Interpretation of uncertainty reduction
The amount of reduction of uncertainty defines the usefulness of a diagnostic test. Ideally, it would be defined in decision analytic context when the “useful” test is the one that affect our decisions and its downstream consequences. This, however, require case-specific decision modeling, which is not the focus of this paper. Alternatively, usefulness of a test can be defined according to magnitude of reduction of diagnostic uncertainty analogously to treatment effects as small, moderate or large19. Thus, we define small reduction of diagnostic uncertainty if it is less than 10%, moderate reduction between 20-30% and large reduction of diagnostic uncertainty if it exceeds 40-50%.