Diagnostic Sensitivity, Specificity, and Likelihood Ratio Dependent on Prevalence of Disease-Underlying Cause
Author(s): Walter Fierz
In the process of medical diagnostics many types of tests are used, among them in vitro diagnostic laboratory tests. The performances of such tests are usually examined in clinical studies with a disease prevalence that is different from the prevalence of the disease in another clinical setting. The question then is whether diagnostic test characteristics like sensitivity, specificity and likelihood ratios are independent on the prevalence of the disease. The answer to this question is quite important when applying the test performance characteristics of a clinical study in a different clinical diagnostic setting. Here, it is demonstrated that, apart from special cases, test characteristics are indeed dependent on the prevalence of the disease. First, the underlying theoretical model of this dependence is demonstrated and, second, the model is validated with three practical diagnostic examples, i.e myocardial infarction, autoimmune disease, and vitamin-B12 deficiency.