How to score a 3

Initial Differential

Breadth before the data, and whether every entry could produce this syndrome in this patient.

Directed History

Whether your questions were driven by your own hypotheses.

Differential Diagnosis

Ranking, and case-specific evidence attached to each entry.

Directed Testing

Tests tied to hypotheses, clustered, without repetition or padding.

A bare list of three correct diagnoses scores a 2. The same three with evidence score a 3. This single habit is worth two criteria, which makes it the highest-value change available to you.

Fix: Write every differential in three parts: the diagnosis, two specific findings that support it, and one that argues against it or that you would need to confirm it. Then one clause on why #1 outranks #2.

You can capture every key finding a case offers and still take a 2 here. Answering what the patient volunteers is not the same as asking what your hypotheses demand — and the gap is most visible on the can't-miss entries, which students list and then never pursue.

Fix: One question owed per differential. If you write it down, you ask about it.

Breadth is only rewarded when every entry survives scrutiny. One diagnosis that could not produce this presentation drags a broad, otherwise-correct list off a 3 — and reflexively ruling out conditions that were never plausible reads as padding, not thoroughness.

Fix: Run the admission test on every item: could this diagnosis, in this patient, produce this exact syndrome? Build by category — psychiatric, substance, medical, neurologic, iatrogenic — which is inherently broad and inherently clean.

"Definitely", "obviously", "I bet", "they already diagnosed her" — these are the written markers of premature closure, and reasoning from a previous clinician's prescription rather than from criteria is how you inherit someone else's anchor.

Fix: Your notes are graded output, not scratch paper. Replace the conclusion with the criterion: not "this is definitely PTSD" but "duration beyond one month, which places this past acute stress disorder." Same answer, but now it shows the reasoning.

Substance intoxication

“What exactly, how much, by what route — and when was the last use? Anything new in the past 72 hours?”

A named substance without a last-use time is not a substance history.

Substance withdrawal

“When was the last drink? Ever had shakes, seizures, or seen things when you stopped?”

Two to three drinks several nights a week is enough to matter.

Neuroleptic malignant syndrome

“Any antipsychotic — or an antiemetic like metoclopramide or prochlorperazine — started or increased in the last two weeks?”

Lead-pipe rigidity, onset over days, hyporeflexia.

Serotonin syndrome

“Any SSRI, SNRI, tramadol, linezolid, triptan, MAOI, or dextromethorphan — and was anything added recently?”

Clonus and hyperreflexia, legs greater than arms, onset in hours.

Anticholinergic toxicity

“Any diphenhydramine, doxylamine, sleep aids, tricyclics, or scopolamine?”

Dry, flushed, mydriatic, retaining urine, absent bowel sounds.

Delirium / medical cause

“Is it waxing and waning through the day? Any fever, head injury, infection, new medication?”

Fluctuating attention is the tell.

Autoimmune encephalitis

“Any viral prodrome, seizures, memory loss, or odd mouth and face movements?”

Worth one question in any new-onset psychosis without a clear cause.

Primary psychotic disorder

“What were they like six months ago — work, hygiene, friendships? Any slow withdrawal before this week?”

Collateral is the only route to the prodrome.

Bipolar spectrum

“Any stretch of days with less need for sleep — up and going, not tired?”

Need, not amount. The word does the diagnostic work.

Borderline personality disorder

“Do the shifts follow a rejection, and do they last hours or days? How do you feel between them?”

Chronic emptiness and identity disturbance sit on the baseline.