Anticipation- when we perceive what we expect to perceive.
This is an example of how we evaluate the visual world. This specific example comes from an article I read many years ago, one that has occupied a central place in my personal learning of pathology. (It is not, I should note, what is taught to pathology trainees.) In this article from 2006, Dr. McLendon, a neuropathologist, wrote about what can be learned about cognitive processes of decision making by studying how CIA analysts process information to make decisions.* Just like in pathology, CIA analysts must take in information that is incomplete and ambiguous and make important decisions.
One way to examine how this is done is to consider these optical illusions that highlight cognitive biases.
Let's look at the image again.
Most people anticipate the sentences and read right through.
Paris in the spring. Once in a lifetime. Bird in the hand.
But this is misreading the articles (the triangle arrangement helps force this).
Did you see them? The-the. A-A. The-The.
One Monday morning, as a surgical pathology fellow, I was reviewing an appendix. It had been removed over the weekend, and the pathologist and I were going over the cases that had accumulated before we started getting the morning specimens from the ORs. Like most of the weekend specimens, this was an unexpected surgery. It was performed on a patient who came in for emergency appendicitis. I took the slide from the histology technician, laid it on my microscope stage, and examined it. I looked for the hallmark of this condition - acute inflammation, or neutrophils, through the full wall thickness of the appendix. I found that and was done.
"Appendicitis", I said, quickly handing it to the attending pathologist, my superior.
She then looked at the slide and within a moment had stood up to go look at the specimen on the gross station. She evaluated that, asked for another section to look at under the microscope, and sat back down next to me.
"Jena, " she said, "if you only look for appendicitis, that's all you will ever see."
The specimen was in fact an appendiceal carcinoma. This is the tumor that is the reason why pathologists even examine the appendix - to ensure this carcinoma is not overlooked. And I had done so. While common things are common, rare things cannot be missed. That moment burned in my chest and mind, and still does. I had anticipated the answer and that is what I saw. A lesson I'll never forget.
The most frequent human cancer is Basal Cell Carcinoma. I see this a lot in my Dermatopathology practice. You could even say I anticipate seeing these tumors when I see a blue nodular growth in the dermis. These are straightforward to identify. Except when they aren't.
One could quickly assess both this case - nodules of blue cells in the dermis. The most likely tumor, the one I am anticipating? A Basal Cell Carcinoma.
Look again, and look more closely.
The higher power (more magnified view) prompts a pathologist to review what else is in the list of blue nodular tumors in the dermis - and here the trained eye can see that the tumor is a Merkel Cell Carcinoma, and not a Basal Cell Carcinoma. Its not just about how different they look; Merkel cell carcinoma is the deadliest skin cancer, whereas basal cell carcinomas are usually almost just a nuisance, treated by either removal or medication.
(For comparison, here are pictures of a basal cell carcinoma with a similar appearance.)
One rule of thumb I have created for myself is to have habitual tricks when my gut response kicks in. Because the stakes are high, I create a new gut response, one that anticipates the worst possible outcome. I put myself on high alert for the diagnosis that can't be missed:
When I look at an appendix, I try to prove that there is not a cancer.
When I see what looks like a Basal cell carcinoma, I instead assume it is a Merkel cell carcinoma.
Lots of what pathologists are trained to do is meant to offset cognitive pitfalls like the anticipation illustrated above. Fear and humility are not necessarily bad teachers during training. All of medicine works to cultivate appropriate differential diagnosis lists. During training, residents and fellows memorize and grow a list of possible explanations for a given condition, and learn the tests to disprove and prove each of these.
Is it uncomfortable to consider the subjective nature of diagnosis? That's what worries me in talking about these questions. I feel like forging on though, because it's not just pathologists and CIA analysts who are taking in incomplete information. All of us imperfectly perceive reality, form an opinion and make a decision. And we come to these decisions largely unaware of the biases that may affect our judgement. If our biases are unconscious, they shape our perceptions and choices without our knowledge. And none of us can afford that.
*McLendon, Roger MD. Errors in Surgical Neuropathology and the Influence of Cognitive Biases: The Psychology of Intelligence Analysis. Archives of Pathology and Laboratory Medicine, Volume 130, Issue 5 (May 2006)