Assessing the clinical utility of diagnostics used in drug therapy

Abstract

There is an ongoing debate over the evidentiary standards that should be applied for introduction of new diagnostics into routine clinical practice. Many call for evidence of "clinical utility," i.e., a positive impact on patient outcomes. A diagnostic, when used with a medicine, has clinical utility if it improves the outcomes of drug therapy. Improved outcomes may be defined broadly, including benefits, harm reduction, and patient-reported outcomes. Much of the controversy centers around the methods of demonstrating clinical utility. For instance, are randomized prospective trials the only acceptable source of data? Practically speaking, many sources of evidence-mechanistic, pharmacologic, and observational-can contribute to a finding of clinical utility, depending on the circumstances. Clinical utility is highly indication specific, and achieving it is dependent on good analytical and diagnostic test performance.