CMAJ Podcasts

How to be better at diagnosing rare conditions.

September 26, 2022 Canadian Medical Association Journal
CMAJ Podcasts
How to be better at diagnosing rare conditions.
Show Notes Transcript

Hereditary angioedema (HAE) is a rare autosomal dominant disorder characterized by recurrent episodes of painful swelling without urticaria that leads to substantial morbidity and even mortality if left untreated.


On this episode, Dr. Victoria Cook, an immunologist working in Victoria, BC, describes the case of a woman who had 6 emergency department presentations, 1 hospital admission and saw 5 different specialists over 11 months before finally being diagnosed with HAE.


Drs. Bigham and Omole, then speak with Dr. Ed Etchells about how physicians can work to strip away biases and cognitive obstacles that can prevent timely diagnoses of rare and challenging conditions. Dr. Etchells works in the division of General Internal Medicine at Women's College Hospital in the University of Toronto.


Article discussed

Transcript

CMAJ 


Join us as we explore medical solutions that address the urgent need to change healthcare. Reach out to us about this or any episode you hear. Or tell us about something you'd like to hear on the leading Canadian medical podcast.

You can find Blair and Mojola on X @BlairBigham and @Drmojolaomole

X (in English): @CMAJ
X (en français): @JAMC
Facebook
Instagram: @CMAJ.ca

The CMAJ Podcast is produced by PodCraft Productions

Dr. Blair Bigham:

I'm Blair Bigham.

Dr. Mojola Omole:

I'm Mojola Omole, this is the CMAJ podcast.

In today's episode, we're going to discuss a practice case that was recently in the CMAJ, entitled “A Challenging Diagnosis: Hereditary Angioedema Presenting During Pregnancy.”

Dr. Blair Bigham:

Hereditary angioedema, Jola, is not something that rolls off the tongue very easily.

Dr. Mojola Omole:

Or something I've ever heard about.

Dr. Blair Bigham:

This comes up on emergency medicine boards, because it's a nice differential along with anaphylaxis and ACE inhibitor induced angioedema. But it certainly isn't something that we see often. This article here was not very typical case presentation of hereditary angioedema. We're going to talk to the author of this paper, who's an expert in hereditary angioedema, but the bottom line is that a woman with an atypical situation wasn't diagnosed for 10 months.

She ended up having three CT scans and seeing many specialists just to get to the point of having a simple blood test that eventually revealed the answer. We'll talk to the author to find out exactly how that happened. And then afterwards we're going to broaden that conversation beyond HAE. We're going to speak to Ed Etchells about how clinicians should approach diagnosing rare conditions. That's coming up.

Dr. Mojola Omole:

Dr. Victoria Cook is the co-author of the practice case in CMAJ entitled “A challenging diagnosis: hereditary angioedema presenting during pregnancy.” Dr. Cook is an immunologist in Victoria, BC. Thanks for joining us today.

Dr. Victoria Cook:

Thank you so much for having me.

Dr. Mojola Omole:

Before we dig into the specifics of the case you describe in the paper, can you give us just a quick primer on what exactly HAE is?

Dr. Victoria Cook:

That's an excellent question, because it's a not commonly seen diagnosis. HAE stands for hereditary angioedema, which as it's in the name is an inherited disorder characterized by intermittent episodes of swelling that can impact the face, it can also impact our intestines, lead to intestinal swelling and obstruction, and other parts of the body as well. Our hands, our feet, any areas of skin.

When we think of swelling, there's often two kinds of swelling that we think of. The more common one is the histaminergic caused by histamine, that quick onset and offset… responds to antihistamines, epinephrine… sometimes associated with allergy. And then there's what we call bradykinin-mediated angioedema, which is what we see in hereditary angioedema. This is an inherited disorder where we either don't make a functional, or don't make enough of a, substance called C1 esterase inhibitor, which is the off switch for a certain pathway of swelling, which results in the production of something called bradykinin, which is this really powerful vasodilator that leads to swelling and pain.

Dr. Mojola Omole:

This might sound like a really basic question, but I'm a surgeon, so I'm kind of basic. Is the swelling... Is it different? Is the bradykinin-mediated angioedema different than a histamine-mediated one?

Dr. Victoria Cook:

When you're looking at swelling, I would say it looks the same, but there are characteristics about them that are different. Primarily one thing we often think about with histaminergic swelling is that it's itchy. There's often an itch and then also swelling that's caused by histamine. There's often people have hives at the same time, so you might have hives and swelling. And then the histamine-mediated one actually happens really quickly. It'll happen over minutes that the swelling and hives can evolve and it goes away a lot more quickly too, so hours or a day or so. And then it's gone versus bradykinin-mediated angioedema swelling tends to have this fairly slow onset over several hours and it lasts for days and it's not associated with any itching whatsoever. And there's definitely no associated hives, although it can be preceded by this serpiginous red-pink rash, which is not itchy.

Dr. Mojola Omole:

Is part of what you've just described to us about the scenario around the swelling, is that part of how you confirm the diagnosis?

Dr. Victoria Cook:

I think the features on history can be helpful in giving you a clue as to what kind of swelling it is, but certainly would not be sufficient to allow you to make the diagnosis which you do with blood and sometimes genetic testing.

Dr. Mojola Omole:

All right. Walk us through this case that you profiled in the paper, what stood out for you?

Dr. Victoria Cook:

This was a really interesting case with lots of good learning points. And basically this was a young woman in her early twenties presenting in pregnancy. And throughout her pregnancy, she was having these episodes of intermittent abdominal pain, nausea, some vomiting, and sometimes associated with a bit of swelling, usually involving her hands, and feet. And a few times she had this red rash that wasn't itchy, that occurred around the time of those symptoms. It happened to her about four times. She went in and she was reassured, these symptoms are super common in pregnancy…once the pregnancy's over, they should go away.

She delivered a healthy child and everyone looked good, she went home. She had a C-section, but she came back a few weeks later and was diagnosed with an infection of her incision. Diagnosed with cellulitis. She was given some antibiotics and went home, but she came back just a few days later and she had really severe abdominal pain, vomiting and signs of hypotension. She was tachycardic, she needed fluid boluses and it was not getting better with, ondansetron or fluids.

They did a really thorough workup on her looking for untreated infections. She actually had two CT scans. They couldn't really find any abscess and the infection, the wound didn't look too bad, and it wasn't necessarily worsening or anything like that. They did a whole bunch of stool samples. They considered reflux… a pretty broad differential.

In the end, the only interesting finding there was on one of the CT scans, her second one, she had swelling of one of the parts of her small bowel in an area that was unrelated to the C-section wound. After this whole workup, multiple physicians, they figured, well, it's probably the infection. They treated her with additional antibiotics. The IV fluids, wound vac, they threw on ranitidine just in case for possible reflux. And they'd found she had a C difficile positive PCR in her stool, toxin negative and they threw on treatment for that just in case, I think being very thorough.

She eventually went home, things cleared, but over the next nine months she continued to have these intermittent symptoms. Primarily it was... What was making her go into healthcare providers in the hospital was abdominal pain, nausea and vomiting. It was lasting three or four days at a time that was associated with swelling and generally the swelling was labial, sometimes peripheral…hands, feet,  fingers, again would last a few days and resolve. She also had episodes of swelling that were unrelated to the abdominal symptoms, which she didn't necessarily present for. But again, that was labial peripheral swelling. And in total it was something like six emergency department presentations, one hospital admission, and I think five or six different specialists in total between obstetrics and gynecology, internal, multiple infectious disease specialists, gastroenterology.

She had a very extensive workup with multiple infectious disease physicians, so looking at stool cultures, looking for sexually-transmitted infections, all sorts of weird and wonderful infections. She had another CT scan, so I think three of those in total, two ultrasounds looking at the vulval swelling in the pelvic regions in the context of that unilateral angioedema of the labia, and H. pylori breath test. And then she eventually also had an upper and a lower endoscopy. And there were various, when you look, going through the records, working diagnoses. She had initially the pregnancy-related symptoms that I mentioned, but also the wound infection, possible C. difficile, gastroenteritis, Helicobacter pylori, possible inflammatory bowel disease and then a nonspecific colitis, so lots and lots of potential diagnoses.

Eventually, she was sent to yet another outpatient infectious disease clinic. And at that time the physician went through her case and went through the patient's really fabulous collection of photographs of her swelling. And that physician ordered the C4 and a C1 inhibitor, which returned abnormal and then referred her on to me where we repeated that testing. And, because this patient didn't have a family history, we also did genetic testing to confirm the diagnosis of hereditary angioedema for her.

Dr. Mojola Omole:

Before the ID doctor raised it, was that the first time when you look back through the notes that it was ever raised as a diagnosis?

Dr. Victoria Cook:

Yeah, it was. And what was really interesting, because I can't stress this enough, the physicians that worked on this case worked really hard for this patient. Everyone was really thinking, the consult notes were really... Everything was well-considered. People were thinking of differential diagnoses, referring things off. But what was really interesting was that swelling was never really considered as a primary presenting complaint, which was really interesting, because people documented, and she had workup for, the labial swelling, this unilateral labial swelling. She had two ultrasounds for that, looking at that, but it was never considered swelling.

No one ever had angioedema and then went through differential for angioedema specifically, even though she had that and it was mentioned and documented in the notes, but it wasn't necessarily considered its own presenting complaint. And part of that she'd presented in pregnancy and then she'd had the infection and then a lot of the symptoms seemed gastrointestinal related, but the swelling itself wasn't considered independently, which is really interesting.

Dr. Mojola Omole:

Do you think maybe that's why physicians were looking in other directions and not thinking of HAE?

Dr. Victoria Cook:

Absolutely. I think I mean there were several features in this case that with the benefit of hindsight and reviewing the records, you're like, Oh, okay. Even though there were elements of this case that were really classic, there were also features that made it more difficult to pick up. And I think one of the key ones is the location of the swelling because I think abdominal symptoms are often really... They have such a broad differential. And if you think someone presenting in pregnancy with abdominal pain and nausea, vomiting - hereditary angioedema is nowhere in the top 100 things on your list to think about. That makes sense that it wouldn't have come up there.

And, similarly, when someone's presenting again, primarily with abdominal concerns, there's so many other things on that differential that this just didn't come up. And I think if she had an episode of swelling that was say facial, like big lip or eye swelling, I feel like it would've gotten picked up much quicker. We get so many referrals for facial angioedema where hereditary angioedema is always considered. And so, I feel like that would get... That's much less likely to be missed in that context, but swelling in other places-

Dr. Mojola Omole:

Did this patient ever have facial angioedema?

Dr. Victoria Cook:

She did not. And still hasn't. And that's not uncommon. But she does have it very obviously in other places. Fingers, parts of the arm, lower limbs, labial-

Dr. Mojola Omole:

And does it sometimes present only with abdominal pain and no external manifestations?

Dr. Victoria Cook:

Absolutely it can. And I think, what's a really interesting thing is … Because often it presents typically in your first two decades of life in about 80% of people and it can be worsened by either endogenous or exogenous estrogens. For young women, presenting with hereditary angioedema, sometimes around the time of onset of menstruation or if they start an oral contraceptive, they might have cyclical abdominal pain. And if that's your only symptom these patients get missed for quite a long time. Because again, no one is thinking of it in that context, but absolutely it can be your only symptom.

Dr. Mojola Omole:

What do you think prompted the final physician to look into it?

Dr. Victoria Cook:

I think what really got her was the photographs. As I mentioned, so there was, again, I think all of the physicians involved in the case were very motivated, trying to think of things. The patient herself really needs to be commended because she kept such fabulous documentation of all of her swelling. She had a phone full of images, which were very classic, including the preceding rash, episodes of swelling in various places. And she had documented when those happened. As soon as you see that… at the time that I saw her, she led me through all of those photographs… she had done that for the previous physician as well. And I think then putting that case together then that made it a little bit simpler.

Dr. Mojola Omole:

How easy is it to incorporate the lab tests into a workup like the C1 esterase and the C4? Are they expensive? Are they available?

Dr. Victoria Cook:

That's a great question. Yeah, they're widely available and they are inexpensive tests to order.

Dr. Mojola Omole:

I've sent off the C1 esterase in the emergency department before. Am I supposed to send off a C4 every time as well? Do you need both?

Dr. Victoria Cook:

That's a great question. And in general in adults, often we say, Oh, you can use the C4 as a screening test and then if it's abnormal move on to the C1 inhibitor. I think practically if you're really considering it, I think most of us just order the C1 inhibitor. And what you get with that actually depends a bit on where in the country you practice. In some places, you can get a C1 esterase inhibitor level and then you separately order a C1 inhibitor function. In other provinces, they fund a single assay, which is a functional assay. And then that'll be low regardless of whether the total amount of protein is low or whether the function is abnormal. But I'd say, functionally, I order both at the same time.

Dr. Mojola Omole:

And once you identify the diagnosis, how challenging is it to treat it?

Dr. Victoria Cook:

This is a disorder that carries significant morbidity and the risk of fatality, right? If you have unchecked upper airway swelling, that can cause fatality for these patients so it is important to catch it. And then when you do, our treatment options generally involve… it depends a little bit on the frequency of patient symptoms, patient preferences, as far as medication use. But what we have is if you're missing, if you're not making either functional or enough of C1 esterase inhibitor, we can give that back to you and we can either do that as needed.

When swelling happens, we can give you C1 inhibitor, or if you have frequent symptoms and the patient wants that you can do prophylactic therapy where you take it to prevent episodes from happening. We have a few other options now for both as-needed management, that isn't C1 inhibitor. Other mechanisms targeting that same pathway to essentially prevent bradykinin from being made or from sticking around that can be used on an as needed or prophylactic basis for patients. And there's several new things coming, hopefully, coming down the pipes for this patient group soon.

Dr. Mojola Omole:

Thank you so much for speaking with us. Dr. Victoria Cook is an immunologist in Victoria, BC.

Dr. Blair Bigham:

Let's zoom out and look at what lessons cases like these have for physicians. Dr. Ed Etchells is a member of the division of General Internal Medicine at Women's College Hospital in the University of Toronto. Ed, thanks for joining us today.

Dr. Ed Etchells:

My pleasure. Thanks for inviting me.

Dr. Blair Bigham:

Ed, how surprised were you by this patient's lengthy journey to eventually get to a diagnosis?

Dr. Ed Etchells:

I was not surprised, primarily because she presents with a rare condition and even the analysis of the experts in the discussion of the paper says that her presentation was not typical for that rare condition. But I think the case does demonstrate some of the common pitfalls in diagnostic thinking. 

There are two overarching tendencies in the diagnostic thinking process. The first is clinicians tend to only seek as much information as is required to form a diagnostic impression. This is actually a really good thing because most of the time clinicians are correct with their initial impressions, so this is a very efficient way for patients to get the correct diagnosis and the correct treatment. When this tendency fails us, it's called “premature closure”. We close our mind before we've nailed the correct diagnosis.

The second tendency is to stick to initial impressions, even though there may be additional new information that does not fit with that initial impression. And that's got a general term of “anchoring”. In this case, I think if you look at the figures, there is very good suggestion of anchoring because the initial impression was infections, either wound infection or Clostridium difficile colitis, both very appropriate, common concerns. But as the situation persisted, you could see there was still a focus or an anchoring on infectious explanations for her condition, when there was other information to suggest there was something else that might be wrong with the patient.

Dr. Blair Bigham:

I want to talk about both early closure and anchoring. But first I want to know, as a specialist, I come to work with a toolbox that I can use, that I've been trained to use. And anything not in that toolbox, my instinct is, well, if I can't make you fit in my toolbox, I'm just going to refer you to somebody else. And that seems to have been what happened here quite a bit, but then it seems like you can just get bounced around different specialists, all of whom have this sort of defined toolbox. How can physicians break out of that anchoring and that early closure and go, Wait a minute - I need to shift gears. I need to think differently here.

Dr. Ed Etchells:

Yeah. So Blair, you're talking about the general field that's called “metacognition”, which is, how do we monitor our own thought processes? I'll just say to you, it's not so easy. It's theoretically attractive, but difficult in practice. And I want to reemphasize that these tendencies that I outlined initially, they're very good tendencies. Only seeking information that's required is very efficient and good for patients. It avoids unnecessary testing for example, and it's efficient. It's only when we miss the diagnosis that people criticize these tendencies and call them biases.

The question is, can we monitor our own thinking? The theory is that we can, if we become aware of common pitfalls, and if we are able to recognize situations where the diagnosis might not be correct. I think a useful diagnostic red flag is that the patient is not improving and they continue to have these same symptoms or signs as their early presentation. To me, this is pretty strong evidence that the diagnosis may not be correct.

In this example, the patient continued to have problems with intermittent limb and labial swelling and abdominal pain, despite many investigations and treatment efforts. This is a diagnostic red flag that there may be a mistake, a misdiagnosis, and then apply some metacognitive tricks to try and get to the correct diagnosis.

Dr. Mojola Omole:

There was one bias that I really wanted to talk about, which was the representation bias. For example, like a heart attack, the presentation we were taught, at least I was taught back then, was that it's going to be a male, usually white, chest pain, short of breath, diaphoretic. And that it's going to be a heart attack, whereas that's probably not what Blair sees on a regular basis.

Dr. Ed Etchells:

Yeah. It's such a good point. Representativeness bias is exactly, as you said, we all form an impression of how that disease should appear to us. That impression can be subject to conscious or unconscious bias or just poor teaching. Speaking personally, my representation of hereditary angioedema is recurrent facial or upper airway edema without urticaria, which is not a bad representation, but it doesn't cover all of the cases even remotely. And this is why the discussants in the paper suggested the diagnosis might have been delayed because that representation doesn't cover the other presentations of HAE, for example, asymmetric, intermittent limb or genital edema.

Now for the experts, that's part of their representation. They had no trouble making the diagnosis. As you also said, our representation for acute myocardial infarction, probably isn't accurate. Women present differently than men. And we should have a different representation of myocardial infarction for a female. Another good example is skin diseases in patients with diverse skin color. The representation is completely wrong for many patients.

Dr. Blair Bigham:

Ed, I can imagine that in the days after reading this article, there were probably a lot of people who ordered C1 esterase labs or things like that, because of that recency bias. Once you hear of a rare diagnosis, it's on top of mind. Talk to me a little bit about how you can learn from cases like this without totally augmenting and beefing up your diagnostic processes for more routine evaluations. And I guess in other words, how do I judiciously order tests and appropriately use healthcare resources while still being mindful and still worrying about missing some of these rare diagnoses?

Dr. Ed Etchells:

Yeah, it's such an important question. I think the first concept is the concept of availability bias that you alluded to. Our diagnostic thinking is unduly influenced by ease of recall of recent similar cases. So because I just read this case today, because I'm going down to our Acute Ambulatory Care Unit on Thursday, the first patient I see with swelling, I'm going to be thinking about this case. That's availability bias. Again, there's no perfect way to combat that, but the best way is to know the base rate of conditions for common clinical presenting syndromes. For example, if I saw a hundred patients in my clinic with unilateral leg swelling, for example, I'm certain that hereditary angioedema would be extremely low on that list and I should make sure they don't have a DVT and varicose veins and all the other common causes.

Having said that, at some point you're going to get a diagnostic red flag that you've gone through the common conditions, the patient still has concerns, and the treatment is not working, and you have not confirmed those diagnoses. Then you need to open your mind to more rare conditions. To combat premature closure, a good question is, what else could it be? And try and generate five to 10 conditions that come to mind and then see what the evidence tells you about those possibilities.

Another trick in the diagnostic expertise literature is to seek disconfirming evidence. A common pitfall is called confirmation bias where we tend to seek information that supports our initial impressions. What we should be doing is seeking information which does not support our impression, that leads us to another diagnosis. For example, if the patient has recurrent unexplained abdominal pain, don't ask about their bowel. I'm not saying you shouldn't ask, but don't focus on their bowel movements, ask them if they've ever had acute neuropathy or their urine changes color in the light, because then they might have porphyria.

Dr. Blair Bigham:

Ed, with all of these cognitive biases that can add up to either what we might think of as being an expected or acceptable delayed diagnosis, or sometimes not when the delay is unacceptable or maybe even dangerous. Or for all those patients out there who might not get to that infectious disease doctor who figured it out or who might not make it to that GIM clinic and get their correct diagnosis. What type of changes, if you were to take a quantum leap a hundred years from now, do you think medicine would have to take to be better at picking up on these unusual presentations of uncommon diseases?

Dr. Ed Etchells:

First of all, it's such an important question. I would emphasize that the most common misdiagnoses are actually common diagnoses. If you look at studies-

Dr. Blair Bigham:

Why is that?

Dr. Ed Etchells:

I think it's probably representativeness bias plus all of the other things we've talked about. Another really good example is pulmonary embolism. That's a common misdiagnosis, because we are trained that it's a patient with acute chest pain, a high GVP, maybe hemoptysis and ECG changes.

Dr. Blair Bigham:

That's none of the last 10 PEs I've diagnosed.

Dr. Ed Etchells:

Yeah, exactly. That just never happens. Those are the easy ones, but those are actually uncommon presentations of a common disease. But for rare diseases, I think having some cognitive humility, recognizing when there are diagnostic delay red flags that we've discussed, forcing yourself to share your thinking with others and “others” could be with the patient or with your colleagues. Often the simple act of trying to explain to others what you're thinking opens your own mind and helps you see the flaws in your thinking.

I love doing team medicine. I have to explain what I'm thinking to the residents and students and vice versa. It's extremely valuable. I also think that there's going to be a role for automated expert diagnostic decision support. There are lots of systems out there. I've not reviewed this literature recently. The last time I reviewed it, the performance wasn't awesome. I took the liberty of typing in the data from this case into an expert system and it does require you to enter the data in a certain way. But if I enter the data as a woman, age 23 with any pregnancy status, with abdominal pain and edema, the first thing that popped up was hereditary angioedema.

Dr. Blair Bigham:

No way.

Dr. Ed Etchells:

Yeah.

Dr. Mojola Omole:

Wow.

Dr. Ed Etchells:

This was impressive to me because last time I did this a few years ago, it didn't work very well. I'm not trying to sell anybody anything, but there's definitely going to be a role for diagnostic decision support that generates lists that help you open your mind.

Dr. Blair Bigham:

Well, you've opened my mind. Thank you so much for joining us today. That was a fabulous conversation. Ed Etchells is a member of the division of General Internal Medicine at Women's College Hospital in the University of Toronto.

All right, Jola, I have lots of thoughts, but first let me hear your take.

Dr. Mojola Omole:

One lasting impression that was left on me, and especially after talking to Dr. Etchells, is whether or not there was implicit bias made about this patient, given her age, her gender identification, and the fact that she in the past used drugs. And whether or not all of those things biased the treating physicians to not take her as seriously. We have lots of evidence about those things affecting our biases. And so, I do wonder if that was part of what made it hard for clinicians to get to the diagnosis.

Dr. Blair Bigham:

I think those biases are exacerbated in the emergency department where we see stereotypes of patients so often that I think it does certainly affect the way we think in a negative way. And it's really, really hard to battle. Sometimes you'll see, just on your screen, it's so obvious now with electronic records that this is the patient's sixth visit in four months and your instinct, sadly, is to roll your eyes as opposed to thinking, Oh, interesting…we still haven't figured it out. And so, I think Ed's message is really important for me to hear. And maybe my colleagues in the ER as well, about really leaning into that unsettled patient complaint and saying, Wait a minute, we've missed something. As opposed to saying, Oh, you're back again. Which unfortunately I think happens all too often.

Dr. Mojola Omole:

A hundred percent agree with you. I would say that it requires, definitely not the emergency department kind of consultations. This requires a clinic where you can spend time with the patient and actually go through every single thing systematically, because I think that often what is missing is that this patient was often seen through the emergency department and then maybe admitted on a busy ward to try and work up for this and that versus going to a specialty clinic. I'm assuming the infectious disease doctor that she saw was in their clinic. And once they were able to sit down and she showed all of the pictures that she's been documenting, it was like, Aha, this is what's missing.

Dr. Blair Bigham:

That's it for this week on the CMAJ podcast. Thanks very much for joining us. If you have a chance, please remember to like or share our podcast, wherever it is that you download your audio. 

I'm Blair Bigham.

Dr. Mojola Omole:

I'm Mojola Omole and, until next time, be well.