Risk of bias assessment
The selected studies were heterogeneous, the overall risk of bias assessment for the included studies is displayed in Figure 2 . Within the patient selection domain, 43.6 % of studies showed high risk, 31.5% showed low risk and 24.8% were of unclear risk of bias. This originates from studies that did not randomly select, consecutively enrol participants, use a case-control design, or not report how exclusion was managed, highlighting the necessity to adhere to standardized procedures. For applicability, 96.0% of studies had low concerns that the included patients did not match the review question.
For the index test domain, 22.1% of studies showed high risk, 38.3% low risk and 36.9% unclear risk of bias. The most common reasons for assigning a high risk of bias were studies with a lack of blinding or failure to establish a threshold for the index tests before conducting the analysis. For applicability, 77.2% of studies had low concerns that the index test, its conduct, or interpretation differed from the review question.
For the reference standard domain, 18.1% of studies showed high risk of bias, 38.9% low risk and 43.0% unclear risk. An increase in the risk of bias was seen in studies which included prior information of the reference standard test while performing the index test. For applicability, 88.6% of studies had low concerns that the target condition as defined by the reference standard does not match the review question.
For the flow and timing domain, 61.1% of studies showed high risk, 23.5% low risk and 15.4% unclear risk of bias. We considered studies to be of low risk of bias if all participants received the same reference standard within six months of having received the index test.
Table 3 shows the risk of bias assessment summary for each domain question and Table S4 shows the individual risk of bias assessment per study.