Understanding Negative Predictive Value

When ordering a medical laboratory test, understanding the test metrics is imperative. I know I can always use a little refresher when it comes to sensitivity/specificity and the less commonly understood negative and positive predictive value. So here is my simplified review of these metrics.

Negative predictive value is the probability that a patient does not have a disease when the corresponding test is negative.

Positive predictive value is the probability that a patient has a disease when the corresponding test is positive.

Important points:

1. If a test is a diagnostic test, then specificity and positive predictive value are most important.
2. If a test is a screening/rule out/triage test, then sensitivity and negative predictive value are most important.
3. Both negative and positive predictive value account for disease prevalence.

Finally easy way to remember: SPIN and SNOUT - Specificity/PPV rules in, Senstivity/NPV rules out.