And that's actually the best process, right,
10:47Because we know there's so much uncertainty.
10:49We know that there's so much new information
10:52that even as you're taking a history,
10:53you know that your differential
10:55is changing literally word to word
10:57by what our learners are saying.
10:59So with that in mind,
11:00how can we assess clinical reasoning?
11:03All right, so warm y'all up a little bit,
11:05everybody's gotten a chance to eat.
11:06This is where I asked for some audience
11:09participation, even if cameras are off.
11:11That's OK.
11:12And I was wondering for y'all,
11:16how do you identify?
11:18If your learner is struggling
11:20with clinical reasoning,
11:21feel free to put it in the chat
11:23or unmute yourself and share.
11:42know your learner struggling
11:43with clinical reason
11:47for you guys who might not have microphones,
11:49Feel free to put it in the chat too,
11:53right? Melissa said. Doesn't
11:55formulate an appropriate differential.
11:56Yeah, that's like kind of like the
12:00the the stop as far as like oh wow,
12:02like now you need to say what
12:04you think is going on and you're
12:06not just presenting information.
12:08Definitely. For sure.
12:08The differential is a great
12:09place to diagnose,
12:15telling you you can't even summarize that.
12:18You know, Kind of like the assessment
12:21where it's just like a whole bunch of
12:24facts kind of put together, all hodgepodge.
12:27Yes, for sure, right when they are
12:30really struggling to like collect
12:32or all that information to what
12:34what matters most, right Penina.
12:36And really determining
12:37that can be challenging.
12:39So at least for each of these steps,
12:40here are some broad ways
12:42that I understand a patient,
12:44when you're listening to our
12:45presentations that you can kind of
12:47like tune into where you might be
12:48able to teach your learners the most.
12:50So first is for data collection.
12:53The quality of the information is very good.
12:56It's not clear.
12:58It's missing information.
12:59And that's the gestalt.
13:00That's when you're like,
13:01you wasn't attending,
13:02you're like, wait, wait,
13:03what's going on with this patient?
13:05That's usually a sign that your learner
13:08hasn't gotten enough information.
13:10Next up, problem representation,
13:11like Panino was mentioning,
13:12most seen in the assessment,
13:14also in sign outs.
13:15That's a great time to figure out or
13:17in consults hearing like a learner
13:19summarize that information and they
13:20just can't put that information together.
13:22They're not including the right stuff.
13:24It's disorganized and all over the place.
13:27It doesn't really give a tempo,
13:29a timing of how the patient
13:32might be evolving,
13:33prioritizing A differential diagnosis.
13:35So obvious, like we said,
13:37either there's not enough diagnosis and
13:39they're not really in a reasonable order.
13:41And I will say this about honestly,
13:43this goes for clinical reasoning,
13:45but specifically for
13:47prioritizing differential,
13:48we're not looking for a
13:50single correct answer.
13:51There's so many analogies.
13:53I'll I'll use some related to like sports,
13:55for example, a strike zone,
13:57especially because we're like
13:59baseball season right now,
14:00but we're not expecting it to
14:02be in a single place,
14:03right.
14:03It could be around a general area
14:05for a differential for a learner's
14:08clinical reasoning to be fairly strong,
14:10but we're looking at to be for it to
14:13be somewhere around there management
14:15even more so actually in management
14:18there often isn't a right answer and
14:22the the answer changes right over time.
14:24So some signs of this can be a
14:26little bit more it can be either
14:28like very blunt and obvious right.
14:30The plan is incomplete.
14:32There's very clear evidence based
14:34guideline interventions that are not
14:37being like suggested recommended ordered.
14:39But then there are like more like faint
14:42signs that I'll notice particularly
14:44in higher stakes such situations.
14:46I attended our step down
14:48unit not infrequently,
14:49so I'll notice here some things
14:53like indecisiveness which might
14:55suggest that a learner has
14:56doesn't have really well honed
14:58testing and treatment thresholds.
14:59It's great to like have
15:01uncertainty and acknowledge that,
15:03but you got to know what you want
15:05to do with that uncertainty over
15:06testing under treatment that my
15:08high value care colleagues will
15:09certainly want me to recommend that.
15:11I include that in in management
15:14reasoning too.
15:15And there's a ton of overlap as well.
15:16There's our patient care
15:18communication aspects around the plan,
15:20but those are some general impressions
15:22of what you're looking at for when a
15:25learner might be struggling in those
15:27areas. All right,
15:28I'm gonna think about some.
15:30I'm gonna share some strategies,
15:32both generally about teaching
15:34clinical reasoning and specific
15:36with the niche strategy.
15:38But I'm gonna pause again.
15:40I'm putting myself out there to
15:42everybody to ask what are some ways that
15:44you like to teach clinical reasoning?
15:46It could be anything,
15:48a very broad open-ended when you
15:51like teaching how you like teaching.
15:54Whatever comes to mind,
16:11Richard, that you're unmuted.
16:12Did you want to say something?
16:14Yeah, I was. I mean it's more
16:16I guess relevant in cardiology,
16:19but I like to usually point them
16:22towards thinking the Physiology of
16:24what we are talking about and then
16:26have that drive towards, you know,
16:28what you're going to do for the patient.
16:30That's great. And it's Rushka,
16:32right, that I, I, I love that.
16:35And that's really good for learners,
16:38especially when they're newish
16:40to a field as well, right?
16:42Especially so your,
16:44your clerkship learners,
16:45they have that basic and science knowledge.
16:46So that's actually kind of fun and
16:49reinforcing for them to see like oh wow,
16:50like I know something or
16:52you're like early fellows.
16:54I'd imagine they're still like really wrote,
16:57you know, close to that,
16:58like that Physiology,
17:00although I imagine this specialty,
17:02you're always close to your Physiology,
17:04but tying that back that can
17:06connect it ground alertness,
17:07clinical reasoning and and almost
17:09act as a framework too for different
17:11options of possibilities, right?
17:13Like a patient might have an
17:15AKI and you can say like,
17:16OK, what are like the parts?
17:18Like let's follow like the urine
17:19all the way to like the ureters,
17:21right?
17:21Like,
17:22So what are the areas where a lesion
17:24could occur to 'cause this AKI?
17:41So using Physiology,
17:42that's a that's a great way
17:43to teach clinical reasoning.
17:45Anything else?
17:46Maybe one more strategy.
17:48I'll wait for
17:50quiet one day group, everyone
17:55or a clinic without a audio
18:00Penina. What's that? What's
18:01the strategy for you to teach?
18:02Yeah. I would just say I I
18:05feel like a four year old or a
18:08three-year old or always say why.
18:11Yeah, you know, it's like not
18:14just regurgitation, but OK,
18:17Why, why did you say that?
18:19What were your thoughts?
18:22Or else asking like the learner
18:25just didn't explain their reasoning
18:27versus just giving an answer
18:29that is amazing.
18:30And that's, are you going to be
18:32the first one that I say for Nina?
18:35So First off is just clarification, right?
18:38Like, what were you thinking behind that,
18:41right, Like because we know from
18:43multiple choice questions, right,
18:44that like an answer doesn't
18:46necessarily show what's going on.
18:47So just asking them to clarify
18:49in different ways, right?
18:50That could be like saying why
18:52watching them do an exam or watching
18:55them ask like those questions.
18:56So getting primary information
18:58arguably is the most important,
19:00though the most time consuming as well.
19:03So you know,
19:04important to note in certain
19:06situations scaffold.
19:07So this goes to what Carol was mentioning
19:09and also Urgeka to a degree as well, right?
19:13So for frameworks, frameworks,
19:14frameworks is the name of the game.
19:16In teaching clinical reasoning,
19:17we simply to go back to deliberate practice.
19:20We don't get like our 10,000 hours anymore
19:24in clinical practice to really have like
19:27exemplars for every single situation.
19:29So giving your learners ways to identify and
19:32figure out what's going on in a patient,
19:35even if they haven't seen it before,
19:37can be helpful.
19:38And so common ones are just
19:41by organ system anatomy.
19:43Some could be like processed like vascular,
19:47neoplastic, infectious, inflammatory,
19:51or even something such as simple as
19:54what's a must not misdiagnosis for this,
19:56what's common even like, you know,
19:57remove like some like data.
19:59Just thinking about this cheap
20:01concern and don't narrow too soon.
20:03Frameworks is a way to really
20:05teach clinical reasoning modeling.
20:07So this is one way I wanted to say
20:09you're allowed to do this as faculty.
20:11This is literally the main feedback
20:13that I got for my first two to three
20:15years as an attending on my emails.
20:17Doctor W,
20:18It's nice that you say that
20:20that was a reasonable plan,
20:21but we'd really want to
20:23hear what your thinking is.
20:24And so clarifying why like
20:26you wanted to make a decision,
20:28what was it about this patient
20:30or this intervention that made
20:32you want to do it or not.
20:33And that can actually be
20:34really quick as well,
20:35because Panino has mentioned that I
20:37I know a lot of y'all are in clinic
20:39and in the outpatient setting,
20:41you're like moving fast and might not
20:42have as much flexibility for teaching.
20:44So even some quick modelling can be helpful.
20:48And then this is one that I after
20:50doing a lot of these workshops with
20:53students that they asked me to put in,
20:56is helping them do some independent learning.
20:59Because you can imagine when
21:00it comes to clinical reasoning,
21:01there are so many different points of
21:04a case that learners can grow from,
21:06but that can be really overwhelming too.
21:09So specifically saying, like, hey,
21:12we saw a lot of patients today,
21:13but could you look up cystic fibrosis
21:16and make an illness script and
21:18talk about it tomorrow, right.
21:19Or could you look into the trilogy
21:22of flow and we can talk about
21:24different management options as well,
21:26right. And and make it really more.
21:29And sometimes what I'll even do is
21:31like you have no more than 10 minutes
21:33to look this up because that can
21:35often give just a sense of relief with
21:38all the things that they need to do.
21:39So anyway, those are some broad categories
21:41as you can imagine for each of those steps.
21:44There are examples for this.
21:46If I had to give data collection
21:49the biggest things that I do when
21:52I'm teaching clinical reasoning in
21:54the clinical setting on the words,
21:57for me
22:00it's tough. I don't get a chance to
22:02like observe as much as I would want to.
22:04I think that's particularly
22:06important for more junilers.
22:07But when I go back bedside,
22:10I'll try my best to like model
22:12some specific maneuvers for them,
22:13spelled NEUVERS, missing any there.
22:16But also I might provide some
22:18like additional resources.
22:20There are great ones.
22:21The Stanford 25 is an excellent one.
22:23Rational Clinical Examiner is another
22:24one which also I think has a few really
22:27nice PEDs files as well one on appendices.
22:30This particularly comes to
22:31mind for teaching strategies.
22:35I was around problem representation.
22:38So broad categories of like when you're,
22:41when you're trying to help your learner
22:43distill what's going on with the patient,
22:45just asking them who is the patient, again,
22:48what matters most about who they are,
22:51their comorbidities,
22:52their risk factors epidemiologically
22:54and what is going on.
22:56Focusing on that chief concern,
22:58the time course and the two to
23:01three symptoms that really help you
23:04narrow your differential diagnosis.
23:06One thing I will say is it can be
23:09helpful to provide your learners
23:10a template particularly for the
23:12assessment just from being around
23:13many different reports and having
23:15so many learners having to present.
23:17To me,
23:17I know it can be really overwhelming
23:19to give that one liner.
23:21And so this is just a template that we use.
23:23It's very mediciney, internal mediciney.
23:25But whatever your framework is that
23:27you want to help your learners with,
23:30just tell them.
23:31For example, I remember Jeremy Mueller,
23:35who's like one of my favorite
23:36educators here at Yale.
23:37He he's really big on.
23:39Tell me exactly what the like
23:41presentation is first and then go after.
23:43Right.
23:43So like,
23:44this is like a cute,
23:48this is acute weakness in the
23:50setting of like whatever,
23:52like symptoms and features are coming right.
23:55And that can be or this is like
23:57a mono neuropathy,
23:58polyneuropathy And having that be
23:59the lead and then going from there,
24:01just tell your learner so that
24:03they Canmore plug and chug their
24:05information as opposed to having that
24:07like germane load of figuring it out,
24:10prioritizing differential,
24:10y'all are all pros at all of these,
24:13frankly, but definitely for differential.
24:14But just so I can tell you this
24:16is what I do with these workshops,
24:17to be honest,
24:18I just like tell the like teachers
24:20what they're doing and they're like
24:22what's what tools they're already
24:24using in their toolbox so they can
24:26use them in the future more confidently.
24:29But the scaffolds,
24:30like we mentioned from Carol earlier,
24:33pneumonic schemas can be really
24:35good for increasing just the number
24:37of diagnosis that are considering.
24:39The biggest one I use when I'm
24:42on the boards is pivot points.
24:44So pivot points means what a part of
24:47the case really LED you one way or the other.
24:50That's what can be really best
24:52modelled and that can really
24:54help you discern how fleshed out
24:57your learner's illness
24:58scripts are. And for independent learning.
25:00There are a bunch of resources here as well.
25:04For example, one that I feel like
25:06there's a Pediatrics podcast I heard
25:07about recently that's really good.
25:09I think it's called the Crib Ciders as well,
25:12but a lot of fun resources to help your
25:14learners learn really easily. All right.
25:18Teaching strategies for management.
25:20Again, this is my area where
25:22I get really jazzed up.
25:23But the teaching strategies that I think
25:26of most when it comes to management
25:29is providing a management script,
25:32even if it's suggesting what
25:34general categories you can think of.
25:36So for example, like today,
25:39this morning I was rounding with my res
25:40and we were really stuck on a patient.
25:42So I was like, all right,
25:42like what else can we do?
25:44Can you think of any labs imagings,
25:46do we need help,
25:47do we need to call somebody or monitoring?
25:50Is there any way we can follow up
25:53information And just that helped us
25:54figure out what a few other interventions
25:56that we might want it to do as well.
25:58I found that's particularly helpful with more
26:01with learners on the more like student range.
26:06But even early fellows can certainly
26:09find help with that specific tier field,
26:11the threshold factors.
26:13This is when things get really nuanced
26:16about why or why not to do something.
26:19I'll really work through my
26:20learners to say like,
26:21what is it about this case,
26:23especially when it's tough.
26:24When there is that indecisiveness,
26:25what is it that matters most?
26:27Is it how sick this patient is
26:29that's going to determine whether
26:30or not we choose antibiotics?
26:32Is it just this patient's goals
26:33of care when we decide whether or
26:36not we want to use chemotherapy,
26:37really figuring out what is it
26:40that makes us decide one way
26:43or the other about a patient.
26:46Another thing I'll say because
26:47I see a lot of specialists on
26:48the call that can be helpful,
26:50particularly especially with learners
26:52who might not necessarily be on
26:55that rotation initially, right?
26:57Like you have a general resident on like
26:59palm consults like Penina or like you know,
27:01a student on cards consults with Ruchika.
27:04I think just saying, hey,
27:07before you start this rotation,
27:09you really need to get down
27:11at like like acute cough,
27:14right?
27:15Or you know narrow complex tachycardia
27:18come up with different things to
27:21consider at least in a plan before
27:24you even start the rotation.
27:25I still do that to some degrees
27:27before I'm on my rapid response
27:28senior before I enter the step down,
27:30I'm like alright,
27:31like what do I do in acute like hypoxemic
27:35respiratory failure or in hypernatremia.
27:37When the Ed calls me I'll like give
27:39them like some suggestions initially,
27:41so giving an initial like framework.
27:43So again it makes it more multiple choice.
27:46All right.
27:48And I have a few more cases,
27:49I've talked a bunch.
27:50I'm starting to get winded so I'm going
27:52to need your help as we go through
27:53these as well. I've tried two times,
27:55but I I think we're warmed up now
27:57to have some more participation.
27:58All right, case one, so remember we
28:00have our four steps data collection,
28:02problem representation,
28:04differential diagnosis and management.
28:07I'm actually end up with a
28:09couple abdominal pain cases.
28:10These are from clinic.
28:11But I want you all to read over this case.
28:15I'll read it as well because I know some of
28:18you all are on your phone and in the chat.
28:21I want you to tell me which one of those
28:24four steps Data problem representation,
28:27Differential and management.
28:29Where you think this learner is struggling
28:31most or where you would consider
28:33doing some teaching for this learner.
28:35All right,
28:36deep breath getting my pretending
28:39I'm a resident or student right now.
28:42All right,
28:42we have a 17 year old girl who
28:44reports A few days ago she started
28:46experiencing sharp pain localized
28:48to her left lower quadrant.
28:50The pain rates as intermittent
28:51lasts for a few minutes.
28:53She's had associated nausea,
28:55but decide denies any fever or chills.
28:57She doesn't have any past
28:59medical history of note.
29:01She's a junior in high school and
29:03she denies any substance use.
29:06Her vital signs were notable
29:08for blood pressure 116 / 82,
29:10heart rate 80, respiratory rate 18,
29:13oxygen saturation 99% on room air.
29:16When I saw her, she was uncomfortable.
29:19Heart rate was regular, lungs were clear.
29:22There is moderate tenderness
29:24in that left lower quadrant,
29:26but no guarding rigidity,
29:29no CVA tenderness and otherwise.
29:32Her exam was unremarkable.
29:35So I'd say this is a previously
29:37healthy 17 year old girl who presents
29:39with acute left lower quadrant
29:41abdominal pain associated with nausea
29:43found to be in moderate distress.
29:45Most likely this is caused by a UT I
29:48I looked at urinalysis, urine culture,
29:51CBC, and maybe start keflex.
29:54We could consider an ultrasound or CT,
29:56abdomen and pelvis if the symptoms persist,
30:00so bear with me.
30:01I'm just a humble internist trying
30:03to design cases for Pediatrics,
30:05but if you got this presentation
30:08in clinic regardless of specialty,
30:11where do you think this learner
30:13is struggling most?
30:15Again, feel free to throw it in the chat.
30:17Or I mean yourself.
30:23Also, So poor Panina doesn't have to
30:25keep answering all of the questions right
30:27now. Although fantastic educator
30:28and we would love to hear from her.
30:30Well, I'm bound to be really wrong,
30:32so I don't want to embarrass
30:34myself. Too bad.
30:36Please. That's right.
30:37Participation will be can help.
30:38Panina. That's right.
30:39And me too. I beat it.
30:42Where is this learner
30:43struggling the most?
30:46Can you go back? One slide?
30:47Just just a reminder of the
30:49four categories. Of course,
30:50Of course. There you go.
30:53Thank you. All right.
30:56Carol puts in differential diagnosis.
30:59Catherine puts in data collection.
31:02And for what it's worth,
31:04all of you know there are
31:06so many different options.
31:07You could pick any four of these
31:09steps and your learner would be like,
31:10wow, that's an amazing teacher.
31:13But if you want to get your
31:15most bang for your buck,
31:16I would agree with the chat,
31:19Catherine, Carol and Rushka about and
31:21Ada about both data collection and
31:24differential diagnosis for this case.
31:27So going back to this case,
31:29so data collection,
31:31incomplete history and physical,
31:32and there's some strategies
31:33that we could do to teach it.
31:34But again, so this patient is
31:36coming in with abdominal pain,
31:38why differential?
31:40But you know they they
31:42don't mention much about.
31:44For example,
31:45if this is a UTI,
31:46you didn't tell us you you told
31:48us that there are any fevers and
31:50chills when she's on history.
31:52But any dysuria hematuria, right?
31:55That'd be nice to know in the physical
31:57exam missing just like a temperature,
31:59that might be helpful.
32:00I'm glad you did a you
32:01checked for CPA tenderness.
32:03But just information that would
32:04be associated with that specific
32:06diagnosis and abdominal pain.
32:08There's like, you know,
32:09plenty of must not miss diagnosis as
32:12well to be including in there too, right?
32:15Some that are more like GI pathology,
32:17some that are even gynecological.
32:19Speaking of history,
32:21getting more about what is her like
32:24sexual history as well would be important.
32:26So for those different steps.
32:28So First off,
32:29for data collection,
32:31some teaching strategies that
32:32might be helpful.
32:33Kind of reinforcing what we talked about
32:36before observing their physical exam.
32:38Probably not as important in this one,
32:41but maybe doing like a hypothesis
32:43driven interview.
32:44Sometimes I'll literally stop my my,
32:47my modeling of interview.
32:48Be like now I'm gonna ask my UTI questions,
32:51now I'm gonna ask my ovarian torsion
32:53questions and it doesn't have to be
32:55that quite that prescriptive but
32:57even showing them how you would
32:59ask specific questions as well.
33:01As far as the differential diagnosis,
33:03this is obviously very limited.
33:04This is an interesting one because
33:06we don't know right what the
33:08differential diagnosis is right Like
33:10we just know what number one was
33:12but maybe this learner had other
33:13diagnosis they were considering
33:15that's when that more open-ended
33:17questioning can be really helpful
33:19like Penina was mentioning just saying
33:21like what else could be going on.
33:23No, like UTI,
33:24that's should definitely be on the
33:26differential but or even numbering
33:28like what are two or three more
33:30diagnosis that could be going on
33:32for with this patient even if
33:33you think they're less likely.
33:35I'm just so you again you know
33:37what their strike zone is.
33:39Again for frameworks they're different
33:41frameworks that you could provide.
33:43This is a way to make it like
33:45more multiple choice like I've
33:46been saying as well.
33:47So maybe they're just struggling
33:48to come up with the differential,
33:50right.
33:50You ask them what else is going on,
33:51They just stare at you blankly. OK,
33:53Can you think of any other infectious causes?
33:57Malignancy causes, vascular causes as well.
34:01And there might not be them for
34:03every category and you can be
34:04judicious about which ones you pick.
34:05But that way the learner can also
34:07feel more ownership about their
34:10clinical reasoning, right?
34:11They leave the encounter, feel like,
34:12hey, like I did like get some stuff
34:14right even though I need to work on it,
34:16which can be motivating for
34:18them in the future.
34:19So great job team.
34:21You are right on track as far as data
34:25collection and differential concerns.
34:27So for case number 2.
34:32All right, so this is a patient on the wart.
34:36So this is a 17 year old.
34:39A female with extra sickle cell
34:41disease and exchange transfusion.
34:43Twice in the past two years I had one of
34:45the Pediatrics fellows give me this case.
34:47It makes it way too complicated.
34:49I don't know what's going on in this.
34:50Anyway, 3 days of fatigue,
34:52Lightheadedness and feeling dehydrated.
34:54Now with a 2 day history of cough,
34:57nausea, vomiting and upper abdominal
34:59pain radiating to the left shoulder.
35:02She was admitted yesterday to Bridgeport
35:04with worsening abdominal shoulder pain.
35:06Refractory to Dilaudid PCA.
35:09So past medical history.
35:12Sickle cell disease.
35:13The disease with multiple crises.
35:14Baseline hemoglobin of nine Has
35:16a history of insulin resistance,
35:18had a cholecystectomy previously,
35:20Takes ibuprofen and oxycodone
35:22for pain but has not filled.
35:24Hydroxyurea and metformin smokes marijuana.
35:28Daily exam notable for a temperature
35:30of 39 four heart rate 107,
35:33blood pressure 98 / 65,
35:35respiratory over 26,
35:37oxygen saturation of 97%,
35:41General to Kipnik with mild distress
35:46and scleral conjunctural Ichteros
35:49diminished breath sounds a sinus
35:52tachycardia On cardiac exam.
35:54The abdomen was with no organomegaly
35:59but diffusely tender.
36:00I mean skin was notable for
36:02pallor and jaundice.
36:03So on labs white blood cell count was 9,
36:06hemoglobin 5.7,
36:07platelet 77 and neutrophil
36:1077% with retic count of 4%.
36:13Mild spot was positive,
36:14total ability was 6,
36:16but the CMP was otherwise normal
36:18and there was a normal lipase.
36:20So again in assessment we have a 17
36:24year old girl with a past medical
36:26history of sickle cell anemia
36:27who presents with abdominal pain
36:29history also notable for fatigue,
36:31light headedness and shoulder pain.
36:32Most likely secondary to
36:35splenic sequestration.
36:36As her variant hemoglobinopathy continues
36:39leads to continued risk of events.
36:41Pneumonia is also possible,
36:42giving the fever to get me in tachycardia,
36:45less likely pancreatitis.
36:47I'd probably give pain medications
36:49and start antibiotics.
36:51So one thing I think is interesting
36:53as I buy time for you all to put
36:54this into the chat or feel free to
36:56like unmute yourself about what
36:57you think is going on with this
37:00learner is that often time I do
37:04direct observations of faculty
37:06and specialties outside of my my
37:09area of internal medicine.
37:11And I will say while you do have
37:13to have a some level of knowledge
37:16in these conditions,
37:18for the most part I feel like I can
37:20get a pretty decent sense of where
37:23a learner's clinical reasoning is
37:25simply by how they present the information.
37:27That's why it can be helpful to
37:29have frameworks to some degree.
37:30So hopefully,
37:31even though this
37:37a little bit more specific of a case,
37:40y'all might be able to share where you
37:41think this learner is struggling the most,
37:43even if you might not necessarily know
37:46how you would teach it given the feel.
37:49So what do you think is
37:50going on with this learner?
37:52Where are they struggling?
37:58Yeah, all four. I agree there's
38:02a lot to unpack in this case.
38:05One thing I'll I'll say with that in mind,
38:07and everybody's different and
38:09sometimes you just absolutely need to
38:11clarify information from a learner.
38:13But I will say usually I'll wait
38:15until the end of a presentation
38:18before I make any comment,
38:19just because there could be so many
38:22different things along the way
38:23that you might want to talk about.
38:25And if you have limited time,
38:28sometimes you can only pick
38:29like certain specific things.
38:38All
38:41right, so we have votes for all
38:44categories, anything in specific.
38:46If you saw this learner what you
38:48would consider working with for them?
38:52So a couple of things that I would
38:54I I feel like would I'd want to
38:57work with this learner on one is
39:00their problem representation.
39:02So this is a very common phenotype of
39:05problem representation that we'll get
39:06in our learners where they essentially
39:08just repeat the history right?
39:10How do they distinguish signal from noise?
39:12We don't know.
39:13And it's like it's really frustrating
39:14because it makes the presentations
39:16a lot longer.
39:17But it is important that they need
39:18to find ways to synthesize it, right.
39:20And so, you know, sometimes in notes,
39:23I like limit learners.
39:24I mean, obviously we have like
39:26really complex patients sometimes.
39:27But all right,
39:27you tell me what are like the two or
39:30three most important comorbidities we have,
39:32All right.
39:32Like if you have to pick two symptoms,
39:34what would the,
39:35what would you include,
39:36what are the two labs you would
39:37want to or how would you categorize
39:39this patient's timing? Right.
39:41It's abdominal pain, like is it acute,
39:43is it sub acute, chronic, right.
39:45Those are whole different
39:46differentials for us as well.
39:48So that would be one thing for
39:51problem representation as well.
39:53The other one is management, right?
39:56And now to be fair,
39:58it's like I don't know how many
40:00learners you would have who
40:02would actually just say give pain
40:03medications and start antibiotics,
40:05maybe an early clerk.
40:07But still y'all know that there
40:09are plenty of times that we need
40:11our learners to specify more.
40:13And so that's when it can be helpful
40:15to give them a prompt like that
40:16management script template that I mentioned.
40:18Like, OK,
40:19you don't have to say everything,
40:20but any tests you want to do,
40:22any diagnostic tests.
40:24Any medications we want to give,
40:26when should we follow up with this patient,
40:28right or how long do we want to
40:30wait until we get the next BMP.
40:32So asking for a little bit more clarity
40:35around the management can lead to really,
40:37really helpful discussions.
40:39And honestly,
40:40that's one of the biggest parts of
40:43our our clinical reasoning on rounds,
40:45right is really helping our learners,
40:47especially the more senior learners
40:49hone their testing and treatment thresholds.
40:51For those early learners,
40:53it's just figuring out what you could
40:55do and for more senior learners
40:57it's figuring out why you should
40:59do one thing versus the other,
41:00knowing that there isn't necessarily
41:03a single right answer.
41:06So with that in mind,
41:07that's a that's a wrap for me.
41:09I wanted to give you all some break.
41:11I know again that this is a a
41:14busy time for Y'all in between,
41:15but I'm around for questions.
41:17But again to summarize today we talked
41:19about a few things in clinical reasoning.
41:22One,
41:22we talked about why we do this,
41:23which is diagnostic error.
41:25We gave some language,
41:26we talked about some theories as
41:29well and really mentioned about the
41:31situativity theory, the context.
41:33I I probably didn't talk about
41:35that as much as we should but you
41:36know that
41:37clinical reasoning is different
41:38in different environments, right,
41:39a busy clinic versus having like
41:41all the time in the world if you're
41:44on like a slow day of rounds.
41:46We talked about strategies to assess
41:48and teach clinical reasoning really
41:49based off of those four steps,
41:51data collection, problem representation,
41:53differential diagnosis and management,
41:55and broad categories of teaching
41:57open-ended frameworks, modeling,
41:59and providing some directed
42:02learning to do afterwards.
42:04And then we had a few fun examples,
42:07a lot of which I'm not completely sure
42:09I knew as a general adult internist.
42:13So yeah, sorry,
42:13I have some questions here,
42:15Feel free to throw them in the chat if
42:16or if you are ready to head out as well.
42:21I just want to remind everybody
42:22what the code is for today.
42:24It's 42047 for CME credit,
42:29but for Carol, as Carol asks,
42:32brings a great point.
42:33I don't know if that's a question
42:35or just like something we
42:36could work on for management,
42:38but I think this triage,
42:40this disposition is huge actually.
42:43So particularly,
42:44we know that diagnostic errors happen
42:46most in these transitions of care.
42:48So really making the decision,
42:50can this patient go home?
42:51Do they need to go to the Ed?
42:53Or if they're in the Ed,
42:55should they like, be admitted,
42:57Should they go home or should we like wait
43:00a little bit longer while they're here?
43:02That can be really helpful,
43:03particularly to model with our learners
43:06to unpacking why that might be so.
43:09OK, great.
43:12Thanks so much.
43:13Sorry, can I
43:14ask a great question?
43:15So for the four different categories
43:17of reasoning that you mentioned,
43:19do you feel I almost feel
43:21like they are interrelated?
43:22For example, if the problem is
43:23not and you might have said this,
43:25I might have missed the
43:26beginning of the talk.
43:28But is that is that true?
43:29Like how I would think about it?
43:31Like I would think like if the if
43:33the learner is struggling with
43:35the problem representation not
43:36he won't he or she won't come
43:38up with the right differential
43:40and hence that will affect the
43:42next steps of management
43:44exactly exactly. All of this is intertwined
43:46and as we've like developed like the side.
43:48So I will admit I'm a splitter and not a
43:51lumper and my colleagues and Co researchers
43:53in the field feel like differently about it.
43:56But I think part of why I do try and split
43:59sometimes is because then at least you have
44:02a better idea of what you can teach, right?
44:05Because then if you're teaching
44:06like generally it can just be hard,
44:08there can just be so much to cover.
44:10But you are absolutely right,
44:11like combine the steps right now.
44:13Hypothesis driven physical examination,
44:15right? You need a differential
44:17to know what like to do,
44:19what questions to ask.
44:21Or for example,
44:23people say a problem representation that
44:26leads towards a differential diagnosis.
44:28They pick specific information, right?
44:30Because they want you to think of these
44:32diagnosis and I will say in like that model,
44:35like a lot of a lot of researchers or
44:40like educators will will post it that that
44:44hypothesis generation actually is the
44:46first step that that's where you start.
44:48And from there you have data collection.
44:51So obviously,
44:52they definitely blend together.
44:54But if it can be helpful just to pick
44:56like 1 area specifically where you want
44:58to focus at least for a given encounter.
45:01Does that make sense?
45:02Yes. Got it. Yeah,
45:03that's what I was trying to like.
45:04In my mind, trying to differentiate.
45:06It makes sense to pick one to
45:07do an intervention in a small
45:09session with this with a learner,
45:10but OK, thank you.
45:14And I will say a lot of you teach medical
45:16students or I think they even do this in
45:18graduate medical education with the Epas.
45:19Like all of these qualities because
45:21map out to the Epas as well.
45:24Like if you go from like a level 1 to a level
45:275 from like pre entrustible to entrustible,
45:29if you like read the fine print,
45:31they essentially include all of these.
45:33They might not say problem
45:35representation because that's probably
45:36too dense of lingo when they were
45:38developing these at the double AMC,
45:39but all of these map out to that.
45:41But at least it's a little bit easier
45:42than saying like I think this learner
45:44has a good clinical reasoning or bad
45:45or I trust them or I don't at least
45:47you can like teach them a little
45:48bit more specifically that way.
45:52And in terms of promoting hypothesis
45:54driven history taking and physical exams,
45:57sometimes I think if they're going
46:00into a room like an outpatient to see
46:02a patient with a certain chief concern
46:04like right off the bat, I'll say to them,
46:06what are you thinking right now just
46:08hearing that age and that chief concern,
46:10what are some things you're thinking about?
46:13And then that helps them say, oh, OK,
46:15well I'm going to go in and you know,
46:18just this, you know,
46:19ask them questions about this,
46:21that or the other because I'm
46:22already thinking of that.
46:23So that's the other thing
46:25that's sometimes is helpful to
46:27them. That's great, Carol,
46:28and I do that as well.
46:30I can't say enough that clinical
46:32reasoning is iterative and I don't
46:34think it necessarily promotes
46:35premature closure and anchoring.
46:37In fact, I think if you keep learners
46:39broad earlier, it can be helpful.
46:40But asking them to start thinking,
46:42OK, like what diagnosis he's heading
46:43into this patient with shortness
46:45of breath do you think about,
46:46because then they'll know like, oh,
46:48I want to ask questions about PE or
46:50want to ask questions about pneumonia.
46:53That's great.
46:53I love that you do that with your learners.
47:08Any other tips and tricks from
47:10the group that you want to share
47:12stuff that Y'all are doing already,
47:14which I'm sure is terrific.
47:18All right, well, I'll take my leave.
47:19Thank you so much for having me. Panina.
47:22Thank you everybody for joining.
47:24It was a blast.
47:25And feel free to reach out if you ever
47:27want to talk about clinical reasoning.
47:29It's my favorite thing to do.
47:31Thank you so much for joining
47:33us in Pediatrics, especially
47:34since I know you're on service.
47:36So thanks so much.
47:37We all really appreciated it.
47:40Bye all. Bye.