## Discriminating Feature: Specificity and Positive Predictive Value **Key Point:** When a test is **positive**, the ability to confirm that disease is truly present depends on **specificity**, not sensitivity. Test B's 95% specificity ensures fewer false positives, yielding a higher PPV — the probability that a positive test result reflects true disease. ### Why Specificity Matters for Confirmation **High specificity** = few false positives = when test is positive, disease is likely present. **High sensitivity** = few false negatives = when test is negative, disease is likely absent. The question asks about identifying women **when the test is positive** — this is the domain of specificity and PPV. ### Comparative Analysis | Property | Test A (92% Sen, 78% Spec) | Test B (75% Sen, 95% Spec) | |----------|---------------------------|---------------------------| | **Sensitivity** | 92% (fewer false negatives) | 75% | | **Specificity** | 78% (more false positives) | 95% (fewer false positives) | | **When test is POSITIVE** | ~22% chance of false positive | ~5% chance of false positive | | **PPV (at 15% prevalence)** | ~52% | ~79% | | **Clinical role** | Screening (exclude disease) | Confirmation (confirm disease) | ### Mnemonic: **SpPin** — **Sp**ecificity rules **In** When you want to **confirm** disease (rule in), choose the test with high **Sp**ecificity. Test B's 95% specificity makes it the superior confirmatory test. **High-Yield:** In a positive test scenario, specificity is the discriminating property. Test B is superior because: 1. Fewer false positives (95% specificity vs 78%) 2. Higher PPV across all prevalence levels 3. More reliable for confirming cervical cancer diagnosis **Clinical Pearl:** In cervical cancer screening, a positive HPV test must be confirmed by cytology or colposcopy. A high-specificity test (Test B) reduces unnecessary referrals and anxiety from false positives, making it the preferred confirmatory test.
Sign up free to access AI-powered MCQ practice with detailed explanations and adaptive learning.