Differential Diagnosis Workup¶
Workflow — instantiates Abductive Explanation Selection
Lists candidate conditions or causes, weighs symptoms and tests, and narrows to a working diagnosis.
The Differential Diagnosis Workup is the clinician's routine for turning a symptom cluster into a ranked list of candidate conditions and then narrowing it by ordered tests. Its defining move — the thing that separates it from a generic comparison — is that it ranks candidates by two axes at once: how common the condition is in this patient (its base rate) and how dangerous it would be to miss. A rare-but-lethal rival stays near the top of the work-up even when a mundane explanation is more probable, and the workflow gates treatment to confidence, so the least-likely killer often gets ruled out before the likeliest nuisance gets treated.
Example¶
A 55-year-old arrives with chest pain, a mild fever, and a productive cough. The clinician records the observation as a symptom cluster, keeping it distinct from any label. The differential: community-acquired pneumonia and musculoskeletal strain (both common at this presentation), pericarditis (less common), and pulmonary embolism and early myocardial infarction (rarer here, but must-not-miss).
Base rates make pneumonia the leading candidate. But the workup does not chase pneumonia first — it orders an ECG and a D-dimer, because a missed embolism or infarction is catastrophic and both are still plausible. Those come back reassuring; a chest film confirms an infiltrate. Pneumonia becomes the working diagnosis, embolism is retained as live until the D-dimer clears it, and antibiotics begin only once the dangerous rivals are excluded and the picture is coherent — treatment gated to confidence, not started on the first plausible story.
How it works¶
- Record the symptom set as observation, before interpretation.
- Enumerate the differential, deliberately including must-not-miss conditions even when unlikely.
- Rank on two axes: prior probability and downside of missing it.
- Sequence tests to rule out the dangerous-and-plausible candidates first, choosing each test for how well it separates rivals.
- Declare a working diagnosis with retained rivals, and hold treatment behind an action–confidence threshold appropriate to the intervention's risk.
Tuning parameters¶
- Differential breadth — a wide net versus a focused shortlist. Wider lowers miss rate but drives over-testing and delay.
- Base-rate versus danger weighting — how much the "worst-case, low-probability" rival is elevated. Tilt toward danger in high-acuity settings; toward base rate in screening.
- Test-ordering aggressiveness — how many discriminating tests fire before committing. More tests buy certainty at cost, radiation, and false positives.
- Empiric-treatment threshold — how much confidence licenses starting treatment before the diagnosis is confirmed.
When it helps, and when it misleads¶
Its strength is disciplined narrowing under time pressure: it captures the clinical maxim that common things are common while forcing an explicit check for the rare catastrophe, so a vivid first impression cannot close the case alone.
Its failure mode is anchoring on the initial diagnosis and reading later data to fit it. A subtler trap is forcing a single unifying explanation when the patient in fact has several concurrent conditions — Occam's razor overreaching where Hickam's dictum ("a patient can have as many diseases as they please") applies.[n1] The classic misuse runs the other way: defensive over-ordering that treats every rare rival as equally live, burying signal in tests. The guarding discipline is to keep the must-not-miss list explicit and to reopen the differential when the patient fails to respond as the working diagnosis predicts.
How it implements the components¶
surprising_observation_record— captures the presenting symptom cluster distinct from any diagnostic label.candidate_explanation_set— the differential, deliberately seeded with dangerous rivals.base_rate_context— prevalence priors that anchor the initial ranking of candidates.discriminating_evidence_plan— the ordered tests chosen to separate rivals, dangerous-and-plausible first.action_confidence_boundary— the rule that gates treatment to confidence and the intervention's risk.
It does not implement the formal explanatory_fit_criteria grid (Inference-to-Best-Explanation Matrix), the durable revision_trigger ledger (Abduction Log), or the explanation_owner handoff record (Explanatory Case Memo).
Related¶
- Instantiates: Abductive Explanation Selection — the domain-specialized clinical workflow of the archetype.
- Sibling mechanisms: Inference-to-Best-Explanation Matrix · Explanatory Case Memo · Anomaly-to-Hypothesis Workshop · Abduction Log · Disconfirming Probe Plan · Forensic Scenario Reconstruction · Model-Debugging Hypothesis Loop
Editorial Notes¶
Form Classification¶
Form family: Protocol, Workflow & Routine
Rationale: Differential Diagnosis Workup operates as a repeatable ordered procedure or handoff sequence that coordinates action because it lists candidate conditions or causes, weighs symptoms and tests, and narrows to a working diagnosis.
Independent corroboration: The frozen evidence defines Differential Diagnosis Workup as 'Lists candidate conditions or causes, weighs symptoms and tests, and narrows to a working diagnosis', so its operative form is Protocol, Workflow & Routine.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Medicine & Healthcare
Origin pattern: Single lineage
Present-day reach: Specialized
Rationale: Clinical practice established the differential workup that ranks conditions by probability and consequence, then gates treatment on ordered testing.
Review resolution: Clinical practice established the differential workup that ranks conditions by probability and consequence, then gates treatment on ordered testing. The workup cohered in clinical medicine, with statistical testing serving the method rather than originating a separate lineage.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Hickam's dictum — "a patient can have as many diseases as they damn well please" — is the customary counterweight to Occam's razor in medicine, warning against collapsing multiple real conditions into one tidy explanation. The differential workup must sometimes hold a conjunction of candidates rather than a single winner. ↩