Time Stamps

  • 00:25 New features in this episode
  • 03:20 Data point 1: Early hypothesis generation?
  • 11:18 Data point 2: Pertinent negatives?
  • 13:28 Data point 3: Representativeness heuristic over-represented in medical education?
  • 19:20 Data point 4: Economical hypothesis generation by expert clinicians
  • 22:10 Data point 5: Diagnosing diagnostic error
  • 27:11 Data point 6: Moving toward a final diagnosis
  • 29:50 Case resolution

HumanDx interactive case: https://www.humandx.org/o/16ilqkaxch8ezqzt113o45n58

Show Notes

  • Expert clinicians tend to generate early diagnostic hypotheses in response to just a few initial cues about the patient and their illness (e.g. the chief complaint).
    • To what extent this behavior in experts is involuntary or voluntary is debatable, and likely contextual. It is clear, though, that experts complement this early hypothesis activation with careful subsequent scrutiny, in the process of hypothesis evaluation/validation.
  • It is important to differentiate between “no past medical history” and “no known past medical history.”
  • Expert clinicians do not simply possess an expansive knowledge base, but also organize this knowledge in hierarchies that complement one another. In other words, the master diagnostician is not only an expert in diseases, but also in symptoms and syndromes, in pathophysiology, and in epidemiology and local patterns and prevalences of disease.
    • Learning to recognize the most representative features of a particular disease is a necessary but by itself insufficient method of clinical learning.
  • Spoiler alert! Gastrointestinal malignancies (gastric, pancreatic, colorectal) are among the cancers most strongly associated with venous thromboembolism.
    • Trousseau’s syndrome refers to migratory superficial thrombophlebitis as a paraneoplastic manifestation of cancer. Coincidentally, Trousseau developed this syndrome himself and was later diagnosed with gastric cancer.
  • Spoiler alert! Owing to a lack of data, there are no firm guidelines in the United States for routinely screening for gastric cancer in immigrants from high-prevalence regions (Latin America, Eastern Europe, and East Asia). As there is some data to support the effectiveness of universal routine gastric cancer screening in certain high-prevalence countries (Korea, Japan, Chile), it is not unreasonable to consider offering screening to first-generation immigrants in the US on a personalized, case-by-case basis.


JOHN Hi everyone, John Hwang here. And welcome back to another episode of Hoofbeats, where we challenge you to solve diagnostically difficult, real-world cases alongside experienced clinicians. As always, I’m here with my partner, Cindy Fang.

CINDY Hi everyone, this is Cindy Fang.

CINDY We actually will be doing a few things differently for this episode. First change: In past episodes, we’ve been presenting you full cases, with all the details available to the original providers, usually in one continuous narrative.

JOHN Over the next couple of episodes, though, we’re going to be using cases that have been stripped down to just a few data points. We’re thinking that presenting these cases reductively will provide a different perspective on the diagnostic process. For this case, we’ll challenge you to solve a case of a man with chest pain, presented in just six chunks of information, provided sequentially.

CINDY They’re still pretty meaty chunks, actually. And as usual, we encourage you to pause during the episode after each bit of information to collect your thoughts, before hearing what our discussants have to say.

JOHN And if you’d like to tackle this case just like the discussants are, chunk by chunk, check out the HumanDx app on your mobile device, or go to the HumanDx website — we’ll include links to this case in the shownotes. And actually HumanDx is dedicating next week to highlighting chest pain cases on their global morning report, so if you tackle the entire week, you should be an expert on this in no time.

CINDY Second change: For this episode, we sat down with two discussants rather than just one.

JOHN And we gave them this case cold — they didn’t get any information beforehand, so you’ll hear their reactions live. Again, we wanted to capture that stage in the diagnostic process, something we haven’t done before.

CINDY CoreIM listeners, meet Drs. Patrick Cocks and David Stern. Both are internists and senior faculty here at NYU.  John, I must thank you for all your incessant stalking and harassing, I mean, convincing, to get them to appear on this podcast.

JOHN No idea why. Dr. Stern has more clinical experience than everyone else put together in the room. And also the administrative influence to black out our unflattering depictions of him.

DAVID STERN [To Cocks] Have you done one of these yet?


DAVID STERN Well, whatever. We can decide that we’ll just totally bail. Like we’ll say we don’t want this to be right… We have editorial…

JOHN You have complete institutional leverage over both of us!

DAVID STERN Oh, excellent [laughing]. Okay. So we have nothing to worry about.

PATRICK COCKS We’ll have security take it away on your way out.

JOHN Meanwhile, Patrick was our program director in residency. So I kind of still need him for a letter of rec if I apply to fellowship.

CINDY We figured, hey, we have two wise discussants, two voices, John and I don’t have to talk at all.

JOHN Also in all seriousness, both Dr. Cocks and Dr. Stern are leaders at our institution in medical education. And so you’ll see much of what they had to say about the case was metacognitive and from the perspective of teachers thinking about how reasoning ought to be taught.

CINDY Should we get started? John, let’s hear about your patient!




JOHN Yes. So. Bullet point / bolus number one. This is a 47 year-old man with chest pain. It’s substernal, it’s aching, it’s worse when he walks, goes away after a while when he rests. And while it started two months ago, over the last few days it’s brought on by the slightest movement, and now it even hurts to breathe.

So, decide what you think about that. And when you’re ready, here is Dr. Cocks giving his immediate reaction to this info, followed by Dr. Stern.


Musical Interlude ♫


PATRICK COCKS I’m hearing a youngish man with, there are characteristics of his pain, which to me sounds like progressive angina and brought on by exertion relieved with rest rapid progression of symptoms in the past several days.

DAVID STERN Yeah, I mean, the way he presents the case, first of all, he’s already done some chunking for us and helped us, like he’s already said that things like the pace and the rate, right, the rate of progression of the illness. For some reason I had this automatic stop after age and gender. Um, so you say 47 year old male and then chest pain. And my brain is already, I’m working pretty hard. On a few things. Uh, you know, and, and 47 to me is just. I understand, I understand you want this to be chest pain from a probabilistic perspective, but, um, I can’t get away from. I can’t for some reason, again, I have no idea why a, a pericardial effusion shows up in my head. Um, but it just, it, it, it does. Um, uh, maybe cause the guy’s just younger and we don’t have any past medical history or anything.

JOHN Cindy, if you remember, this actually immediately led to a bit of a debate about the chief complaint. Dr. Stern is saying hypothesis generation begins with the chief complaint. Age, gender, primary symptom, duration. He’s showing his old school roots. Basically what I was told as a med student by every older gray-bearded attending. Usually while sternly looking at me down his glasses.

CINDY Yeah, while Dr. Cocks pushes back against this a bit.

DAVID STERN I often find that I often really struggle hard with the, the sort of stepping away from my anchor. I just, I like and I hear you doing that. You saying, okay, I am still in that space of undifferentiated chest pain and I’m living there and holding on to that even though like for me it’s hard.

PATRICK COCKS And because I was pushing back on you a little before because I actually like to hear the whole chief complaint and HPI because I’m fearful in that process of generating, embracing our type one thinking as we put forward the age and demographics, we are going to start to anchor and create a story that meets the hemorrhagic pericardial effusion from tuberculosis disease that he got there. And so I think it’s an interesting that way with which we produce this.

CINDY We have a tendency to seek confirmation. The fear here is that simply having a working hypothesis after a chief complaint is enough to poison your subsequent information gathering.

JOHN Do you try to actively fight it or have you learned to let your mind wander before you bring it back course? Since you were saying you don’t actively fight system 1?

DAVID STERN I don’t fight system one. What I do is I will always write it down. Because you know, in the seven plus or minus two world, I’m a five, you know what I mean? Sorry, there’s a wonderful paper from 1956 [John: Oh, it’s about working memory.] It’s about working memory, that you can carry seven plus or minus two items in your brain at any one time. And I’m a five, I’m at least one standard deviation below the mean, and I cannot, I can’t remember all those things. And so when I think of something, I always write it in the margin. So it is very helpful for me, uh, to simply write pericardial effusion in the margin, draw a line under it, and then keep going. And I know it’s there. I don’t have to worry about forgetting about it and don’t have to keep it in my working memory. Then now have a longer conversation about all the other things that it’s far more likely to be.

JOHN My rational brain wants to agree in Dr. Cocks. Get all the information, then get to work. But listen to the way Dr.  Stern talks about this. His early hypothesis generation — at least for this case — appears to be driven by his fast thinking; he describes it as involuntary, almost intrusive. It’s like he has to acknowledge pericardial effusion, for all the flaws in that theory, before moving on with this thinking. It’s like having a kid: It says what it wants to say. Most of the time it’s wrong. And you’re not going to be able to move on with your day unless you patiently listen to whatever it is that they’re saying, even if it’s ridiculous. He’s just going to keep saying it louder. The payoff in listening, of course, is that every now and then, it manages to stumble on something profound.

So I would question — is it that Dr. Cocks is actually not coming up with early hypotheses? Or is he simply not indulging his brain in voicing them aloud? In hopes that it allows him to move forward more objectively.

CINDY Some of this debate is probably an artifact, though. Remember the context here. Right now we are learners trying to improve our reasoning skills, not clinicians trying to take care of a patient.

JOHN Right, Cindy. The way we’re presenting this case, prompting hypothesis generation after each discrete piece of data, that is a valuable way to force learners to practice certain skills. If I simply tell you, 30 year-old man with diarrhea, the learner has to remember to ask questions at whether that this might actually be hyperdefecation, or fecal incontinence. They need to remember to ask about HIV status, and if that’s not provided, to ask about risk factors like a history of STIs or anoreceptive intercourse.

CINDY Whereas if I present the entire case to you, where I simply say, the patient reports 5-6 watery non-bloody bowel movements per day, and his social history is notable for a history of multiple stds… those muscles aren’t being engaged in the same way. More recall, less retrieval.

JOHN Exactly. But this kind of exercise isn’t really representative of the real world, where patients don’t always come as stepwise data points. There are benefits in using this strategy with real patients to be sure, but as Dr. Stern says, it’s contextual.

JOHN So when you say that you stop after the chief complaint in the classic sense, is that something you do for every patient in reality or that is also a product of the academic context that we’re in right now?

DAVID STERN  Uh, I do it anytime I, anytime that I’ve got a diagnostic dilemma, you know, when, when, when I’m staffing cases in primary care clinic, I’m not doing that. Um, because that’s not the way they come. They come with a long past medical history. I probably know them already, you know, I’ve seen them in clinic with the residents for years and so there’s already a fair amount of background. But the cases that come in report, or the cases that come up from the emergency room, or the cases that come up to the ward and get presented for the first time as a, you know, this is a patient who got admitted for chest pain. Those people, I definitely stop after 47. I stop after male. And I stopped after chest pain.


Musical Interlude ♫


JOHN Okay, second data point. Which admittedly is more of an absence of a data point… the patient was previously healthy, with no medical or surgical history, and he takes no medications.

CINDY This didn’t really tip our discussants’ needle one way or the other.

PATRICK COCKS I mean, I think that that piece of information continues to place us… on guard is probably not the right term, but again, you used the term that you’re working really hard, like working really hard to tease out where is the signal where the noise, and the fact that he’s not carrying traditional risk factors that we may attribute to coronary artery disease to me it makes me think a little deeper and expand my differential diagnosis much, much deeper. If he was a smoker with diabetes and hypertension and his father was 47 and I’m going to lead us down a very different route than where we have now. So still represented to me as undifferentiated progressive chest pain.

JOHN A common axiom I’ve heard from experienced clinicians: pertinent positives are much more helpful than pertinent negatives.

CINDY Or non-pertinent negatives.

JOHN Yes. Med students, please remember that the next time you admit someone with altered mental status, and you started with canned lines like “regular rate and rhythm, no murmurs rubs or gallops.”

CINDY I have to mention my personal bias from training at Bellevue, where a lot of patients do not have access to the healthcare system regularly. When someone comes in and tells me he does not have any medical problems, I wonder if he actually goes to his PCP yearly and receives a clean bill of health, or he actually has decades of HTN/hyperlipidemia but is just never diagnosed because he never sees a doctor? Going back to the whole “listen to your patient, don’t trust your them” thing again…

JOHN Fair. So, I will clarify, he does get sporadic routine medical care, as young patients are wont to do. So let’s say he has no known medical history, as opposed to an unknown medical history.

JOHN Third data point. He was born in Mexico and immigrated to the United States 30 years earlier.

CINDY This led the discussants to think about TB and TB-like diseases in that part of the world. Which I think a lot of us would also be considering reflexively.

JOHN Cindy, a few weeks ago we have a listener write in, questioning whether it’s really a good idea for shelf and boards questions to reinforce this kind of reflexive association.

CINDY You mean, “see immigrant from Latin America (endemic regions), think TB?”

JOHN Yes. Or how on exams or review materials, the patient with bilateral hilar infiltrates and noncaseating granulomas on biopsy is invariably African American. The patient with polyarthralgia, evanescent rash and fever of unknown origin is a young woman. Our listener asked, are we doing a disservice to learners when we emphasize these associations repeatedly? Are we finishing medical training with too much of our knowledge base in the form of these associations? Taking this case in point, the vast majority of patients who immigrated from Mexico 30 years ago are not going to have TB. Yet seven years of medical training have made that association in my head involuntary, and again, almost intrusive.

CINDY I think the rationale is, learn the classic examples of diseases first, then you’re positioned to learn the rest.

JOHN Maybe. No doubt these associations are helpful to the clinician. But one wonders how much medical education should revolve around developing representative conceptualization of diseases, particularly when it comes to ethnicity or gender.

PATRICK COCKS How you teach that, I guess, is more of the metacognitive approach to diagnostic reasoning. To acknowledge that an individual who comes from Mexico is a trigger that is sometimes deeply rooted in our medical education system, that triggers TB. 30 years ago is a pretty long time to have that, to have him reactivate, unless we are now about to find out some other elements of his history that may predispose to concurrent illness or drugs.

DAVID STERN I sort of see multiple ways of — It’s a really tough question — multiple ways of teaching and learning these things, none of which are perfect.

I’m particularly enamored of symptom based education. It’s not a very common way of teaching medical students. Um, they did it at University of British Columbia probably 20 – 30 years ago. They revised the curriculum based on, I believe there were 105 ways that the human body can be abnormal. Dyspnea, chest pain, hyperkalemia.

JOHN I thought there were just ten.

DAVID STERN  Ha. So I think. And then there was some argument about whether there are really 108 or 105… Anyway, there was this and then they devised a curriculum around that and so that students were learning the diseases, learning medicine based on how patients present: knee pain, chest pain, shortness of breath. And then you dive into it from the physiology of the complaint, to the biology of the disease, and even into the sort of molecular genetics. I’m not sure whether it really worked well. It certainly didn’t catch on so much. But um, medical schools have gone from discipline based education to, like, you know, a biochemistry, physiology, anatomy to things like organ-based, to problem-based… a variety of different ways to get at it.

JOHN I’m surprised it didn’t catch on. Organizing teaching and learning by symptom, rather than by disease or organ system, actually makes a lot of sense to me.

CINDY Right. Because patients don’t present with diseases. They don’t come to you and say, I have a problem with my kidneys.

Because my patients definitely come in with the chief complaint of “ i have a problem with my kidneys and I need you to work up my AKI”

JOHN Yeah.

CINDY But I didn’t get the sense at all that he thinks this method of learning is necessarily superior, or sufficient on its own.

JOHN I agree. He talked about how experts have a nuanced understanding of certain diseases because study it from different perspectives — their knowledge is highly organized, compiled not just by diagnoses and disease associations, but also by symptoms, by pathophysiology, by epidemiology and local prevalence, and so on.

DAVID STERN And I think ultimately they all get you into the same space, which is a lot of experience with a lot of cases from a variety of directions eventually fills the gaps in your knowledge and adjusts your probability such that when you tell Patrick, well he’s from Mexico. The trigger of TB pops into his head and then, and then TB triggered other things, right? I heard him say TB and then he was like histo and he’s probably thinking of, you know, a few other things that sort of pop along those lines. Uh, and then he backs off of that with the 30 years, you know, sort of comes and goes and because he’s constantly adjusting that based on his knowledge and experience, so he ends up in a space. You could start in a variety of places and get there and the common denominator for all of them is a lot of cases, a lot of cases.


Musical Interlude ♫


JOHN Fourth data point. On exam, the patient has tenderness to palpation in the epigastric and left upper quadrant regions of the abdomen. Again, Dr. Cocks.

PATRICK COCKS I mean focality is important. I think right? Like one of the aspects of the case, which I did not think of was anatomic processes, whether it be splenomegaly that we now see in that area leading to his progressive exertional symptoms. Right. And so I’m going to anchor actually on now Mexico and you know, could he have hereditary gastric cancer and what we’re witnessing now is a growth within his stomach that’s leading to either physiologic changes into his cardiac output or return or is it truly just he’s uncomfortable and you can’t take deep breaths because we did hear in his chief complaint that he had left side pleurisy.

JOHN By the way, I just want to point out that, this is only the second time the discussants have proposed a specific unifying diagnosis, the first being TB.

CINDY It’s actually notable. We haven’t heard many specific diagnoses at all, period, aside from the early thought of pericardial effusion by Dr. Stern and the brief consideration of TB and TB-like syndromes earlier. I feel like if this was a resident report, there’d be a long DDx on the whiteboard by now.

JOHN It could be an artifact of this setting. Highly academic, two co-discussants, maybe they don’t want to tip their hand and feel foolish. Or maybe this is the natural reaction when presented with minimal data.

CINDY Again, thouh, this is something that has been found in formal studies — and its been observed that expert clinicians tend to have only a small number of working hypotheses active at any given time.

JOHN Yes that’s true, though I don’t think that this observation has necessarily yet been explained. Is it purely involuntary, does this reflect limitations in working memory? Or is this a semi-conscious strategy that experts adopt? Do they constantly reevaluate the top diagnoses on their differential — is this a way of economizing their cognitive bandwidth? I don’t know, and I’m not saying necessarily that we ought to be purposely restricting our differential. But certainly in resident report when we write that 25, 30 item differential on the chalkboard, that’s not an accurate representation of what’s going on — at least what’s going on in the expert clinician’s mind. It’s more of an academic exercise. Or perhaps as Dr. Stern said earlier, a road map of roads traveled, of thoughts thought.


Musical Interlude ♫


JOHN Data point number 5. The patient has bilateral leg swelling. There’s no erythema, edema, or tenderness.


I’m going to anchor on what Dr. Stern said earlier. Could this be pericardial disease? Hard pressed to put the palpable pain that he has. But if it were to be constrictive pericarditis or pericardial effusion affecting output, wouldn’t you expect to see some JVD? That’s right. And we don’t see that at all. Right. So it’s very helpful. Again, coming back to the anatomic process of thinking about how could this individual have a lower extremity edema in the absence of right sided heart failure.

Uh, you said, what did she say? Lower extremity edema? Yeah. He said no edema. I thought I heard it’s leg swelling without edema. But again, that’s, you know, it’s one of those things that you hear that and there’s not a whole lot of things that come to mind when you say leg swelling, no edema, right? I mean there’s two things that immediately popped to my head and there’s probably more because Patrick’s better than I am at this step, but… and why he called me Dr. Stern, I’m really not sure. It’s like so respect. Just because I’m the oldest person in the room, it doesn’t mean that I deserve the respect. The first thing that always comes to my mind, and maybe it’s because people always forget it, it’s thyroid disease, right? Is this hypothyroidism and this is really doughy edema that’s not edema, that’s really from thyroid disease. Um, which, you know, I like because it fits some other elements. The other thing is DVT, bilateral DVT, which was really weird, but you’ve got some sort of abdominal complaints going on and so for him to have some sort of obstructive something… I would expect that to be edematous — but you know, DVT is probably early, it can cause swelling. He could just have fat legs. Lipidema. I like that. Lipedema or lymphedema, we don’t know. What else do you, what else does that trigger for you?

PATRICK COCKS You’re, you’re right. I, I was not listening. So my mind, my data collection process was faulty right there.

JOHN Just to interject here Cindy, I know he’s being a little tongue in cheek (“data collection process is faulty” sounds like an error message). But knowing him, he’s deliberately modeling an important metacognitive skill here: diagnosing one’s own diagnostic process.

CINDY Right. Simply recognizing you’ve made a diagnostic error is half the battle. It requires conscious follow-up and humility. But equally important is localizing the lesion. Where did I go wrong?

JOHN Right. Did I make a faulty assumption when evaluating my hypothesis? Was there a gap in my knowledge about a particular test? Did I misinterpret the data? Or, as in this case — and as happens all too frequently in the real world — did I simply not gather enough of the right data?

CINDY So with that mistake cleared up, Dr. Cox considers a few more hypotheses.

PATRICK COCKS It would be important for me to examine and see if there are other signs of portal hypertension, whether it be from a venous occlusion below the diaphragm and the liver, whether it be from a cirrhosis, you know, whether… his symptoms that… the focality of his pain to me does not fit with how I think about hypothyroid or myxedematous process. Could he have retroperitoneal… cancer or something that’s leading to progressive lymphedema in his lower extremities? Sure. But I would expect it to be a little weeping and we’re not hearing that. I’d expect it to be much more chronic. We haven’t heard anything about him reporting his legs have been swelling over a period of time. Um, so that, that, those are thoughts that I have. Hasn’t really helped me formulate a conclusive next diagnostic step, so to speak. But I am thinking more anatomically as we’ve heard this.

DAVID STERN But where are you like, like where are you in terms of diagnosis? Because I heard you throw a bunch of things. Are you at a place where you are thinking about diagnoses? Or are you still like in the, I want more data before I begin to come up with a list, sort of thing?

PATRICK COCKS I’m thinking… So coming back to this idea, it sounds like heart failure without right sided heart failure. I’m thinking of processes below the diaphragm that can lead to swelling of his lower extremities. The focality of his pain to me makes me think about a mass.

DAVID STERN You mean his discomfort on palpation?


DAVID STERN And not the exertion?

PATRICK COCKS So I think that this is again coming back to like our experiences and hearing and seeing cases is that… Anemia can lead to dyspnea on exertion. Valvular processes can lead to dyspnea on exertion. Pulmonary hypertension can lead to lead to dyspnea on exertion. Neuromuscular problems can lead to lead to dyspnea on exertion. Myopathies will lead to dyspnea on exertion. So while we jumped to the young man with progressive chest pain with exertion thinking about coronary artery disease, I’m thinking much more about that being a secondary process perhaps going on in his abdomen, right.

JOHN The sixth and last data point: His CBC. He is severely anemic, with a hemoglobin of 3.4, MCV 58, a WBC count of 12, with slight neutrophilia, otherwise normal differential, and platelets of 335.

PATRICK COCKS I think we have a reason why he… It explained, a symptom that the subacute to chronic progressive dyspnea; his hemoglobin is three. I mean, it doesn’t, we don’t know what it was three months ago, but why does he have a hemoglobin of three days? The more relevant question now…

DAVID STERN And why isn’t he more symptomatic? I mean three is really low.

PATRICK COCKS That’s a great point. So that to me argues that this process is going a lot longer. Right? So he passed his stress test, so, uh, and that he, this has been going on for a long enough time that he can upregulate the counterregulatory hormones to the reduction in his red blood cell mass. So this is now a chronic process. That’s a great point. Like that triggered… This is now a chronic process. He didn’t go from seven to three overnight, right?

DAVID STERN And the MCV of 58. I mean, that’s just, to me, that’s the chronicity that you’re talking about. And was like, would it be some iron deficiency is a part of this. That’s where I am.

PATRICK COCKS So what would you want as your next steps?

DAVID STERN Uh, imaging is what I am interested in

PATRICK COCKS Um, but I think you’re right. We Would look into his abdomen for any mass lesions, primary splenic processes, diffuse retroperitoneal adenopathy.

DAVID STERN Yeah, or something… I wrote portal vein thrombosis at some point here. Abdominal imaging was where I was heading as well. I wouldn’t be surprised if someone wanted to get a scan for a DVT, or someone else wanted an echo.

JOHN Alright Cindy, that’s all the information we’ll be hearing about this case. To recap: 47 year old man, an immigrant from Mexico, with progressive exertional chest pain with a pleuritic component, epigastric and left upper quadrant tenderness on exam, bilateral leg swelling and severe microcytic anemia.

CINDY And our discussants have reached something close to a consensus. So, do you agree with them? What’s your diagnosis? What would you do next? Finalize your thoughts and we’ll compare notes after the break.


Brief music pause — a few seconds


CINDY So John, you want to walk us through the resolution of this case?

JOHN Sure Cindy. So the next test was indeed a CT of the chest and abdomen, which showed a large mass that appeared to arise from the stomach and abutted the spleen and left hemidiaphragm. It wasn’t timed as a contrast study, but incidentally showed multiple left-sided lobar pulmonary emboli. That led to a duplex which confirmed bilateral DVTs. The EGD showed a large friable mass in the gastric fundus, and biopsy confirmed gastric adenocarcinoma.

CINDY Unfortunate… So our discussants were pretty much on the money.

JOHN Yeah.

CINDY Seems like a solid bread-and-butter case. Any particular reason you chose it?

JOHN Yeah, a couple of reasons. I thought it was striking how throughout the course of this illness, the patient never had a complaint of abdominal pain, or nausea, vomiting, weight loss, early satiety, any of that. I mean, he was tender if you mashed, but he had no abdominal pain per se. Denied melena or hematochezia. Instead he’s presenting with secondary phenomena of his cancer — secondary myocardial ischemia from severe anemia induced by chronic GI bleeding. And I think he had pleurisy from pulmonary emboli which might have contributed to that acceleration of his symptoms.

CINDY Worth pointing out that gastro-intestinal cancers are among the most strongly associated with thromboembolism, especially pancreas and colorectal, but gastric is no exception.

JOHN Right. You might remember Trousseau’s sign from medical school — migratory superficial thrombophlebitis heralding the onset of cancer. Apparently Trousseau diagnosed himself with his own sign before dying of gastric cancer two years later. I didn’t know that, probably because I perennially skipped medical jeopardy in residency.

But anyway, reflecting back, I don’t think the diagnosis of gastric cancer would have been high on my differential were it not for the finding of severe anemia. So it’s interesting to see how Dr. Cocks triggered that diagnosis of gastric cancer much earlier — remember at that point he had been given just that this was a middle-aged man from Mexico with chest pain and epigastric tenderness.

CINDY Something else that stands out to me: He’s pretty young.

JOHN Yeah. And he’s not the first patient in his 40s or even 30s I’ve had with gastric cancer.

CINDY A question this case brings up: Should this patient have had screening endoscopy? As a Hispanic immigrant from a region with a high prevalence of gastric cancer?

JOHN In some high-risk countries like Japan and Korea, it’s standard to screen everyone, and while that has always been controversial, there is observational data suggesting it improves early detection and mortality. So it’s probably not unreasonable to consider offering screening to patients who have immigrated from one of these regions.

CINDY The problem is, selectively screening in immigrant populations is a strategy that hasn’t been fully studied. Whether to do this, whether it’s cost-effective, even how to do it — these are all areas of active research right now. Do you start at 40, or at 50, do you scope or use an upper GI series; do you repeat it every 2 years, or every 3 years, or only if it was initially abnormal?

JOHN Right. So there aren’t formal guidelines in the US. Which means as an outpatient provider you’re going to have to decide on your own, or with your patient, whether and how you’re going to do screening.

And we’re only talking about first generation immigrants here. Population studies in Japanese immigrants to the US show over subsequent generations, gastric cancer risk decreases to the baseline rate of Caucasians, so presumably early environment exposures are playing as much if not more of a role than genetics per se.


♫ Music to indicate end of episode ♫


JOHN Alright listeners, that’s about it for this episode. Hope you enjoyed listening to two of our bosses struggle as much as we did.

CINDY We hope you’ll join us next time as we continue experimenting with different case formats. Next episode’s twist: One discussant, but multiple cases.

JOHN Again, a shout out to HumanDx for hosting an interactive version of this case, which we hope some of you solved. And again, check out their app or website for other great chest pain cases throughout next week. Symptom-based learning, like Dr. Stern said.

CINDY Speaking of which, CoreIM finally has a website of its own. (claps and whoops)

JOHN That’s right. Come visit us at CoreIMPodcast.com, and subscribe for the latest updates. Remember, if you have a case you’d like to submit for discussion, or someone you’d like to come on and hear as a discussant, or if you’re interested in developing and hosting an episode, please, get in touch with us! Send us an email at CoreIMpodcast@gmail.com. We are also on facebook, and twitter, at @CoreIMpodcast.

CINDY Thank you to Drs. Patrick Cocks and David Stern for weighing in on this episode. Special thanks to our audio editor for this episode Richard Chen, along with our CoreIM colleagues Shreya Trivedi, Marty Fried, Amy Ou… And an honorable mention as always to Dr. Steven Liu.

JOHN Opinions expressed in this podcast are our own, and do not represent the opinions of NYU or other affiliated institutions, nor should they be construed as medical advice.

CINDY Thank you for joining us. With CoreIM, I’m Cindy Fang.

JOHN And I’m John Hwang. See you next time.



  • Cash BD, Banerjee S, Anderson MA, et al. Ethnic issues in endoscopy. Gastrointestinal endoscopy 2010;71:1108-12.
  • Kassirer JP, Gorry GA. Clinical problem solving: a behavioral analysis. Annals of internal medicine 1978;89:245-55.
  • Kassirer JP, Wong JB, Kopelman RI. Learning clinical reasoning. Baltimore, MD: Wolters Kluwer/Lippincott Williams & Wilkins; 2010.
  • Mandin H, Harasym P, Eagle C, Watanabe M. Developing a “clinical presentation” curriculum at the University of Calgary. Academic Medicine 1995;70(3):186–93.
  • Nendaz MR, Raetzo MA, Junod AF, Vu NV. Teaching Diagnostic Skills: Clinical Vignettes or Chief Complaints? Advances in health sciences education : theory and practice 2000;5:3-10.
  • Noble S, Pasi J. Epidemiology and pathophysiology of cancer-associated thrombosis. British journal of cancer 2010;102 Suppl 1:S2-9.
  • Saumoy M, Schneider Y, Shen N, Kahaleh M, Sharaiha RZ, Shah SC. Cost Effectiveness of Gastric Cancer Screening According to Race and Ethnicity. Gastroenterology 2018;155:648-60.
  • Miller, G. A. (1956). “The magical number seven, plus or minus two: Some limits on our capacity for processing information”Psychological Review63 (2): 81–97. CiteSeerX doi:10.1037/h0043158. PMID 13310704.
  • Mandin H, Harasym P, Eagle C, Watanabe M. Developing a “clinical presentation” curriculum at the University of Calgary. Acad Med. 1995 Mar;70(3):186-93.




Tags: , , , , , , , ,