Npv

[taylor6902]

---

Members don't see this ad.

If prevalence increases, what happens to NPV?

Also, if prevalence decreases, what happens to NPV?

 

[PBandJ]

NPV and prevalence follow the opposite trend as PPV and prevalence.
Increase in prevalence means a decrease in NPV
Decrease in prevalence means an increase in NPV.

 

[TMP-SMX]

Think of it this way. If a lot of people in the population have the disease and the test shows up positive, your PPV is higher just because there is a higher risk overall to begin with. If you screen for diabetes in a population in the mall your PPV will be lower than if you screen a middle aged overweight population with symptoms of diabetes. This is why we don't usually test unless there is a reason to do so because of false positives.

NPV is the opposite. It is inversely proportional to prevalence. If you increase prevalence you decrease NPV.

PPV is directly proportional to prevalence. If you increase prevalence PPV increases.

Draw a 2X2 table if this doesn't make sense. The number of people in the disease column vs. the no disease column (prevalence) will heavily change the NPV and PPV while having no bearing on sensitivity and specificity. You can be 90% sensitive and specific but still have issues with the predictive value of a test depending how common the disease is.

 

[taylor6902]

thanks for the replies