QUICK-REFERENCE BOX — The Mnemonics That Actually Help
- SnNOut: a highly Snensitive test, when Negative, rules a disease Out.
- SpPIn: a highly Specific test, when Positive, rules a disease In.
- Sensitivity & specificity are properties of the test; predictive values depend on prevalence.
- Likelihood ratios are prevalence-independent — best for bedside probability updating.
- LR+ >10 or LR− <0.1 → large, often decisive change in probability.
1. Overview
The Core Idea
Diagnostic test performance is built from a 2×2 table comparing the test result against the true disease state. Sensitivity and specificity describe the test itself; predictive values tell you what a result means for this population and depend on prevalence; likelihood ratios combine both and are the most portable tool at the bedside.
2. The 2×2 Table
| Disease + | Disease − |
|---|
| Test + | True Positive (TP) | False Positive (FP) |
| Test − | False Negative (FN) | True Negative (TN) |
3. Sensitivity & Specificity
Definitions
- Sensitivity = TP / (TP + FN) — of those with disease, the fraction the test catches. High sensitivity, when negative, rules out (SnNOut).
- Specificity = TN / (TN + FP) — of those without disease, the fraction correctly cleared. High specificity, when positive, rules in (SpPIn).
- Both are (largely) properties of the test and do not change with disease prevalence.
4. Predictive Values
Definitions
- PPV = TP / (TP + FP) — given a positive test, the probability of disease.
- NPV = TN / (TN + FN) — given a negative test, the probability of no disease.
- Predictive values depend on prevalence: as prevalence rises, PPV rises and NPV falls. The same test performs differently in a screening vs a high-risk population.
5. Likelihood Ratios
The Bedside Tool
- LR+ = sensitivity / (1 − specificity) — how much a positive result raises the odds of disease.
- LR− = (1 − sensitivity) / specificity — how much a negative result lowers the odds.
- LRs are prevalence-independent. Pre-test odds × LR = post-test odds.
- Rules of thumb: LR+ >10 or LR− <0.1 causes large, often conclusive shifts; LRs near 1 are useless.
6. Cutoffs & the ROC Curve
The Trade-Off
- Moving the cutoff trades sensitivity against specificity — you cannot maximize both at once.
- The ROC curve plots sensitivity vs (1 − specificity) across all cutoffs.
- Area under the curve (AUC) summarizes overall discrimination: 0.5 = coin flip, 1.0 = perfect.
7. Applying It at the Bedside
How to Use It
- Estimate the pre-test probability from the clinical picture.
- Apply the test's likelihood ratio to get the post-test probability.
- Choose sensitive tests to screen/rule out; specific (or confirmatory) tests to rule in.
8. Key Pearls
High-Value Points
- SnNOut / SpPIn — the two mnemonics worth memorizing.
- Sensitivity/specificity = test properties; PPV/NPV = population-dependent.
- Likelihood ratios are prevalence-independent and best for updating probability.
- Lowering a cutoff raises sensitivity at the cost of specificity.
- AUC quantifies how well a test discriminates disease from non-disease.
9. Common Mistakes to Avoid
Misreadings & Better Practice
Frequent errors with diagnostic statistics.
| Mistake | Why it's wrong | Better reading |
|---|
| "PPV is a fixed property of the test." | It changes with prevalence. | Report PPV/NPV with the population in mind. |
| Using a sensitive test to confirm disease. | Sensitive tests rule out, not in. | Confirm with a specific test. |
| Ignoring pre-test probability. | Post-test probability depends on it. | Combine pre-test probability with the LR. |
10. Board-Style High-Yield Summary
Key Takeaways
- Sensitivity = TP/(TP+FN); Specificity = TN/(TN+FP); SnNOut / SpPIn.
- PPV and NPV depend on prevalence; sensitivity/specificity do not.
- LR+ = Sn/(1−Sp); LR− = (1−Sn)/Sp; prevalence-independent; LR+ >10 or LR− <0.1 is powerful.
- Cutoff choice trades sensitivity vs specificity; ROC/AUC summarize discrimination.
- Post-test probability = pre-test probability updated by the likelihood ratio.
11. References
- 1.Deeks JJ, Altman DG. Diagnostic tests 4: likelihood ratios. BMJ. 2004;329:168–169.
- 2.Akobeng AK. Understanding diagnostic tests 1: sensitivity, specificity and predictive values. Acta Paediatr. 2007.
- 3.Guyatt G, et al. Users' Guides to the Medical Literature.
Back to Clinical Guideline Hubs