Fig. 2

Summary of simulation results. Bias in estimates of the contrasts in diagnostic accuracy from the proposed meta-analytical models applied in the simulation study. The boxplots present the bias in \(\hat {S}_{D12}\) and \(\hat {C}_{D12}\) (first row), \(\hat {S}_{RR12}\) and \(\hat {C}_{RR12}\) (second row), \(\hat {S}_{OR12}\) and \(\hat {C}_{OR12}\) (third row). The first column presents Scenario 1 where a common imperfect reference standard with moderate S and high C was used, the second scenario 2 where systematic bias is induced by differing reference standards. Full explanation of the model is in the text; full explanation of the simulation setup and results in Additional file 3. Note for Scenario 1: For models 1 to 3 disease status was estimated from the results of T₄. Note for Scenario 2: Models 4 and 5 were applied both assuming it is known that the reference tests differ across studies (4a and 5a) and ignoring the difference in reference tests (4b and 5b)