How to evaluate a diagnostic marker test. Lessons from the rise and fall of dexamethasone suppression test

To understand why the dexamethasone suppression test (DST) for the diagnosis of depression became widely accepted and later rejected, we reviewed the sequence of publications in the DST literature. To evaluate the events, we developed and applied concepts of a five-phase process that can be used to assess the clinical utility of diagnostic marker tests. The review showed that when the DST was introduced into the clinical arena, the initial and final two phases of testing (I, IV, and V) had not been adequately conducted. When these phases of testing were suitably checked many years later, the Phase I studies (exploring basic mechanics of test procedures) showed that dexamethasone had variable bioavailability and that the cortisol assay procedure was unreliable. The Phase IV and V studies (examining test results in groups with suitably broad spectrums of cases and controls) showed that the test did not differentiate depression from most pertinent comorbid conditions. Beyond application to the specific problems of the DST, the proposed five phases of development and evaluation for diagnostic marker tests can be used to plan suitable research and avoid similar problems in the future.