Makindo Medical Notes

🧠 Understanding Sensitivity, Specificity, and Predictive Values

💡 Key idea: Every diagnostic test gives results that may be true or false relative to the real disease state. The goal is to quantify how well a test distinguishes between health and disease.

🧩 Basic Terms

• True Positive (TP): Patient has the disease ➜ test correctly positive.
• False Positive (FP): Patient does not have the disease ➜ test incorrectly positive.
• True Negative (TN): Patient does not have the disease ➜ test correctly negative.
• False Negative (FN): Patient has the disease ➜ test incorrectly negative.

🔬 Sensitivity

Formula: Sensitivity = TP / (TP + FN)

• Measures a test’s ability to correctly identify those who have the disease.
• High sensitivity = few false negatives → useful for screening.
• Example: If 19 of 20 diseased patients test positive → sensitivity = 19/20 = 95%.

🩺 Tip: “SnNout” — when a test is highly SeNsitive, a Negative result rules disease out.

🧪 Specificity

Formula: Specificity = TN / (TN + FP)

• Measures a test’s ability to correctly identify those without the disease.
• High specificity = few false positives → useful for confirmation.
• Example: If 80 of 100 healthy people test negative → specificity = 80/100 = 80%.

💡 Tip: “SpPin” — when a test is highly SPecific, a Positive result rules disease in.

📊 Using Two-Step Testing

• Begin with a high sensitivity test (broad screening) → captures nearly everyone with disease.
• Follow with a high specificity test (confirmatory) → filters out false positives.
• Example: HIV testing uses this principle — ELISA (sensitive) followed by Western blot (specific).

📈 Positive Predictive Value (PPV)

Formula: PPV = TP / (TP + FP)

• Answers: “Given a positive test, what is the probability the patient actually has the disease?”
• Strongly influenced by disease prevalence — PPV rises as disease becomes more common.
• Example: In a high-prevalence population, even modest tests yield high PPV.

📉 Negative Predictive Value (NPV)

Formula: NPV = TN / (TN + FN)

• Answers: “Given a negative test, what is the probability the patient truly does not have the disease?”
• Also depends on prevalence — NPV rises when disease is rare.

⚖️ Likelihood Ratios (LR)

Formulae:
LR⁺ = Sensitivity / (1 − Specificity)
LR⁻ = (1 − Sensitivity) / Specificity

• Express how much a test result changes the probability of disease.
• LR⁺ > 10 → strong evidence to rule in disease.
  LR⁻ < 0.1 → strong evidence to rule out disease.
• Useful for Bayesian reasoning in clinical decision-making.

🧮 Example Table (2×2 Matrix)

🧠 Teaching Commentary

• Sensitivity and specificity are intrinsic properties of the test — they do not change with prevalence.
• PPV and NPV depend on disease prevalence — a low-prevalence condition produces many false positives.
• In acute medicine, start broad (sensitive test) and narrow down (specific test).
• Likelihood ratios integrate both metrics, helping clinicians interpret tests across populations rather than within one study.

📚 References

• Lalkhen AG, McCluskey A. Clinical tests: sensitivity and specificity. CEACCP 2008;8(6):221–223.
• BMJ Best Practice. Diagnostic test evaluation.
• Akobeng AK. Understanding diagnostic tests 1: sensitivity, specificity and predictive values. Arch Dis Child 2007.