Objective: The lack of a standard methodology in diagnostic research impedes adequate evaluation before implementation of constantly developing diagnostic techniques. We discuss the methodology of diagnostic research and underscore the relevance of decision analysis in the process of evaluation of diagnostic tests.
Study design and setting: Overview and conceptual discussion.
Results: Diagnostic research requires a stepwise approach comprising assessment of test characteristics followed by evaluation of added value, clinical outcome, and cost-effectiveness. These multiple goals are generally incompatible with a randomized design. Decision-analytic models provide an important alternative through integration of the best available evidence. Thus, critical assessment of clinical value and efficient use of resources can be achieved.
Conclusion: Decision-analytic models should be considered part of the standard methodology in diagnostic research. They can serve as a valid alternative to diagnostic randomized clinical trials (RCTs).