‘Never meant to be elementary’: Why rheumatologists should avoid diagnoses of exclusion
Click Here to Manage Email Alerts
Although the phrase “diagnosis of exclusion” may conjure images of Sherlock Holmes solving mysteries through deductive reasoning, the practice of arriving at a diagnosis by excluding others should be avoided, a group of experts wrote.
Michael Putman, MD, MSci, of the Medical College of Wisconsin — alongside colleagues Jayshil Patel, MD, also of the Medical College of Wisconsin, and Anisha Dua, MD, MPH, of the Northwestern University Feinberg School of Medicine — recently published a paper in Rheumatology discussing the potential drawbacks of relying on diagnoses of exclusion in rheumatology.
Characterizing the phrase as a “‘pseudo-probabilistic aphorism,” Putman and colleagues argue that diagnoses of exclusion are inherently subjective and conditional, and as such defy the “principles of diagnostic reasoning,” and can lead to unnecessary testing and premature closure.
“Invoking the ‘diagnosis of exclusion’ aphorism immediately begs the question, ‘to the exclusion of what?’” they wrote. “In rheumatology, a prespecified list of all possible ‘exclusions’ does not exist. If asked to provide an exclusionary list before arriving at Behçet’s syndrome, for instance, five rheumatologists may opine seven different conditions. If asked weeks later, the same group of rheumatologists may well provide different lists entirely. This is because any list of ‘exclusions’ will necessarily be subjective and conditional.”
Healio sat down with Putman to further discuss diagnosis of exclusion, the potential pros and cons of the strategy, and alternatives rheumatologists can employ to draw more accurate conclusions for challenging conditions.
Healio: What is a diagnosis of exclusion? Where did that label come from?
Putman: It comes from a phrase that is commonly heard on the internal medicine wards when discussing the diagnostic workup for rheumatic diseases. I’ve heard it used to describe a number of diseases, including seronegative RA, polymyalgia rheumatica and fibromyalgia, to name a few. I don’t think any of those are diagnosis of exclusion. I think they are diagnoses that you come to through your measured diagnostic process.
Healio: What prompted the editorial in Rheumatology?
Putman: Hearing the phrase repeatedly said by fellows and residents prompted it. I write things because I want to contribute some perspective that I think would be worthwhile.
Healio: What is the advantage of seeing certain rheumatic diseases as diagnoses of exclusion?
Putman: I don’t think there is one to be totally honest, and the reason is that there is no such thing as a diagnosis of exclusion. Diagnostic reasoning should be Bayesian. What I mean by that is you start by formulating a probability that someone has the disease based on their presenting concerns — ie, “pretest probability.” Then you perform a physical exam, obtain laboratory testing, or get imaging studies.
Each of these has performance characteristics — ie, “likelihood ratios” — which allow you to update the likelihood that your patient has a disease — ie, “posttest probability” — depending on whether they are present or absent. At no point does it make sense to think of diagnoses as obtained by exclusion, because clinical reasoning does not work that way and modern testing was not designed to function in that manner.
Healio: If this is such a problematic technique, how did it become such an accepted and ubiquitous process?
Putman: This is a very seductive form of reasoning. Approaching a complex problem and boiling it down toa simple formula, which goes something like, “Not these things, ergo must be this other thing,” allows physicians to compartmentalize a very complex process.
It’s the formula Sir Arthur Conan Doyle used repeatedly for Sherlock Holmes, and I think many of us like to think of ourselves as a little Sherlockian. We gather clues and use brilliant deductive insights to find a definitive answer to vexing and complicated problems. It’s intrinsically appealing but it’s also not how diagnosis actually works, so it’s very problematic.
Healio: What are some of the theoretical problems with diagnoses of exclusion thinking?
Putman: This is where I really have strong feelings about this. There are a couple problems. The first one is that it is a subjective and conditional idea. When I talk to trainees about this, I ask what their list of exclusions includes, and everyone’s lists are always completely different. That’s because patient presentations vary and physician imagination varies. The second problem is that it conveys a false sense of confidence and leads to premature closure.
Healio: What are some of the practical problems with diagnoses of exclusion?
Putman: Likewise, I think the diagnosis of exclusion perspective leads people to send for unnecessary testing. If you’re going to say, “We have to exclude lupus” for a patient getting worked up for fatigue, you don’t need to exclude lupus, you just need to test for lupus if there is a plausible situation where the patient may have it. Establishing fatigue or fibromyalgia as “diagnoses of exclusion” requires you to exclude things like SLE, which is in many cases inappropriate. It winds up leading over testing and a lot of confusion for patients.
Healio: What are the dangers of over-testing related to diagnosis of exclusion?
Putman: I think this is an especially important issue. Over-testing is dangerous for society. We send for a lot of expensive labs and imaging that are almost certainly not necessary. To me, though, that is not the main concern. The main concern is over-diagnosis. When you start sending tests that probably should not be sent in the first place, you will inevitably get positive results.
Let’s say you had a “diagnosis of exclusion” perspective that you took toward fatigue and sent an ANA for every patient who presented with that concern. We know that an ANA will be positive in 5% to 20% of healthy outpatients, so we will wind up having a lot of positive ANAs. There are two ways that goes. One way is that you tell the patient about the result, but that you don’t think they have lupus. That is a very frustrating and confusing conversation for the patient. I think that adds to the sense that patients can’t trust doctors when they are doing workups.
The second and worse thing that happens is that you tell them they may have lupus, even if you don’t think they do. Then, when the next provider sees them, the patient explains their history of lupus and all of a sudden they are carrying this diagnosis that they probably don’t have.
Healio: Are there any potential benefits to thinking through the diagnosis process through exclusionary thinking?
Putman: No, I don’t think there are. I just think that it is not useful, and it can be harmful. I don’t mean to sound like a crazy person, but I don’t have anything positive to say using this framework. I think the thing people are trying to accomplish is to ensure they and trainees are thinking of every possibility, but you don’t need to call anything a diagnosis of exclusion to accomplish that. Much better to focus on principles of diagnostic reasoning and focus more on learning about your patients and how testing works.
Healio: In the editorial, you and the other authors mention patient-centric skills. What can rheumatologists practice to reduce a reliance on diagnosing by exclusion?
Putman: It’s not really fair for us to critique this practice without offering an alternative. I think the alternative is a combination of patent centric skills — like narrative medicine — understanding precepts of diagnosis reasoning, and deliberately honing those skills over time.
We need to be great history takers and great epidemiologists, so we can adequately assess the pretest probability that a patient before us has a disease. Then we need to be great diagnostic reasoning-ers so we understand what positive and negative tests will mean. I know many of these skills are more difficult than relying on heuristics and “lists of exclusions,” but like we say in the article — diagnosing rare and amorphous diseases was never meant to be “elementary!”
Reference:
Putman M, et al Rheumatology. 2022;doi:10.1093/rheumatology/keac278.
Collapse
Healio Interview
Read more about
Click Here to Manage Email Alerts