How Doctors Think

A master cabinetmaker and a novice are given two identical planks of wood, each with a subtle, S-shaped grain running through its core. The novice, eager to create a perfect tabletop, sees the grain as a defect. He spends hours trying to sand it down, plane it flat, and ultimately hide it under a dark stain, fighting the wood’s nature at every turn. His final product is functional, but lifeless; the wood’s character has been erased. The master, however, sees the same grain as the wood’s story. She designs the entire piece around it, letting the S-curve guide the shape of the table legs and the edge of the surface. The grain becomes the central, celebrated feature—a line of beauty that makes the piece unique and alive. Her approach was about listening to the material and collaborating with its inherent nature.

This same divide in thinking—seeing a patient’s story as either a deviation from a textbook case or as the central, guiding feature of their diagnosis—is the puzzle that drove Dr. Jerome Groopman to write this book. As a distinguished professor and practicing physician at Harvard Medical School, Groopman had seen firsthand how even the most brilliant doctors could fall into the trap of the novice, sanding down the unique details of a patient’s life to fit a standard diagnosis, sometimes with disastrous consequences. After his own harrowing experience of being misdiagnosed, he became obsessed with understanding the hidden thought processes, the biases, and the flashes of intuition that shape a doctor's mind. He realized that the most effective physicians were more like master artisans; they knew the textbooks, but their true expertise came from their ability to read the unique grain of each person’s illness.

Module 1: The Cognitive Traps of Diagnosis

Medical school drills a specific method into students. It's slow, analytical, and systematic. You take a history. You do a physical exam. You create a list of possible diagnoses. But in the real world of a busy clinic or a chaotic ER, this deliberate process breaks down. Doctors rely on mental shortcuts, or heuristics. These shortcuts are essential for efficiency, but they are also packed with hidden dangers. The book reveals that most diagnostic errors arise from flaws in thinking.

One of the most common errors is the Representativeness Error, where a doctor's thinking is guided by a stereotype of a patient or disease. A classic example is Dr. Pat Croskerry, an ER physician. A fit, athletic man in his forties named Evan McKinley comes in with chest pain. He looks healthy. He doesn't fit the prototype of a heart attack patient. Croskerry attributes the pain to muscle strain and sends him home. The next day, McKinley returns with a massive heart attack. The doctor was blinded by the prototype. His mind saw a healthy man, not the dangerous symptoms.

Another trap is the Attribution Error, which happens when a doctor stereotypes a patient and dismisses their complaints. Think of Charles Carver, a retired sailor who came to the ER with fatigue and a swollen abdomen. He was unkempt and had alcohol on his breath. The medical team immediately labeled him an "old, foul-smelling, rum-swilling sailor." They assumed alcoholic cirrhosis and wanted to discharge him quickly. Dr. Donald Redelmeier, however, recognized the team's disgust. He forced them to look past the stereotype. It turned out Carver had Wilson's disease, a treatable genetic disorder. The attribution error almost killed him.

Here's the thing. Positive feelings can be just as dangerous as negative ones. This leads to the Affective Error, where a doctor's affection for a patient leads to under-investigation. Groopman shares a personal, painful story. He was treating a young professor, a fellow runner, for bone cancer. He liked this patient immensely. One morning, the patient had a fever. Groopman, wanting to spare his friend discomfort, broke his own routine. He didn't ask the patient to roll over for a full exam. He missed a developing abscess that led to life-threatening septic shock. The patient survived, but Groopman learned a hard lesson. Caring too much can make you see what you want to see.

Module 2: The Tyranny of the First Impression

Once a doctor forms an initial idea, it can be incredibly difficult to dislodge. This is a powerful cognitive bias known as Anchoring. The first piece of information, or the first diagnosis considered, becomes an anchor that pulls all subsequent thinking toward it.

Groopman introduces us to Dr. Harrison Alter, an ER physician. A woman named Blanche Begaye comes in with fever and rapid breathing. The community is experiencing a viral pneumonia outbreak. That diagnosis is "available" in Dr. Alter's mind. He anchors on it. He then engages in Confirmation Bias, selectively hearing information that supports his anchor. The patient mentions taking "a few aspirin." Perfect. That fits the story of a cold. He ignores contradictory evidence, like a normal chest X-ray. The real diagnosis? Aspirin toxicity. Blanche had taken a life-threatening overdose. The anchor of "viral pneumonia" nearly drowned the truth.

This problem gets worse in a team setting. It creates a phenomenon called Diagnosis Momentum, where an initial, unverified diagnosis is passed along and gathers force. At Boston Children's Hospital, an infant named Shira was diagnosed with SCID, a severe immune disorder. The diagnosis was "consistent with SCID." That phrase became a mantra. It was repeated on rounds every morning. The diagnosis took on a life of its own, rolling like a boulder toward a dangerous, unnecessary bone marrow transplant. The team stopped thinking critically. They were just following the momentum.

So what's the antidote? One patient's mother, Rachel Stein, shows us the way. Her daughter was Shira. Rachel, a layperson, educated herself. She saw inconsistencies the doctors missed. She noticed her daughter's immune cells were actually rising, which contradicted the SCID diagnosis. She politely but persistently questioned the team. She forced them to re-examine the data. The retest showed Shira's immune system was completely normal. A persistent, informed patient can be the most powerful force against diagnostic momentum.