Understanding Sensitivity and Specificity
Yes, you are correct: sensitivity is the proportion of individuals with disease who test positive (true-positive rate), and specificity is the proportion of individuals without disease who test negative (true-negative rate). 1, 2
Core Definitions
Sensitivity measures how well a test catches disease when it is actually present:
- Calculated as True Positives / (True Positives + False Negatives) 2
- Represents the percentage of diseased patients correctly identified by the test 1, 2
- Answers the question: "Of all the people who actually have the disease, what percentage will test positive?" 1
Specificity measures how well a test correctly identifies those without disease:
- Calculated as True Negatives / (True Negatives + False Positives) 2
- Represents the percentage of disease-free patients correctly labeled as negative 1, 2
- Answers the question: "Of all the people who do not have the disease, what percentage will test negative?" 1
The Inverse Relationship
Sensitivity and specificity are inversely related—when one increases, the other decreases: 1
- As you adjust the diagnostic cutoff to catch more true cases (increasing sensitivity), you inevitably catch more false positives (decreasing specificity) 1
- All diagnostic tests have a range of inversely related sensitivities and specificities that can be selected by choosing different diagnostic cutoff points 1
- For example, exercise ECG testing using 1mm ST-segment depression demonstrates sensitivity of 68% and specificity of 77% for coronary artery disease, but these values would shift if a different cutoff were chosen 1
Factors That Influence These Metrics
Sensitivity is influenced by: 1
- Disease severity (tests perform better with more severe disease) 1
- Patient effort level during testing 1
- Use of anti-ischemic drugs or other medications 1
Specificity can be affected by: 1
Critical Distinction from Predictive Values
Sensitivity and specificity are characteristics of the test itself, not measures of disease probability in individual patients: 2
- These metrics remain constant regardless of disease prevalence in the population 2
- In contrast, positive predictive value (PPV) and negative predictive value (NPV) tell you the probability that a patient actually has or doesn't have disease given their test result 2
- PPV and NPV vary dramatically with disease prevalence, while sensitivity and specificity do not 2