A diagnostic test for COVID-19 has a sensitivity of 85% and a specificity of 95%. Which statement correctly describes the relationship between these two parameters?

Explanation

## Sensitivity–Specificity Trade-off **Key Point:** Sensitivity and specificity are **test characteristics** that are inversely related. Changing the diagnostic cutoff threshold to improve one typically worsens the other. ### The Trade-off Mechanism When you lower the diagnostic threshold (e.g., lower the cutoff CT value for COVID-19 PCR): - **Sensitivity increases** — more true positives are caught, but also more false positives - **Specificity decreases** — more false positives occur Conversely, raising the threshold: - **Sensitivity decreases** — some true positives are missed - **Specificity increases** — fewer false positives ### Visual Representation: ROC Curve The **Receiver Operating Characteristic (ROC) curve** plots sensitivity (y-axis) against (1 − specificity) (x-axis). The curve demonstrates this trade-off: as you move along the curve, improving sensitivity comes at the cost of declining specificity. **High-Yield:** The **Area Under the Curve (AUC)** summarizes overall test accuracy: - AUC = 1.0 → Perfect test - AUC = 0.5 → No better than chance - AUC > 0.8 → Good discrimination ### Clinical Application | Clinical Scenario | Preferred Metric | Rationale | |-------------------|------------------|----------| | **Screening** (e.g., TB, cancer) | **High Sensitivity** | Minimize false negatives; catch all potential cases | | **Confirmation** (e.g., confirmatory COVID test) | **High Specificity** | Minimize false positives; avoid unnecessary treatment | | **Serious treatable disease** | **High Sensitivity** | Cost of missing disease > cost of false positives | | **Serious untreatable disease** | **High Specificity** | Cost of false positive diagnosis is high | **Mnemonic:** - **SNOUT** — **S**ensitivity rules **OUT** disease (use for screening) - **SPIN** — **SP**ecificity rules **IN** disease (use for confirmation) **Clinical Pearl:** A test with 85% sensitivity and 95% specificity is appropriate for initial screening (good sensitivity to catch cases) but should be confirmed with a more specific test before diagnosis.