The positive predictive value (PPV) tells you how likely it is that you actually have a disease if you test positive for it. It is defined as the number of true positives (people who test positive who have the disease) divided by the total number of people who test positive. It varies with test sensitivity, test specificity, and disease prevalence.

You can see how these aspects change a test's PPV in the example below. PPV is a tough thing for doctors to estimate, which can be hard on patients. However, it depends very highly on disease prevalence in the communities where they work. Knowing the sensitivity and specificity isn't enough to tell what the PPV is. You also have to have a good idea of how common the disease is that you're testing for.

### Examples

Imagine that a chlamydia test has 80% sensitivity & 80% specificity. In a population of 100 with a chlamydia prevalence of 10%:

8/10 true positives will test positive

72/90 true negatives will test negative

Out of 26 positive tests, 8 are true positive. 18 are false positive. Therefore the positive predictive value (PPV) would be 31% (8/26). Only one third of people testing positive would actually have chlamydia.

On the other hand, if the prevalence of chlamydia was 30%:

24/30 true positives would test positive

56/70 true negatives would test negative.

In this situation, the PPV would be 24/38=63%. Two thirds of the population testing positive would have an accurate test result.

What about a test that is 80% sens. and 95% spec. in the 20% population?

16/20 true + will test +

76/80 true - will test -

And the PPV would be 16/20=80%

### Factors That Increase PPV

Increasing the percentage of true positives and the test PPV require one or both of two things. The test could have a high specificity. Most people who didn't have chlamydia would test negative. Then, there would be very few false positives. The other thing that increases test PPV is high prevalence.

The greater percentage of people who are infected, the higher the PPV.

In most cases, everyone would prefer just increasing test specificity. Having more people be sick just to improve the performance of a diagnostic test is somewhat counter-intuitive.