And so basically, as medical students we're taught pattern recognition and probabilities. Have a person in the hospital that has sudden kidney failure a few days after an aggressive bacterial infection? That's the classic pattern for aminoglycoside toxicity.
Or maybe you're told your patient who has a terrible cough that won't go away. She's worried because she watched a movie with someone who had lung cancer that coughed in similar way. Without any other information, sure, some sort of lung or throat cancer is on your differential. Find out she's never smoked and hasn't had unexplained weight loss? Suddenly the chances of her having cancer are much lower. And then it turns out she has high blood pressure and is on an ACE Inhibitor. The odds are completely different now. Cancer is way down and side effect of her medication is really high on the list.
But these are all likelihoods. It's not impossible that our hospital patient wasn't given an aminoglycoside and has kidney disease independent of his/her previous infection. Furthermore, no doctor would ever tell you that there's no chance that the cough is cancer. 40 year-olds who have never smoked sometimes get cancer and not all people on ACE Inhibitors cough. We start with the most likely and move our way down.
My point is that in order for a diagnosis to be made, it has to make it onto the differential in the first place. Furthermore, in order for a diagnosis to be made promptly, the physician has to have an accurate understanding of probabilities.
So what happens when those very probabilities are generated by historically white, male populations and they're taught as though they apply to everyone? Diagnoses get missed and care suffers. Certainly, medical institutions are starting to recognize this. We're routinely told to scrutinize treatment studies that only study Caucasian males because unless your patient is a white guy, the study might not actually apply. And just this week, during our dermatology sequence we were given a lecture entitled, “Dermatology of Pigmented Skin.” In our cardiovascular sequence we were given a lecture entitled, “Cardiovascular Disease in Women.”
I have no doubt that this is a vast improvement from medical curricula in the past. However, we can do better.
The lecture objectives from the dermatology talk:
-Define skin type designations
-Understand the biologic differences in darker skin types
-Discuss common conditions in skin of color
-Illustrate different presentations of diseases in various skin types
I think objectives 1-3 are fantastic and well placed. It makes sense that if there are only certain issues that darker pigmented folks get, that they are presented in the lecture about darker skin color. Furthermore, because the overwhelming majority of American medical students have light skin, they are unaware of certain benign phenomenon that folks of color sometimes get. However, the fourth lecture objective is what I take issue with. Why aren't illustrations of different presentations of disease given to us when we learn about that disease? If we have a whole lecture dedicated to acne and it turns out that acne can look different in folks with darker skin, why wasn't that part of the acne lecture? Furthermore, if we had a lecture in neurology about dermatomyositis, a non-dermatologic disease, which is often identified by a characteristic rash around the eye, why wasn't it brought up then?
It may not seem like that big a deal. If you learn it, you learn it, right? I disagree for a few reasons. The first is mostly rhetorical. By dedicating an hour to acne and countless other hours to other diseases and then segregating out how they all appear in other folks, the implicit assumption that white and male is the norm continues to be perpetuated. This hegemony not only effects the ability for folks outside that culture to succeed, but also impacts the kind of research that gets done and how care gets developed.
For example, for a reason that we don't understand, women often deposit fat in their coronary arteries (the arteries that deliver blood to your heart muscles) in different patterns than men. Men tend to deposit in more local area that cause a discrete narrowing that can be easily seen. Women, on the other hand, frequently distribute deposit fat more equally. Both patterns can ultimately lead to reduced blood flow and thus chest pain. However, when women used to present with chest pain, doctors would tell them that nothing was wrong because when their coronary arteries were imaged, they used the male norm to interpret the images. When they saw no localized narrowing, they assumed that nothing was wrong and told their patients they were crazy. This was so institutionalized that it was even given a name, “cardiac syndrome X.” In fact if you google it, you'll come up with websites explaining how it's psychogenic. These days, many leading cardiologists are finally recognizing that it's perhaps not that women are crazy, our bodies are just different and those symptoms are real.
But perhaps you're not political and rhetorical arguments don't move you; another reason why teaching “atypical” presentations separate is detrimental. As I outlined before, medical care is all based on probabilities. You start with the most likely and then work your way down. However, when something is labeled “atypical” I would argue that it immediately drops down lower on our differentials (if it makes it at all). The word itself means “not representative” which in my mind translates to “not likely.”
And the main problem that I have is that “atypical” is subjective. If the things in medicine that are described as “atypical” or “other” were actually rare, then maybe I could go along with it. However, the reality is that there's no such objective cut off.
For example, in our “Cardiovascular Disease in Women” lecture, we're presented with the idea that myocardial infarction (heart attack) – incidentally, there was a three hour block two weeks before dedicated solely acute chest pain - can present “atypically” in women because they might not get chest pain (a classic symptom). A study published in JAMA looked at nearly half a million patients and showed that one third of patients that present with myocardial infarction do not have chest pain. The study also showed that these folks were more often female or non-white. In my mind, 33% does not say “rare” to me.
However, by teaching this presentation as atypical and separate from the MI lecture, it only enforces the instinct that this is rare. This means it either doesn't make it to the differential as often, or it's put lower on the list. That means you check it later. In an health emergency where they say time = muscle, it's not hard to imagine that these patients receive worse care. The same study showed that these patients were twice as likely to die in the hospital because the amount of time it took to get the right diagnosis and then appropriate care was longer.
And more infuriating, as I said before, atypical is completely subjective based on norms. For example, during our infectious disease unit we learned about the measles. Typically, measles is associated with children and usually it's quite treatable. Children most often recover and go on their merry way (especially these days because we have a vaccination for it.) However, there are rare complications such as pneumonia and encephalitis (inflammation of the brain). Overall, the complication rate in school-aged children is somewhere between 3-5%. In our measles lecture we were taught these complications This was true for all the diseases that we learned and they were not set aside for a separate lecture entitled, “Atypical Presentations/Complications of Infectious Diseases.”
According to the most recent census, non-whites make up 32% of the US population and 1) many dermatologic diseases afflict all races at equal rates and 2) many non-dermatologic diseases have manifestations on the skin. “Atypical” heart attack makes up 33% of cases that present to the hospital. 33%, 32% versus the 3-5% in measles? Those statistics make it hard for me to believe these separations aren't anything more than continued hegemony.