The Universal Differential Diagnosis
If you are reading this book, chances are that you are fairly smart and already know lots of useful medical stuff. In becoming a radiology resident, you have clawed your way to the top of the medical food chain (in my opinion), and are a force to be reckoned with. Even so, when you are shown cases in conferences, you probably still stumble around a bit coming up with any kind of plausible differ ential diagnosis. If you don’t do this in case conferences, it is probably because:
- you have seen a case of that disorder before.
- your mother had that disorder.
- you read this syllabus.
- you are a Vulcan, and so inhumanly smart and organized that you don’t need this syllabus. If so, go away, and pass this syllabus on to someone who really needs it.
So, how can it be that someone as smart and knowledgeable as you undoubtedly are still blows cases in teaching conference? Generally, it’s because under stress, you have trouble accessing the information that’s already in your head. Chances are that you already know many pertinent facts about most of the disorders in the differential for the cases that you miss — it’s just that you can’t pull them out in an organized fashion when you’re on the spot. This situation is going to arise many times a day, for the rest of your professional life, that is, if you plan to actually read films, and not work in some radiological monastery or convent somewhere. On our musculoskeletal service, we formally present and discuss unknown cases several times a week to our clinical colleagues. Clinicians drop by our alternators informally all during the day for consultations and differential diagnoses on unknown cases. We also do acute readings of cases from the Emergency room all day long. In addition to this, lawyers call us up every month or so and want our opinions on medicolegal cases — sometimes just on the phone, sometimes in the form of a deposition, and sometimes in court, in front of a jury.
Someday, after reading thousands of films and many books and journals, you’ll be able to do this as well as or better than your attendings. By then, you’ll have seen most of the common and many of the unusual disorders in musculoskeletal radiology, and you’ll have many differential diagnoses ready at the tip of your tongue. This syllabus, however, is concerned with right now. How can you, a beginning resident, begin to access some of this stuff that you already know, particularly when you are under stress?
One handy tip is: mnemonics. While some people learn lists of facts perfectly well without them, I personally find mnemonics veryhelpful — especially the right mnemonics. A good mnemonic radically changes the way you use your brain. Instead of wandering around aimlessly in your head hoping for inspiration, a good mnemonic lets you systematically look one by one in various drawers in your brain, sorting through facts that you already know.
The mnemonic that I think you should learn first is probably the most useful mnemonic in all of medicine: the universal differential diagnosis mnemonic. This is also known as the “categories of disease” differential diagnosis. Now, you don’t need to use this differential diagnosis all the time. If you have a differential diagnosis for a specific radiographic problem, use it instead. However, if you don’t know a specific differential diagnosis for a particular problem, use the universal one. The universal differential diagnosis that I use most often is: VINDICATE.
There are lots of other possible mnemonics for the universal differential diagnosis. I have at one time or another used “VINDICATE”, “VITAMIN C+D”, “ATOMIC DDT “, and “KIIIITTEN” (four-eyed kitten). They all work just fine. I present VINDICATE here because it happens to currently be au courant at the University of Washington.
Now, when someone is consulting you, you don’t want to just blindly regurgitate this differential in their lap. Instead, you want to run through it quickly in your mind, and apply any specific clinical, laboratory, or other radiographic knowledge you have about the patient. As you apply this extra knowledge, you will be able to mentally discard some of the entities in the universal differential and only actually utter the ones that make some sense.
Let’s try VINDICATE with diffuse sclerotic bone disease. We want to start by considering that diffuse disease is most often caused by entities that are generalized or systemic in action. Put these entities at the top of your differential. Below them on the list should be the entities that usually are more focal in nature, but which can occasionally spread diffusely throughout the musculoskeletal system. Omit any entities that don’t make any sense or don’t cause bony sclerosis. So, here’s the raw differential one might come up with:
- infarcts (sickle cell)
- chronic osteomyelitis
- Vitamin D
- bone islands
- fracture (stress)
- Paget’s disease
The next thing to do is to reorder these entities so that they are now arranged in order of their likelihood. Here’s how that list might look if rearranged this way for a particular patient:
Differential diagnosis in order of prevalence
- Generalized or systemic processes
- diffuse skeletal infarcts (sickle cell)
- drugs (vitamin D, fluoride)
- congenital (osteopetrosis, pyknodysostosis)
- Usually focal processes presenting as diffuse disease
- metastatic tumor (prostate, breast, other)
- Entities that are pretty darned unlikely to present as a diffuse process
- primary neoplasms (osteoma, osteosarcoma)
- congenital (bone islands, osteopoikilosis)
- metabolic (Paget’s)
- chronic osteomyelitis
If a disorder or finding has its own special differential, use it
Sometimes, a specific category of disease will have its own intrinsic logic that dictates the easiest way to approach disorders in that category. Examples of this will be presented for each major finding in this syllabus. For now, I’ll present two examples of this: one for vertebra plana (a vertebral body that is as flat as a pancake) and one for jaundice.
First, let’s look at what the universal differential diagnosis gives us for vertebra plana.
- vertebral osteomyelitis (bacterial, TB, etc.)
- eosinophilic granuloma
For vertebra plana, the differential diagnosis I like to use is: MELT. I like this differential not only because it is useful and succinct, but also because it actually spells out a real word!
- Metastasis / Myeloma
- Eosinophilic granuloma
- Trauma / TB
With a specific differential diagnosis like this, you don’t waste your time considering a lot of unlikely causes of disease.
This shorter and more specific differential diagnosis came up with the same list of possibilities, as the longer and more comprehensive one, just in a different order. The down side of using the longer universal differential diagnosis is that it forces one to consider lots of possible etiologies that just don’t make that much sense as causes for vertebra plana.
Sometimes, though, it does make more sense to lengthen the differential diagnosis, rather than shorten it. A good example of this is in the workup of jaundice, which has jillions of causes. Just as an exercise, shut the syllabus for a moment and see how many causes of jaundice that you can think of off the top of your head. Go ahead, dig around up there, and when you finish racking your brain, come back here and we’ll look at a way to expand VINDICATE to look at jaundice.
Ah, you’re back. Hopefully, you came up with a reasonably long list of causes for jaundice (I realize that most right-thinking radiologists have long since stored what little they once knew about jaundice up in hermetically sealed crates up in the musty attic of their brains. Some of it has no doubt spoiled, some of it is out of date, and some of it has probably been eaten by mice. However, lots of it is still up there somewhere — and even radiologists should know something about jaundice). Now let’s see how we do with VINDICATE. To expand a mnemonic like VINDICATE, it is helpful to first apply a little specific knowledge about the underlying pathophysiology of the disorder in question. When we look at the underlying nuts and bolts of jaundice, we find that it is due to too much bilirubin in the wrong places. It is also pretty clear that the liver plays a major role in world of bile. Therefore, it is convenient to divide the universe of jaundice into three broad categories of causes: prehepatic, hepatic, and posthepatic. If we take these three pathophysiological categories and multiply them by the nine categories of VINDICATE, we now have 27 theoretically possible causes of jaundice, and 27 little drawers to rummage through up in your brain. As you can see in the table on the next page, some of these drawers may turn up empty, but at least we looked in them.
Differential Diagnosis of Jaundice
|V||Pulmonary embolism with infarction||hemobilia|
|I||generalized sepsis, malaria||viral hepatitis||cholangitis|
|N||neoplastic syndromes||hepatoma & hepatic metastases||pancreatic carcinoma & cholangiocarcinoma|
|D||alcohol, solvent exposure, contraceptives|
|C||hereditary spherocytosis & other causes of hemolysis||Gilbert, Rotor, Dubin-Johnson & Crigler-Najar syndromes||biliary atresia|
|T||soft tissue injury with large hematomas||hepatic trauma|
Hopefully, VINDICATE helped you to dredge up a few more entities in your differential diagnosis that you might have otherwise missed. It certainly helps me to dredge them up.
One last thing about mnemonics: sometimes they are appropriate and helpful and sometimes they’re not. I offer the following two examples as a reductio ad absurdum proof of this point.
1. OMONTETE I learned this one on a pediatric rotation while I was in med school. It was given to me as a memory aid for use in newborn physical examinations.
2. ” Frank sings those fine, fine songs — sings even new tunes ever tunefully.” I learned this one sometime during my radiology residency, just in time for boards…..
Wise sayings in musculoskeletal radiology
As Captain James T. Kirk keeps saying, “Spock, sometimes logic is not enough!” Once one has applied all of the logical rules, one still may not have a definite diagnosis. In such cases, it is helpful to apply various empirical rules that have been noted and collected over the years by various wise and venerable musculoskeletal radiologists. These are the radiological equivalent of some of the wise sayings from other walks of life shown below.
|Red sky in morning, sailors take warning — Red sky at night, sailor’s delight.||Sailors||Weather prediction|
|Beer and whiskey, mighty risky — whiskey and beer, never fear.||Drunkards||Estimation of the a priori probability of big time yorfing given excessive consumption of short chain hydrocarbons|
|Red and yellow, kill a fellow — red and black, OK, Jack.||Boy Scouts||Distinguishing the highly poisonous coral snake from the similar but nonpoisonous king snake|
|A long lesion in a long bone.||Radiologists||A finding suggestive of fibrous dysplasia|
There is an awful lot more that one should know about fibrous dysplasia besides the fact that it often presents as a long lesion in a long bone, but this can still be a very helpful little rule at times. Therefore, in each chapter, after I have laid a logical framework for approaching a radiologic problem, I will then try to list as many “wise sayings” about that problem that I can remember.
A final word
Finally, remember the old joke about the famous violinist who came to New York to perform in Carnegie Hall. His plane got in late and then his cab driver got lost. In desperation, he stuck his head out of the cab window and yelled to a woman passing by: “Hey, lady! How do I get to Carnegie Hall?” She yelled back, “Practice!”
So, as Jean-Luc Picard often says, “Make it so!”