Assessing risk of bias in comparative diagnostic accuracy studies using QUADAS-C
Bada Yang, Cochrane Netherlands, Department of Epidemiology and Health Economics, UMC Utrecht
Background: Diagnostic Test Accuracy reviews addressing comparative questions may include studies directly comparing index tests in the same study group. Because these comparative accuracy studies aim to estimate the difference in accuracy between tests (rather than solely the accuracy of each test), we need to consider additional sources of bias specific to test comparisons. The QUADAS-C tool, an extension to QUADAS-2, considers these additional sources of bias and is endorsed by Cochrane for assessing risk of bias in comparative accuracy studies.
Objectives: To learn how to assess risk of bias in comparative accuracy studies using QUADAS-C.
Description: First, we will briefly discuss possible study designs for comparing the accuracy of two tests (10 min). We will then brainstorm, in small groups, which shortcomings could be present across the domains Patient Selection, Index Test, Reference Standard and Flow and Timing (20 min). Subsequently, the structure of and items in QUADAS-C will be presented (10 min). Lastly, we will evaluate a comparative accuracy study in small groups using QUADAS-C (30 min), followed by a plenary discussion of any discrepancies between assessors (20 min). Importance to patients: Although this concerns a methodological topic and does not contribute directly to patient health, using biased research may lead to misleading recommendations about diagnostic tests.