A screening test for gestational diabetes mellitus (GDM) in Indian pregnant women is being validated. The test has a sensitivity of 88% and a specificity of 92%. All of the following statements about this test are true EXCEPT:

Explanation

## Likelihood Ratios and Clinical Utility **Key Point:** Likelihood ratios quantify how much a test result changes the probability of disease. A test with high sensitivity but lower specificity is better for ruling OUT disease (high NPV), not ruling IN disease. ### Calculating Likelihood Ratios Given: - Sensitivity = 88% - Specificity = 92% **Positive Likelihood Ratio (LR+)** = $\frac{\text{Sensitivity}}{1 - \text{Specificity}} = \frac{0.88}{1-0.92} = \frac{0.88}{0.08} = 11$ **Negative Likelihood Ratio (LR−)** = $\frac{1 - \text{Sensitivity}}{\text{Specificity}} = \frac{1-0.88}{0.92} = \frac{0.12}{0.92} ≈ 0.13$ ### Interpretation of Likelihood Ratios | LR+ Value | Clinical Interpretation | |-----------|------------------------| | > 10 | Strong evidence to rule IN disease | | 5–10 | Moderate evidence to rule IN disease | | 1–5 | Weak to moderate evidence to rule IN disease | | 1 | No diagnostic value | | 0.1–1 | Weak to moderate evidence to rule OUT disease | | < 0.1 | Strong evidence to rule OUT disease | **High-Yield:** - **LR+ = 11** → Strong evidence to RULE IN GDM (good for confirmation) - **LR− = 0.13** → Moderate evidence to RULE OUT GDM (good for exclusion) This test is actually **better for ruling IN disease** (LR+ = 11 is strong) than for ruling OUT (LR− = 0.13 is only moderate). Option 3 states the opposite. ### Verification of Other Options **Option 1:** False Negative Rate = 100% − Sensitivity = 100% − 88% = 12% - Out of 100 GDM patients, 12 will test negative ✓ (Correct) **Option 2:** False Positive Rate = 100% − Specificity = 100% − 92% = 8% - Out of 100 non-GDM individuals, 8 will test positive ✓ (Correct) **Option 4:** LR− = 0.12/0.92 ≈ 0.13 ✓ (Correct) **Clinical Pearl:** A test with LR+ > 10 and LR− > 0.1 is useful for both ruling in and ruling out. However, this test's LR+ (11) is much stronger than its LR− (0.13), making it primarily a confirmatory test, not an exclusionary one. **Mnemonic:** **SnNOut, SpPIn** - **Sn**Nsitive test, **N**egative result → rules **Out** disease - **Sp**ecific test, **P**ositive result → rules **In** disease This test has moderate sensitivity (88%) and good specificity (92%), so it's better for ruling IN (SpPIn) than ruling OUT.