Case Study 1: Dr. Okafor's Cross-Specialty Transfer

How Diagnostic Reasoning Travels Between Medical Specialties

Background

When you first met Dr. James Okafor in Chapter 2, he was a medical student learning how memory works — specifically, how encoding depth determined whether he'd remember drug mechanisms or merely recognize drug names on a multiple-choice exam. In Chapter 6, you saw him struggle with the effects of sleep deprivation on clinical reasoning, learning the hard way that biology constrains learning no matter how motivated you are.

Now James is in his third year of residency in internal medicine, and he's facing a new challenge — one that has nothing to do with fatigue or memorization. James has spent the last four months on the cardiology service, where he's become genuinely skilled at diagnosing cardiac conditions. He can look at an electrocardiogram and spot the subtle patterns that distinguish one arrhythmia from another. He can hear a heart murmur and narrow the differential to two or three possibilities. He can take a patient's history of chest pain and construct a diagnostic plan that his attending physician, Dr. Ndiaye, describes as "thorough and efficient."

James feels confident. Competent, even.

And then he rotates to the pulmonology service.

(Dr. James Okafor is a composite character based on common patterns in medical education — Tier 3, illustrative example.)

The Problem

James's first day on pulmonology is humbling. A patient presents with shortness of breath, and James instinctively reaches for his cardiology knowledge — could this be heart failure? A valvular problem? An arrhythmia causing poor cardiac output? The pulmonology attending, Dr. Vasquez, gently redirects him: "The chest X-ray shows hyperinflated lungs and flattened diaphragms. The spirometry shows an obstructive pattern. What does that tell you?"

James knows the answer in the abstract — those are signs of COPD. He learned it in his second-year lectures. But he doesn't feel it the way he feels cardiology. In cardiology, the patterns are vivid, automatic, richly connected. In pulmonology, the patterns are thin, effortful, fragile. He's constantly looking things up. He second-guesses himself. He takes twice as long to formulate a plan.

"I feel like a first-year student again," James tells his co-resident, Priya, at the end of the first week. "All that cardiology expertise I built — none of it seems to apply here."

Priya, who rotated to pulmonology before James, gives him a look. "Really? None of it?"

"Well, the knowledge doesn't transfer. Different organ, different diseases, different tests."

"The knowledge doesn't transfer," Priya agrees. "But the reasoning does. You're just not seeing it yet."

The Turning Point: A Bridging Conversation

The breakthrough comes during a case conference with Dr. Ndiaye, James's cardiology attending, who happens to be visiting the hospital for a grand rounds lecture. James mentions his struggle, and Dr. Ndiaye asks him a question that changes everything.

"Tell me about your best cardiology case. Not the diagnosis — the reasoning process you used. Walk me through the steps."

James thinks about a complex case from two months ago — a 58-year-old woman with atypical chest pain, borderline troponins, and a confusing ECG.

"Okay," he says. "First, I gathered the presenting symptoms and history. Chest pain — but atypical, not classic angina. Risk factors: hypertension, family history, smoking. Then I generated a differential — acute coronary syndrome, pulmonary embolism, pericarditis, musculoskeletal pain, and anxiety. I prioritized: ACS was most dangerous even though not most likely, so I wanted to rule it out first. I ordered serial troponins, a repeat ECG, and a D-dimer. The serial troponins trended up slightly, the ECG showed dynamic T-wave changes, and the D-dimer was normal. That moved ACS to the top and dropped PE. I recommended a cardiology consult and stress test."

Dr. Ndiaye nods. "Now strip away the cardiology. What are the steps?"

James pauses. "Gather information. Generate possibilities. Prioritize by probability and danger. Test with targeted investigations. Update based on results. Act on the revised assessment."

"And is that process specific to the heart?"

The insight hits James like a physical sensation. "No. That process works for any diagnostic problem. Lungs, kidneys, brain — the organs change, the diseases change, the tests change. But the reasoning architecture is the same."

"Now tell me," Dr. Ndiaye says, "about a pulmonology case you found difficult this week. But don't tell me the pulmonology. Tell me the reasoning."

The Schema Becomes Visible

James reconsiders his difficult pulmonology case — a 65-year-old man with progressive dyspnea — through the lens of the abstract reasoning schema instead of through the lens of "I don't know pulmonology well enough."

Step 1: Gather. Presenting symptom: progressive shortness of breath over six months, worse with exertion. History: 40-pack-year smoking history, occupational exposure to asbestos, no cardiac history. Exam: diminished breath sounds bilaterally, clubbing of fingertips.

James recognizes that he did this step well. He gathered a thorough history and exam. The skill of knowing what questions to ask and what physical findings to look for — that transferred from cardiology. The specific questions were different (pulmonary risk factors instead of cardiac risk factors), but the process of systematic history-taking was identical.

Step 2: Generate. James had struggled here — his differential felt thin. But when he applies the same approach he used in cardiology (cast a wide net, include common and dangerous diagnoses, organize by category), he generates: COPD, interstitial lung disease (asbestosis or idiopathic pulmonary fibrosis), lung cancer, heart failure with pulmonary congestion, combined pulmonary-cardiac disease.

"I actually had a reasonable differential," James realizes. "I just didn't trust it because the diseases were less familiar."

Step 3: Prioritize. Most likely based on history: COPD (smoking) or asbestosis (occupational exposure). Most dangerous to miss: lung cancer (smoking + asbestos = very high risk). James recognizes this step as identical to his cardiology reasoning — prioritize by probability and danger, address the dangerous-to-miss diagnoses first.

Step 4: Test. Spirometry (assess the pattern of lung dysfunction), chest CT (look for masses, fibrosis, or emphysema), and basic labs. James had ordered spirometry and a chest X-ray but hadn't thought to order the CT. When he applies the cardiology heuristic — "What test would most change my management?" — the CT becomes obvious, because it's the test that distinguishes cancer from benign disease.

Step 5: Update. The spirometry shows a mixed obstructive-restrictive pattern (suggesting both COPD and fibrosis). The CT shows bilateral lower-lobe fibrosis consistent with asbestosis and a suspicious 2-cm nodule in the right upper lobe. James's differential narrows dramatically: likely asbestosis with possible concurrent lung cancer.

"You just diagnosed a complex pulmonary case," Dr. Ndiaye says. "You did it using reasoning skills you built in cardiology. The specific knowledge was new — you had to learn what spirometry patterns mean, what asbestosis looks like on CT. But the reasoning architecture that organized your approach? That was already there. You just hadn't recognized it."

What Transferred (and What Didn't)

James's experience illustrates a critical distinction: knowledge and reasoning are different types of things, and they transfer differently.

What did NOT transfer: - Specific disease knowledge (COPD pathophysiology, spirometry interpretation, asbestosis imaging findings) - Specific test ordering patterns (when to order spirometry vs. ECG) - Specific physical exam skills (auscultating lungs vs. auscultating the heart) - Pattern recognition for specific conditions (the "look" of a COPD patient vs. an ACS patient)

These are domain-specific facts and pattern libraries. They must be learned fresh for each specialty. There's no shortcut.

What DID transfer: - The diagnostic reasoning schema (gather-generate-prioritize-test-update) - The heuristic of prioritizing by both probability and danger - The metacognitive skill of recognizing when a differential is too narrow - The habit of asking "What test would most change my management?" - The ability to tolerate uncertainty and continue reasoning without a definitive answer - The skill of organizing clinical information into a structured presentation - The habit of explicitly stating and then revising probabilities

These are abstract reasoning skills and metacognitive habits. They're not tied to any specific organ or disease. They're structural, not surface. And they dramatically shortened James's learning curve in pulmonology — not because he knew the pulmonology content, but because he already had the reasoning architecture into which the new content could be organized.

The Transfer Accelerator

After the conversation with Dr. Ndiaye, James begins doing something deliberate at the end of every clinical day. He takes five minutes and writes in a notebook — not clinical notes, but transfer notes:

  1. What reasoning process did I use today? (Not what diagnosis — what process)
  2. Is this process specific to pulmonology, or would it work in any specialty? (Identifying the abstract schema)
  3. Where have I used this same process before? (Bridging backward)
  4. Where else could I use it? (Bridging forward)

Within two weeks, James has identified a dozen transferable reasoning patterns. Some are obvious (differential diagnosis generation, test-ordering heuristics). Others surprise him:

  • The "anchor and adjust" pattern: In cardiology, he learned to start with the most common diagnosis and then adjust based on features that don't fit. He's now doing the same thing in pulmonology — and he realizes it's a general pattern of reasoning under uncertainty, not a cardiology trick.

  • The "red flag" scan: In cardiology, he automatically screens for life-threatening conditions before investigating less urgent possibilities. This transfers directly — the specific red flags change (pulmonary embolism instead of myocardial infarction), but the habit of screening for emergencies first is universal.

  • The "sick or not sick" assessment: Within the first 30 seconds of seeing a patient, James makes an intuitive judgment: is this person seriously ill or not? This gestalt assessment, which took months to develop in cardiology, transfers almost immediately to pulmonology. The specific visual cues differ (cyanosis instead of diaphoresis), but the cognitive process — rapid pattern matching against a "seriously ill" template — is structurally identical.

Analysis: The Mechanics of James's Transfer

1. James's transfer was primarily near transfer — within the same broad domain (medicine), between related subdomains (cardiology and pulmonology). The conceptual distance was moderate: different organ system, same patient population, same clinical reasoning framework. This makes James's transfer easier than, say, transferring medical reasoning to legal reasoning (far transfer).

2. The transfer required a bridging prompt. James didn't spontaneously recognize that his cardiology reasoning transferred. He needed Dr. Ndiaye's question — "Walk me through the reasoning, not the diagnosis" — to shift his attention from surface features (specific diseases and tests) to structural features (the reasoning process). This mirrors the Gick and Holyoak finding: the knowledge was there, but the connection wasn't. The bridging question made the connection visible.

3. The abstract schema was the transfer vehicle. Once James articulated the gather-generate-prioritize-test-update schema in abstract terms, it became portable. The schema wasn't tied to any organ or disease. It was a reasoning template that could be filled in with any domain's content.

4. James combined bridging with hugging. His daily "transfer notes" practice was bridging (explicitly connecting reasoning across specialties). His clinical work itself was hugging (practicing diagnostic reasoning under real clinical conditions, not just in classroom simulations). The combination — abstract reflection and concrete practice — produced faster and more robust transfer than either alone.

5. Metacognitive awareness was essential. James's transfer improved dramatically once he started monitoring his own reasoning process rather than just executing it. By stepping back and asking "What am I actually doing when I diagnose?" he made the implicit explicit — and explicit knowledge transfers better than implicit knowledge.

Discussion Questions

  1. Dr. Ndiaye asked James to describe his reasoning process rather than his diagnosis. Why was this reframing so powerful? What does it reveal about the relationship between surface and structural features in clinical reasoning?

  2. James initially felt "like a first-year student again" in pulmonology. Was this feeling accurate? Using the concepts of near transfer and abstract schemas, explain why James's subjective experience of incompetence was misleading.

  3. James's "transfer notes" practice takes only five minutes per day. Design a similar five-minute practice for your own learning: What would you write? What questions would you ask yourself? How would this practice promote transfer?

  4. Not all of James's knowledge transferred — specific disease knowledge, test interpretation, and pattern libraries had to be learned fresh. Does this mean those types of knowledge are inherently non-transferable? Or could they become transferable under different conditions? (Hint: think about levels of abstraction.)

  5. James needed Dr. Ndiaye's bridging prompt to see the transfer. Could he have discovered it on his own? What metacognitive habits would have made spontaneous transfer more likely?

Your Turn

Apply James's transfer strategy to your own learning:

  1. Choose two related domains from your current learning (e.g., two courses, two aspects of a skill, two professional tasks).

  2. Describe your process in one domain — not the content, but the reasoning steps you follow.

  3. Abstract the process — strip away domain-specific details and describe the reasoning in general terms.

  4. Map the process onto the second domain — where does each step appear? What changes, and what stays the same?

  5. Write a "transfer note" for today: What reasoning process did I use? Is it domain-specific or general? Where else could I use it?

This case study connects to: Chapter 2 (encoding depth, schema formation), Chapter 6 (sleep and clinical reasoning), Chapter 12 (deep processing — attending to reasoning structure rather than surface content), Chapter 16 (self-testing and clinical skills), Chapter 21 (learning by doing — clinical learning as experiential), Chapter 25 (expertise development, adaptive expertise, Dr. Okafor's trajectory from novice to expert), Chapter 28 (Dr. Okafor's culminating reflection on his learning journey).