How Doctors Think – Jerome Groopman
During daily rounds, the author, an oncologist and hematologist, became concerned that members of his medical staff were not thinking deeply or broadly, running to their mobile to consult decision trees and lists of symptoms. Like an old guy would, he first thought that These Kids Today weren’t up to the high diagnostic skills levels of The Days of the Giants, when he was an intern. But being a smart guy, he got over that knee-jerk nostalgia and started looking at the research on different ways of thinking among doctors.
Linear thinking is when step by step we go through data and draw a conclusion. This is not the method in medicine. Research among physicians indicate they think while they do. From their first step into an exam room, doctors blend observations of the patient, the history, the risk factors, and the exam. Their minds draw a conclusion from all this data. Doctors create a pattern in their mind and superimpose on the individual patient the template of a typical case of whatever disease they think they are encountering. This process comes up with the correct diagnosis in 80% of cases.
This means in 10 to 20% of cases diagnoses are delayed or incorrect. A 2023 study concluded because of diagnostic errors, total serious harms annually in the USA add up to about 795,000 people permanently disabled or dead in a year.
The vast majority of medical errors are thinking errors. Research in cognitive psychology indicates that heuristics – using shortcuts or rules of thumb as decision strategies – are a major source of error.
Anchoring is when the human brain anchors itself on the first bit of data it meets. This could be related to our human nature, hardwired into us from eons ago: sense movement in bush, assume snake not sparrow, run, escape. In a modern hospital far from the savannah, however, doctors fixate on certain features of the patient’s presentation too early in the workup.
Availability is another source of error. It is a bias that relies on what immediately comes to mind to make quick decisions and hasty judgements. For instance, the doctor sees 12 cases of flu in two days and figures the 13th person he sees with the same symptoms is the flu. But really it is a something less frequent, for example aspirin toxicity which has similar symptoms to flu.
Attribution. Our culture teaches us stereotypes as a short-cut to sizing people up. Unshaven dude in the ER shows up in smelly clothes, griping his belly has swollen up, and claims he has only one drink a day. The intern can’t believe this old rummy has only one drink a day. But the attending says, Check for Wilson’s disease. And that is what it is. And the family confirms he has only one drink a day.
Confirmation bias is when we selectively gather and interpret information to conform with our beliefs. A type of confirmation bias in medical settings is called diagnostic momentum, the tendency of a diagnosis to be accepted and passed on, with little or no examination of the underlying evidence for its validity.
How then do we answer, “When you hear the hoofbeats, it’s horses, not zebras?” Patients should ask doctors open-ended questions, such as “What else could explain the cause of my problem,” or “What does not fit the initial diagnosis,” or “What might be another problem in addition to this symptom.” Help the doctor think outside the box and out of the course of day to day routine.