Diagnostic test results don't exist in a vacuum - the population and setting directly impact how results should be interpreted to guide clinical care. Positive and negative predictive values highly depend on disease prevalence within the tested sample. A 99% negative predictive value converts to 80% when normalized to a just moderately higher prevalence of 50%. Clinicians must apply nuance rather than taking predictive values at face value, carefully considering the patients and scenarios where the test was studied. While diagnostic testing guides diagnosis and treatment, results require thoughtful interpretation within each unique clinical context. Blanket reliance on reported predictive values without considering prevalence risks misapplication. Instead, understanding how population factors impact results allows the artful practice of evidence-based medicine. Tests rule in and rule out disease to varying degrees based on context. Rather than just accepting predictive values as fixed, clinicians optimally normalize and adapt evidence to the patient in front of them.
Heston TF (2009). Detection of myocardial infarction by CT angiography. Heart 2009;95:1108. http://dx.doi.org.offcampus.lib.washington.edu/10.1136/hrt.2009.169508