81 Diagnostic utility of molecular and imaging biomarkers 2 Quality assessment Two of the authors (DV and EJK) independently appraised quality of the selected articles using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) criteria. This 7-item tool assesses risk of bias and concerns about applicability regarding four domains: patient selection, index testing, reference standard, and flow and timing. The tool allows for a consistent and reliable assessment of the quality of diagnostic accuracy studies included in a systematic review [346]. Signalling questions used to assist the evaluation of the four domains were modified to fit our review topic, as appropriate (Table 2)[346]. Disagreements between the investigators were resolved by consensus after re-evaluation of the references. A graphical overview of the QUADAS-2 assessment is provided. As appropriate, two questions that were less relevant to our systematic review were omitted from the applied modified QUADAS-2 tool [346]: - Domain 1: “Was a case control design avoided?” Case control studies were already excluded during full text screening. - Domain 4: “Was there an appropriate interval between index test(s) and reference standard?” Due to the indolent nature of thyroid carcinoma, the precise interval between index test and reference standard is not deemed the most relevant to assess the quality of the included studies. Data extraction Summarized patient data and data on true positives (TP), false positives (FP), false negatives (FN) and true negatives (TN) were extracted from included articles by both reviewers. If a study assessed multiple diagnostic procedures, data were extracted into separate contingency tables for each of the investigated index tests. Additionally, if a study primarily investigated a combination of genetic alterations or a gene mutation panel (GMP) but results for the individual mutations were reported, too, these separate data were also extracted. Where possible and appropriate, data for Bethesda III and Bethesda IV nodules were also separately extracted, including cytology defined in accordance with one of these two categories. For statistical reasons, studies were excluded from meta-analysis when there were no benign (TN+FP=0) or no malignant (TP+FN=0) histopathology results. Data presented can differ from published results: data of individual thyroid nodules with other than indeterminate cytology were not considered, and only data of nodules using histopathology as the reference test were included in the present meta-analyses. Moreover, in several studies we reclassified the reported histopathology to homogenize definitions for two possible reasons. Firstly, papillary microcarcinoma (mPTC) was reported by multiple studies. In accordance with definitions
RkJQdWJsaXNoZXIy MTk4NDMw