Hobbies And Interests
Home  >> Science & Nature >> Science

How to Calculate PPV and NPV

Epidemiologists and doctors characterize tests for a disease or condition by each test's specificity and predictive accuracy with respect to that disease or condition. The positive predictive value (PPV) of a test identifies the accuracy of a test in identifying test subjects with a disease or condition. The negative predictive value (NPV) of a test identifies the accuracy of a test in identifying test subjects without a disease or condition. In both cases, a number close to 1 is best.

Things You'll Need

  • Pencil
  • Paper
  • Calculator
Show More

Instructions

  1. Positive Predictive Value

    • 1

      Write the number of test subjects without the disease or condition who test positive for that disease or condition. To "test positive" means the test has indicated the test subject has the disease or condition. For example, write, "17."

    • 2

      Write the number of test subjects with the disease or condition who test positive for the disease or condition. For example, write, "34."

    • 3

      Add the two numbers you have written using a calculator. For example, 17 + 34 = 51.

    • 4

      Find the quotient of your answer and the number of patients with the disease or condition who test positive for the disease or condition. For example, 34/51 = 0.6666. The positive predictive value (PPV) of this test is approximately 0.667.

    Negative Predictive Value

    • 5

      Write the number of test subjects without the disease or condition who test negative for the disease or condition. To "test negative" means the test has indicated the test subject does not have the disease or condition. For example, write, "31."

    • 6

      Write the number of test subjects with the disease or condition who test negative for the disease or condition. For example, write, "72."

    • 7

      Find the sum of the two numbers you have written using a calculator. For example, 72 + 31 = 103.

    • 8

      Find the quotient of your answer and the number of test subjects without the disease or condition who test negative for the disease or condition. For example, 31/103 = 0.3009. The negative predictive value (NPV) of this test is approximately 0.301.


https://www.htfbw.com © Hobbies And Interests