Illness Scripts
Experts keep clinical knowledge as typical patterns, called illness scripts. They diagnose by matching the case to a script — not by checking off facts one by one.
Observe
An illness script is a mental picture of a condition. It holds three things: who tends to get it, what goes wrong inside the body, and how it usually looks. Take strep throat — who: a school-age kid; what goes wrong: a strep germ infects the throat; how it looks: sore throat, fever, no cough. Experts store many scripts and match a new case to one, instead of reasoning out every fact from scratch.
You build scripts by seeing many real, varied cases. Each one sharpens the pattern in memory, until the right script comes to mind on its own — once you frame the case in a clear line.
This shapes how we think about error. Researchers call it the knowledge-access view. It asks two questions: Is the right script even in your memory? And did you call it up for this case? A miss means the script was missing, or it was there but did not come to mind.
A miss can happen in fast thinking AND in slow, careful thinking. So error is not just a fast slip that a slow double-check is sure to catch. If the script is not in memory, slowing down will not find it.
Say the one-line problem representation first — age, key features, time course. This framing is what decides which scripts come to mind.
Go deeper — worked examples & detail
What a script is made of
Every script holds three parts. Here they are, with a strep-throat example so each part is concrete.
- Setup — what raises the odds: age, history, exposures. (Strep: a school-age kid, sick classmates.)
- The fault — what goes wrong inside, and why the features hang together. (Strep: a germ infects the throat.)
- How it looks — the signs and symptoms you actually see. (Strep: sore throat, fever, no cough.)
- Scripts grow from many real cases over time — not from memorizing facts alone.
Working a case by script
Reason in this order. Frame the case, see which scripts light up, then find the one feature that tells your top two apart.
- Say the one-line problem representation first: age, key features, time course.
- Ask which scripts that line brings to mind — and which it argues against. One line can light up two or three at once.
- Name the discriminating feature for your top two scripts, and ask about it. That is what separates them.
- A shared finding cannot separate them. It may fit both, but it does not move one ahead.
Why a shared finding settles nothing
This is the engine of the whole method. When two scripts both fit, only the right kind of finding can pull them apart. Here is the reasoning laid out as structure.
- Premise 1A finding shared by both scripts raises each one by the same amount, so it moves neither ahead.
- Premise 2A discriminating feature is expected in one script and not the other, so finding it raises one and lowers the other.
- Premise 3Telling two scripts apart means moving one ahead of the other.
The conclusion follows from the premises by their plain meaning — that makes it valid. The premises hold for how scripts compare, so it is sound here. The takeaway: stacking up shared findings feels like growing certainty, but it separates nothing. Hunt for the feature expected to differ.
A note on what kind of reasoning this is
Matching a case to a script is not proof. It is a best-fit guess that you then test — what logic calls abductive reasoning (reasoning to the best explanation). The script that fits best is a lead, not a verdict. You confirm it by asking after the discriminating feature, not by counting how cleanly it fits.
- Best-fit match = a strong lead, never a sure thing.
- You test the lead by asking the one question that could move your top scripts apart.
- A clean, fast fit is a reason to keep checking — not a reason to stop.
Key terms — 4
- Illness script
- A stored, typical pattern for one condition — its usual setup, what goes wrong inside, and how it shows up. Example: panic attack — a young adult, a surge of fear, racing heart and short breath that peak in minutes.
- Problem representation
- A one-line summary of the case — age, key features, time course — built before you name any diagnosis. Example: "teen, two weeks of low mood plus poor sleep, no clear trigger." This line decides which scripts come to mind.
- Discriminating feature
- A finding you expect in one script and NOT in the other. Because it should differ between them, asking about it tells the two apart. A finding shared by both cannot. Example: a past manic episode — expected in the bipolar script, not in the plain depression script.
- Knowledge-access view
- A research idea that diagnostic error is mostly about which scripts are in memory and whether you call them up — not one fast bias that a slow check always fixes. On this view, a missing script stays missed even when you slow down.
Practice
A script can fit too fast. One problem-representation line often lights up two or three scripts at once. The richer and more familiar one feels, the easier it is to stop looking — exactly when an unasked feature could have pointed elsewhere.
Run the discriminating-feature move on a case
- Premise 1A 14-year-old (fictional) has weeks of irritability, a sleep change, and falling grades, with no clear trigger.
- Premise 2That one line brings up three scripts at once: plain depression, bipolar-spectrum, and adjustment reaction.
- Premise 3Each has its own discriminating feature: a past distinct manic or hypomanic episode points to bipolar; a clear recent stressor with a tight time course points to adjustment; long duration with no trigger fits plain depression.
Irritability, poor sleep, and falling grades sit in all three scripts, so listing them again favors none. The discriminating features — a past manic episode, a recent stressor, the duration — are what move one script ahead. "No clear trigger" is itself a clue: it argues against adjustment. Naming these out loud keeps every option open instead of locking onto the first clean fit.
Check your reasoning
A fellow (fictional) builds this problem representation: "adolescent, several weeks of irritability plus a sleep change plus a drop in grades, no clear trigger." It activates a depressive script and a bipolar-spectrum script at once. Which next move best reflects reasoning by illness script — keeping both scripts live rather than locking onto one?
Commit to one — then see the reasoning behind each.
Implement
- Say the one-line problem representation out loud first: age, key features, time course.
- Name which scripts that line brings to mind — often two or three — and which it argues against.
- Pick your top scripts. Name the one feature you expect in one but NOT the others.
- Ask about that feature. Let the answer move a script ahead — do not just re-count shared findings.
- Treat a fast, clean match as a reason to ask one more separating question, not a reason to stop.
Watch for
A script that fits too fast. The richer and more familiar it is, the easier it is to stop looking. A clean early match raises your confidence and quietly ends the search — exactly when an unasked discriminating feature could have closed the case on the wrong pattern. Pair this with the debiasing layer so an early match does not shut the door.
Make it stick. Recall these, grade yourself, and they come back on their own schedule.
Spaced recall
5 cards for this tool
Recall it, then reveal — grading honestly is what makes it stick.
Grounded in the research — 5 sources
- Bordage, G. (1994). Elaborated knowledge: A key to successful diagnostic thinking. Academic Medicine, 69(11), 883-885. https://doi.org/10.1097/00001888-199411000-00004
- Schmidt, H. G., & Rikers, R. M. J. P. (2007). How expertise develops in medicine: Knowledge encapsulation and illness script formation. Medical Education, 41(12), 1133-1139. https://doi.org/10.1111/j.1365-2923.2007.02915.x
- Pelaccia, T., Tardif, J., Triby, E., & Charlin, B. (2011). An analysis of clinical reasoning through a recent and comprehensive approach: The dual-process theory. Medical Education Online, 16, 5890. https://doi.org/10.3402/meo.v16i0.5890
- Si, J. (2022). Strategies for developing pre-clinical medical students' clinical reasoning based on illness script formation: A systematic review. Korean Journal of Medical Education, 34(1), 49-61. https://doi.org/10.3946/kjme.2022.219
- Norman, G., Pelaccia, T., Wyer, P., & Sherbino, J. (2024). Dual process models of clinical reasoning: The central role of knowledge in diagnostic expertise. Journal of Evaluation in Clinical Practice, 30(5), 788-796. https://doi.org/10.1111/jep.13998