Professional Documents
Culture Documents
Predictive Value of Diagnostic Tests
Predictive Value of Diagnostic Tests
Sensitivity is the proportion of patients with disease who test positive. Sensitivity = True Positives/True Positives PLUS False Negatives
Specificity is the proportion of patients without disease who test negative. Specificity = True Negatives/True negatives Plus False Positives
Positive Predictive Value is the probability the person with a positive test result actually has the disease being tested for. PPV = True
Positives/True Positives +False Positives
Negative Predictive Value is the probability the person with a negative test result does not have the disease being tested for. NPV = True
Negative/False Negative +True Negative
Sample 2X2 table. Test sensitivity is 98%. Test specificity is 95%. Prevalence of disease is 1 in 100. 10,000 people are screened with the test.
disease no disease
test + 98 492
test - 2 9408
The positive predictive value (PPV) is 98/(98+492) = 16.6%. Even though the screening test has a sensitivity of 98%, a patient with a positive
test result in this low prevalence population has only a 16.6% chance of having the disease.
Some commonly used screening tests have poorer test characteristics than this example. For example, PSA for prostate cancer screening has the
following test characteristics, depending upon the cutoff for a positive test: