In medicine we use all of our senses to evaluate patients. If I open up an abscess and I’m overwhelmed by the smell of rotted cheese, I can be pretty sure the abscess started as a sebaceous cyst. If I hear a “whooshing” sound over the abdomen in a smoker, I look more closely for an aortic aneurysm. A subtle buzzing sensed through my fingers can indicate an arterial fistula. And the glowing yellow of jaundiced skin can point me straight to the liver.

But what about subtle changes? With just our senses, distinguishing normal from abnormal is a coveted skill, one that some argue is being lost. Experience makes a difference and can help save patients from misdiagnosis and unnecessary testing. Where we train can make a big difference. As I wrote earlier, skin diseases look different depending on the underlying skin tone.

I bring this up because of the “flesh-colored dress” debacle. Michelle Obama wore a dress to a state dinner which an AP reporter described with the 1960s Crayola term “flesh color”, despite the fact that the dress’s color bore little resemblance to Mrs. Obama’s skin.

Distinguishing “normal” from “abnormal” is no trivial matter in medicine: one person’s “flesh color” is another person’s anemia. I frequently have residents report to me that a patient has scleral icterus, a yellow discoloration of the white of the eyes due to liver disease. Often enough, there is no icterus, but simply what we call “muddy sclera”, a common variant seen mostly in darker-skinned older people. “Flesh-colored” is a descriptor that would never fly in medicine. If anything, “identical in color to the surrounding skin”, but never “flesh-colored”.

The news story on Mrs. Obama revealed a deeply-ingrained bias on the part of the reporter and the editor. In medicine, these same biases exist of course. It’s interesting that even dark-skinned doctors often make these same errors (“I cannot tell if there is any redness because the patient is black.”) In training doctors we try to root out these biases and encourage objective descriptions with clear referents, but in medicine, it’s just as hard as everywhere else.