Differential diagnosis Study Guide

📖 Core Concepts Differential Diagnosis – A systematic analysis that separates one disease from others that share similar signs or symptoms. Hypothetico‑Deductive Method – Treats each possible disease as a hypothesis; evidence raises or lowers confidence in each. Pre‑test Probability – The clinician’s estimate (often based on epidemiology) that a disease is present before any test. Likelihood Ratio (LR) – A numeric value that tells how much a test result shifts disease probability. Odds ↔ Probability – Odds = p / (1 − p); Probability = odds / (1 + odds). Empiric Treatment – Starting therapy based on the most likely diagnosis when definitive proof is still pending. 📌 Must Remember Differential diagnosis is both a diagnostic tool and a safety net for life‑threatening conditions. Stepwise Process: Gather → List → Prioritize → Test (rule‑out) → Empiric treat. Bayes’ theorem underlies probability updates: $$ P(\text{Disease}|\text{Finding}) = \frac{P(\text{Disease})\times P(\text{Finding}|\text{Disease})}{P(\text{Finding})} $$ Post‑test odds = Pre‑test odds × LR. Convert back to probability for decision‑making. Test selection: High specificity → “rule‑in” (increase probability of already likely disease). High sensitivity → “rule‑out” (decrease probability of competing diseases). 🔄 Key Processes Gather Information – History, physical exam, basic labs. Generate List – Write or mentally note all plausible diseases (use mnemonic to avoid blind spots). Prioritize – Weigh each candidate by: Baseline incidence (epidemiology). Severity if missed (life‑threatening vs. benign). Order Tests – Choose the test with the greatest LR impact for the top candidates. Update Probabilities – Convert pre‑test probability → odds, multiply by LR, convert back → new probability. Decide – If a single disease now has a high enough probability → treat definitively; otherwise consider empiric therapy for the most likely cause. 🔍 Key Comparisons Epidemiology‑Based Method vs. LR‑Based Method Epidemiology: Starts with population incidence → rough pre‑test probability. LR: Refines that probability using objective test performance. High‑Specificity Test vs. High‑Sensitivity Test Specificity: Best for confirming a suspected disease (↑ post‑test probability). Sensitivity: Best for excluding diseases (↓ post‑test probability). ⚠️ Common Misunderstandings “If a test is positive, the disease is present.” Forgetting the pre‑test probability → LR may still leave post‑test probability low. Treating empirically without a “best guess.” Empiric therapy should follow a ranked list, not be random. Assuming the first plausible diagnosis is correct. Systematic ruling‑out is required, especially for high‑risk conditions. 🧠 Mental Models / Intuition “Probability → Odds → LR → New Odds → Probability” – Think of a see‑saw: each test tips the balance according to its LR. “The most dangerous missed diagnosis wins the list.” – When in doubt, prioritize conditions where delayed therapy is catastrophic. 🚩 Exceptions & Edge Cases Mild laboratory abnormality – May stop after excluding common serious causes; exhaustive list unnecessary. Severe pain or life‑threatening presentation – Require broader differential and more definitive testing before narrowing. Rare diseases with high severity – Even with low epidemiologic probability, they stay on the list if the clinical picture fits strongly. 📍 When to Use Which Use Epidemiology‑Based Estimates when no test results are yet available and you need an initial ranking. Apply Likelihood Ratios after you have a test result and want to quantitatively update probability. Choose High‑Specificity Tests when your top candidate already has moderate/high pre‑test probability. Choose High‑Sensitivity Tests when you need to exclude a serious alternative before committing to therapy. 👀 Patterns to Recognize “Red‑Flag” symptoms (e.g., sudden severe headache, chest pain) → automatically elevate life‑threatening candidates. Clusters of findings that match a single disease’s typical presentation → high pre‑test probability before testing. Discrepancy between epidemiologic risk and clinical picture → signals need for broader differential. 🗂️ Exam Traps Distractor: “A test with LR = 1 does not change probability.” → True, but the trap is forgetting that pre‑test probability still matters for management decisions. Distractor: “High sensitivity always rules‑in disease.” – Sensitivity only helps to rule out; misreading this leads to false confidence. Distractor: “If a disease is rare, you can ignore it.” – Rare but high‑mortality diseases stay on the list when the presentation is compatible. Distractor: “Empiric therapy equals guessing.” – In fact, empiric therapy follows the educated best guess after systematic ranking and partial testing.