Effect of disease prevalence on predictive values

The Diagnostic 2x2 - Test Measures & Truths

• Intrinsic Test Properties (Stable)

  • Sensitivity & Specificity are fixed characteristics of a diagnostic test. They are NOT affected by how common the disease is.

• Prevalence-Dependent Values (Variable)

  • PPV & NPV are heavily influenced by the pre-test probability, which is the prevalence of the disease in the population being tested.

• Effect of Prevalence on Predictive Values

  • If Prevalence ↑ (e.g., testing in a high-risk group):
    • PPV ↑: A positive result is more likely to be a true positive.
    • NPV ↓: A negative result has a slightly higher chance of being a false negative.
  • If Prevalence ↓ (e.g., screening the general population for a rare disease):
    • PPV ↓: A positive result is more likely to be a false positive.
    • NPV ↑: A negative result is very likely to be a true negative.
  • 📌 Prevalence & PPV are Positively correlated.

⭐ A screening test for a rare disease (low prevalence) will have a low PPV, even with high sensitivity and specificity. This means a positive result is more likely to be a false positive than a true positive.

Prevalence Power - The Predictive Value Swing

• Core Principle: While Sensitivity and Specificity are fixed, intrinsic characteristics of a diagnostic test, its real-world utility, measured by predictive values (PPV & NPV), hinges on the prevalence of the disease in the tested population. Prevalence is the most important determinant of a test's predictive value.

• Sensitivity & Specificity vs. Predictive Values:

  • Unaffected by Prevalence: Sensitivity, Specificity.
  • Affected by Prevalence: PPV, NPV.

• Positive Predictive Value (PPV):

  • Relationship: Directly proportional to prevalence.
  • As prevalence ↑, PPV ↑. A positive result is more trustworthy.
  • As prevalence ↓, PPV ↓. A positive result is more likely to be a false positive.
  • 📌 Mnemonic: Prevalence Positively predicts PPV.

• Negative Predictive Value (NPV):

  • Relationship: Inversely proportional to prevalence.
  • As prevalence ↑, NPV ↓.
  • As prevalence ↓, NPV ↑. A negative result is more reassuring.

⭐ High-Yield Pearl: Screening tests for rare diseases in the general population (low prevalence) will have a low PPV, even if the test is highly sensitive and specific. This is why confirmatory tests are crucial and screening is often targeted at high-risk (high prevalence) cohorts to increase the PPV.

• Clinical Bottom Line: The answer to "How good is this test?" depends entirely on "Who are you testing?". A test with 95% sensitivity/specificity is powerful in a specialty clinic (high prevalence) but can be misleadingly alarming in a general screening setting (low prevalence) due to a low PPV.

• Sensitivity and specificity are intrinsic to the test and do not change with disease prevalence.
• Positive Predictive Value (PPV) is directly proportional to prevalence; as prevalence ↑, PPV ↑.
• Negative Predictive Value (NPV) is inversely proportional to prevalence; as prevalence ↑, NPV ↓.
• In high-prevalence settings, a positive test is likely a true positive.
• In low-prevalence settings, a positive test is more likely to be a false positive.