Positive Predictive Value vs Negative Predictive Value

Positive predictive value (PPV) is the probability that a person with a positive test truly has the disease; negative predictive value (NPV) is the probability that a person with a negative test truly does not. Unlike sensitivity and specificity, they depend strongly on how common the disease is in the tested population.

Definitions

• PPV = TP / (TP + FP)
• NPV = TN / (TN + FN)

Illustrative example with worked numbers

Invented for teaching. A test has sensitivity 90% and specificity 90%. In 1,000 people with 10% prevalence (100 with disease): TP = 90, FN = 10; of 900 without disease, TN = 810, FP = 90. PPV = 90 / (90 + 90) = 50%; NPV = 810 / (810 + 10) = 98.8%. If prevalence falls to 1% (10 with disease, 990 without): TP = 9, FN = 1, TN = 891, FP = 99, so PPV = 9/108 = 8.3% and NPV = 891/892 = 99.9%.

Why prevalence matters

When a disease is rare, even a good test produces many false positives relative to true positives, so PPV falls. When the disease is common, NPV falls. PPV and NPV from one study apply only to populations with similar prevalence.

Using them clinically

• Estimate the pre-test probability for your patient.
• Use sensitivity, specificity and likelihood ratios to update it (Bayes' theorem), or use predictive values that match the prevalence in your setting.
• Consider the consequences of false positives and false negatives.

Reporting

Diagnostic accuracy studies should state the prevalence in their sample and, ideally, show predictive values at several plausible prevalences.

Common pitfalls

• Quoting PPV from a high-prevalence study to a screening population.
• Confusing PPV with sensitivity.

Key takeaways

• Predictive values depend on prevalence; sensitivity and specificity do not (in principle).
• Always consider the setting.

This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.