Methodological challenges and problems associated with the evaluation of diagnostic tests were identified within a narrative review and the results were described according to an evidence hierarchy model developed by Fryback and Thornbury. Methodological guidelines of selected institutions (MSAC, IQWiG, EUnetHTA, NICE) were used to elicit similarities and differences in the evaluation of diagnostics between these institutions. These findings were also presented based on the hierarchy model. Finally, a checklist was developed as an aid for decision-makers for structuring questions on benefits, harms and costs of a test in order to facilitate the assessment of diagnostic technologies.

For the evaluation of diagnostic test from a decision-makers perspective, the assessment of patient-relevant outcomes is of utmost importance. In order to assess these patient-relevant outcomes, not only consequences of the test itself but also of any subsequent therapies have to be taken into account. This can be done directly, if studies are available which compare therapeutic consequences of the new test with the therapeutic consequences of the reference test. As these studies are rare, „linked evidence“ offers an alternative by linking studies on diagnostic accuracy to studies on therapeutic effectiveness. However, many methodological challenges are associated with the assessment of the quality of studies on diagnostic accuracy and several requirements have to be met for „linked evidence“. For example, unique methodological challenges of studies on diagnostic accuracy necessitate specific tools for judging the study’s quality. Moreover, the study population of the diagnostic accuracy study and the therapeutic efficacy study have to be comparable and a commonly accepted reference standard has to exist by which these studies can be linked.

The investigated institutions based their evaluation of diagnostics primarily on benefits arising for patients. Their preferred method for assessing patient relevant outcomes is a systematic review and three of the institutions mention “linked evidence” as a possibility to indirectly establish these outcomes. Instruments for judging the quality of studies on diagnostic accuracy are QUADAS, a checklist developed by the Cochrane Organisation and the STARD Statement. Three institutions additionally consider the costs in relation to benefits when assessing tests.