Diagnosis of Pulmonary Embolism, Aortic Dissection Sharpened by Hindsight: An Expert Interview With David Manthey, MD

Daniel M. Keller, PhD

February 23, 2010

February 23, 2010 — Editor's note: Hindsight makes events seem more predictable than they were, and it can be a great teaching tool to enable behavior change and improvement. In this Expert Interview with Medscape Emergency Medicine, David Manthey, MD, associate professor of emergency medicine and vice chair of emergency medicine education at Wake Forest University School of Medicine in Winston-Salem, North Carolina, discusses what hindsight teaches us about the diagnosis of pulmonary embolism (PE) and aortic dissection, messages he delivered at the American Academy of Emergency Medicine 16th Annual Scientific Assembly, held February 15 to 17, 2010 in Las Vegas, Nevada. Dr. Manthey gives a lesson on how physicians can simulate hindsight while trying to reach a diagnosis.

Medscape: How does hindsight work to improve diagnostic capabilities?

Dr. Manthey: Anytime we miss a disease process, we look back at the evaluation to see if we could have prevented the miss. The problem is we're looking back through a set of eyes that already knows the diagnosis. So now all those "maybe" clues take on more importance and weight.

You have overlapping presentations that have different evaluations. If we talk about chest pain, we have acute coronary syndrome (ACS), PE, and dissection. Most of the time, we can differentiate them, so we are not talking about most of the cases. But there are times when any 1 patient has features of all 3. Each evaluation only rules out 1 disease, and it is often at odds with another evaluation — like if I [computed tomography] CT them for a PE, do I get the right timing, the right bolus [of contrast agent] for a dissection? Can I really rule it out? Once I have given them a radiation load and a contrast load, can we go do the catheterization afterward?

Medscape: Do emergency physicians delay diagnosing these kinds of conditions and, if so, why?

Dr. Manthey: [They do] for several reasons. One, the history is not classic. You wish they would read the textbook and come in the way the textbook says they should come in. I wish every case of appendicitis would come in with pain that started around the umbilicus then radiated to the right lower quadrant, that they had pain first followed by nausea and vomiting, and they had no diarrhea, they had a fever, and they didn't want to eat. I'd be able to hit that every single time. The problem is they don't have that presentation.

If you look at a rare disease, the classic presentation is not that common. If you look at somebody who has a Beck's triad for pericardial tamponade . . . Beck's triad is rare to see all of it. So you've got to look for atypical presentations.

In people who have aortic dissection, if it doesn't take off a vessel that will show you symptomatology — i.e., have a stroke, have a loss of pulse in 1 arm, have numbness and tingling down the leg, take out the artery of Adamkiewicz and become a hemiplegic — if they don't do that there is no definitive physical exam findings. How many of those do you chase?

Or [there is the case of] a dissection masquerading as an [myocardial infarction] MI, where the dissection went backward and took out a coronary artery, so they come in with the full-blown MI with ST-segment elevation. But guess what? It was actually the dissection that caused it. When it mimics a much more common disease, you're going to go after the more common disease. That's why we delay.

Why else do we delay? When you start thinking about the disease process, you think [about] the test that's going to make the difference [and] if it's expensive. Do you radiate the patient? [We delay because] I've got to hold the patient here because I need the labs. Before I do the CT, or if you're going for a transesophageal echo in the case of an aortic dissection, it requires the cooperation of a consultant. I've got to call them and convince them that I actually need this. It adds to the length of stay. Human nature would be that I really don't need to look for it. I've got a diagnosis and am admitting them for ACS.

We may also delay treatment if I'm trying to decide between PE and aortic dissection. To treat one kills the other.

Medscape: Does your diagnosis in the emergency department (ED) in any way bias the services upstairs?

Dr. Manthey: I don't know the literature on that, but I would say, anecdotally, yes. It always biases us. We know that if someone is diagnosed with gastroenteritis, let's say, when the next person sees them, they say, "Oh, you had gastroenteritis," and we don't necessarily restart [a diagnosis]. [For example], when you turn over a patient in the ED to another doctor, you say, "I'm looking for a kidney stone, so I'm sending him to CT." The CT comes back negative for kidney stone, and you're done — no kidney stone — so you go home. What we should say is, "no kidney stone — what are we missing?" That's not a ding against our colleagues in other specialties. They are as much under the gun as we are.

What can we do with hindsight to prevent that from happening? I think a lot of physicians, when we write up a chart, we write up a chart to prove that it's what we think it is. . . . Perhaps we could do a little hindsight right then and say, "What am I missing? What are the bad things that I could miss?" You would be going back and reviewing your medical decision-making — not as to why you thought it was one disease, but why you didn't think other diseases were there. I think that prevents us from putting blinders on.

[A patient] may give you a pristine case for having a heart attack. What you should then do is say, "What are the other things that it could be? Could it be a PE?" Go down the Wells criteria. Ask him about pleuritic chest pain. Look for the tachypnea, the tachycardia, the anxiety. Look at the [electrocardiogram] for more than ischemia. Think of aortic dissection as the cause of the ischemic pain, and ask about symptoms like Marfan's or Ehlers-Danlos syndrome. Look for physical exam, look more heartily at the x-ray, and think of why you don't think it is aortic dissection.

When you look at medical malpractice, a lot of times what the reviewer is doing is not saying, "You missed it." We already know you missed it. That's why it's a malpractice case. What they're doing is saying, "Should a reasonable person have found this?" If you've got good documentation that's reasonable as to what you did, why you did it, and why you didn't chase it further . . . I think you're much better protected. . . . I think you have to document why you thought it was what it was, but if there are other things that are overlapping in your disease process, or you're letting somebody go home and you don't know what it is, it's very nice to go down and say, "These are the emergencies I've ruled out, and this is how I've ruled them out. This is what I've told him to come back for, and this is the follow-up I've done to make sure that I'm not missing something."

Medscape: D-dimer tests are being used more and more in the diagnosis of PE. What are some of the pitfalls of using it?

Dr. Manthey: We didn't know who had PE. You had no idea of how to figure it out, and we came up with Wells criteria. Wells criteria, however, [categorized] everybody into different risk panels but didn't rule anybody out. . . . What's the next thing I'm doing? I'm chasing down a D-dimer. Now, a D-dimer can be falsely positive. In a couple pieces of literature, the D-dimer in a normal person without any disease process is negative in only about 65%. Hence, it's positive in 35% for a multitude of reasons — recent surgery, trauma, pregnancy, postpregnancy, cancer, [disseminated intravascular coagulation], cardiovascular disease, congestive failure, infection, sepsis, inflammation, vasoocclusive episodes of sickle cell disease, decreased clearance secondary to liver disease, renal disease, and venous malformations that clot and unclot.

Medscape: Emergency physicians are very aware of PEs and try to track them down aggressively using Wells or other criteria. If the criteria scores indicate that a PE is unlikely, they turn to D-dimer tests. Do you have any advice about the use of D-dimers and their interpretation?

Dr. Manthey: Know what the D-dimer [cut-off] is at your own institution. It depends on which assay they use and which machine they use to run that assay.

Having been trained in looking at Wells criteria, Wicki criteria, Geneva scores, we know those scores, and so our gestalt is as good as any one of those scoring programs. Using gestalt and having a very low [threshold of] suspicion and using a D-dimer would be OK, too.

Medscape: Let's talk about aortic dissection and some of the diagnostic tests for it.

Dr. Manthey: Chest x-ray is not very good for aortic dissection. I would say that 30% of aortic dissection x-rays are normal or nonspecific. So what do we look for in aortic dissection? We look for obvious evidence, a wide mediastinum greater than 8 to 10 cm. But there are lots of things that cause widened mediastinum: lymphadenopathy, ectatic aorta, the way the picture is taken. But you look at other things. Do they have an indistinct aortic knob? That's a subtle finding.

If the calcification on the intima vs the adventitia is greater than 6 mm, you worry. That could be a dissection. If you see a left-sided pleural effusion, that could be a hint they ruptured and bled. They usually bleed out to the left side. If you have depression of the left main-stem bronchus because the aortic arch is over it, and if you get a hematoma there, it will push down on the left main-stem bronchus, but that's a subtle finding. Does it mean aortic dissection every time? No. These are all secondary signs that may make you scratch your head but don't say you've got a dissection.

I don't think any ED physician is going to miss the widened mediastinum with somebody who's got pain that's tearing through to their back. It's the [patient] that has a nondescript chest pain who has a subtle finding on chest x-ray [that is the problem].

Medscape: Do you have any bullet points to impart about PE and aortic dissection and quick ways to zero in on a diagnosis?

Dr. Manthey: I don't know that I do. I think for most emergency medicine physicians, PE has become so high on their differential now. The literature suggests that even with the onset of Wells [criteria] and D-dimer, we still miss a significant number. In a study done in 2009, PE [was found to be] 1 of the five most commonly missed diagnoses — in their study, 4.5% of the total. We know that there's a delayed diagnosis in about 16% of people with PEs.

Medscape: How good are physicians at finding aortic dissections, and what advice can you offer?

Dr. Manthey: My argument for aortic dissection is, you've got to think of it. If you look at the literature, even the cardiothoracic surgeons, who are the experts in this, say that missing it is probably the norm because it's found [only] postmortem in 27% to 55% of patients with it. So the standard of care is probably to miss aortic dissection.

We probably ought to be thinking a little bit more about doing another study to look at it. There's some literature out there about D-dimer. I'm not sold on it yet, as to whether or not it can rule it in or rule it out. Does that mean we're going to be chasing more and doing more CTs? Absolutely it does. Does it mean we're going to be doing more transesophageal echoes that are negative? Absolutely it does. But we've got to do something to catch more aortic dissections, and nobody has a good answer to that.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.