Presentation & problem representation
Compress symptoms, signs, history and context into a representation that preserves discriminating information without prematurely naming a diagnosis.
Subject
Purpose
Human disease and care studied through diagnosis, mechanism, intervention, prognosis, uncertainty and the integration of evidence with individual context.
Structure
Entities → interactions → mechanisms → scales → measurement
Medicine is a decision discipline built on biological knowledge but constrained by incomplete information, competing risks, patient values and variable treatment effects.
Compress symptoms, signs, history and context into a representation that preserves discriminating information without prematurely naming a diagnosis.
Generate and rank plausible causes using prevalence, mechanism, severity and the expected value of further information.
Treat tests as evidence that shifts probability rather than as binary truth machines, with sensitivity, specificity and pretest probability explicit.
Compare benefits, harms, interactions, feasibility and patient preferences rather than optimizing one clinical endpoint in isolation.
Update decisions over time as disease evolves, treatment response appears and new evidence changes the working model.
These separations prevent nearby ideas from collapsing into one another before the subject is understood.
diagnosis ≠ certainty
statistical significance ≠ clinical importance
guideline ≠ patient-specific command
Use these to test whether the model is becoming explanatory rather than merely familiar.
How should diagnostic probability change after a test result?
When is watchful waiting a better intervention than immediate treatment?
How should population evidence be adapted to a patient whose risks and values differ from the study average?
Clinical claims should distinguish randomized evidence, observational evidence, diagnostic accuracy, mechanistic plausibility and expert consensus, with outcome relevance and absolute effects visible.