Introduction/Objective: Before the introduction of a new diagnostic technology in clinical practice its effectiveness should be assessed. Unlike the evaluation of new pharmaceutical compounds, there is no consensus on how such an evaluation should proceed.

Methods: Systematic review of literature. In the Medline database articles describing strategies for the evaluation of diagnostic technologies were retrieved.

Results: Twenty-five articles were retrieved. Six of these articles described a strategy for the evaluation of a test, from its development until its introduction into clinical practice. The different phases of test evaluation were very similar. These revealed an analogy to the four phases of pharmaceutical treatment assessment. A synthesis of these six strategies into these four phases results in: I. Test development, II. Diagnostic accuracy, III. Clinical impact, IV. Societal efficacy. Phase III is subdivided by some authors into (1) impact on diagnosis, (2) impact on therapy and (3) patient outcome.

Discussion: This strategy in four phases has some limitations. First, it assumes the existence of a reference test for the test under examination. Such a test may not always be available. Second, it requires that the test is used to obtain a diagnosis. The latter precludes the evaluation of diagnostic tests used for other purposes such as prognosis or monitoring of therapy. Current strategies of evaluation of diagnostic technologies are thus limited in applicability. Elements for an alternative evaluation strategy will be presented.