Adriaan de Groot studied chess grandmasters in 1946 and found something that shouldn’t have been surprising but still was. When shown a board position for a few seconds, grandmasters could place nearly every piece correctly. Average players placed a handful. The difference wasn’t effort. The grandmasters weren’t thinking harder. They saw the board as a single recognized pattern rather than a collection of individual pieces. They weren’t calculating. They were recognizing.
What Expertise Actually Looks Like
Gary Klein spent years watching firefighters make decisions in burning buildings. His finding: experts don’t evaluate options the way anyone expects. They don’t generate alternatives and compare them. They recognize a situation as a type of situation, generate one course of action, run a quick mental simulation to check whether it’ll work, and either act or adjust.
That’s how expert clinical judgment works. The experienced PA who hears a history and knows, before the stress test, that this isn’t typical angina isn’t running through a differential. She’s pattern-matching. The pattern was built across hundreds of encounters, refined by feedback, corrected by being wrong and eventually finding out. Novices calculate; experts recognize; there’s no shortcut between those two states.
What NP and PA Education Teaches Instead
Most NP and PA education is organized around information transfer. Here’s the guideline. Here’s the dosing. Here are the diagnostic criteria, the contraindications, the red flags.
That content matters. A student who doesn’t know the HFpEF criteria can’t reason about heart failure. But knowing the criteria and applying them to a patient who also has COPD, decompensated renal function, and an LVEF of 48% are not the same activity. The guideline doesn’t tell you what to do when three of its own recommendations conflict in one patient.
I teach the ACNP program at SJU and practice cardiology at WellSpan. The gap between what graduates know and how they reason in the room is visible from both ends. Students come out of programs knowing more than I knew at their stage. The reasoning catches up more slowly, and the trajectory is longer than most programs acknowledge.
Why the Gap Persists
Expertise researchers put the threshold at roughly 10 years of deliberate practice in most complex domains. Clinical training compresses some of that through volume and direct supervision, but the basic constraint holds: pattern recognition requires patterns. Enough patients. Enough variation in how the same diagnosis presents. Enough feedback when you got it wrong.
Programs can’t give students 10 years of clinical experience in two years of graduate school. That’s a constraint, not a criticism. But what programs can do is design around the gap. Simulations with immediate, specific feedback. The same diagnosis presented three different ways. Cases where the correct answer isn’t in the guideline because the patient violates the study population. Assessments that test reasoning, not recall.
Collins and colleagues described the instructional framework for this in 1989. Most programs have read about it. Most don’t build for it.
The Part That’s Easy to Miss
I watch a version of this play out every year. A student knows the guideline. She can recite it on the written exam. In the sim lab, she misses the scenario where it applies because the presentation didn’t look like the chapter.
She wasn’t reasoning from the guideline. She had stored the guideline as a fact. The connection between the fact and the clinical situation takes time to build, and it builds through exposure, not memorization.
You can’t assign that connection. You can create the conditions where it develops: enough clinical exposure, enough variation, enough feedback specific enough to reshape thinking rather than just confirm the answer.
What Doesn’t Fix It
More content doesn’t fix it. NP programs are already too dense for most students to fully absorb. More guidelines, more hours, more assigned readings don’t move the needle on reasoning.
Board prep doesn’t fix it either. Recognizing the answer pattern on a multiple-choice item is a real but narrow skill. It helps at the margins on the exam and transfers to clinical practice to the degree that the underlying reasoning was already there.
What closes the gap is time with clinical problems that don’t resolve cleanly, combined with feedback specific enough to tell a student not just that she was wrong but what she missed and why the correct answer followed from the clinical picture.
The grandmaster looking at that chessboard wasn’t born seeing patterns. She built them, thousands of positions at a time, most of them in games she lost. That’s the mechanism. There’s no faster version of it. The question is whether the curriculum is designed to accelerate the process or just to document that the content was covered.