98 chapter 2 Funnel plots regarding the publications on the BRAF (a, Egger’s test for funnel plot assymmetry p=0.60) and RAS (b, p=0.06) point mutations, RET/PTC (c, n.a. due to small number of observations) and PAX/PPARg (d, n.a. due to small number of observations) fusion, 7-gene mutation panels (e. p=0.12), Afirma® GEC (f, p=0.22), immunocytochemistry using Galectin-3 (g, p=0.94), HBME-1 (h, p=0.48), or CK-19 (i), suspicious ultrasound characteristics including solitary nodule (j, p=0.14), solid nodule (k, p=0.83), hypoechogeneity (l, p=0.15), ultrasound size ≥ 4 cm (m, p=0.73), taller-than-wide shape (n, p=0.58), absence of hypoechoic halo (o, p=0.56), irregular margins (p, p=0.80), increased intranodular vascularization (q, p=0.08), microcalcifications (r, p=0.34), and macrocalcifications (s, p=0.25), elastosonography (t, p=0.96), [99mTc]Tc-MIBI (u, n.a. due to small number of studies), and [18F]FDG-PET/CT (v, p=0.45). Main results Figure 4 provides an overview of the pooled test performance estimates of all the additional diagnostic procedures under investigation in this meta-analysis. Besides estimated pooled test performance results for all indeterminate thyroid nodules with an available histopathological correlation, we report a best-case and worst-case scenario for each index test. These scenarios take account for the varying rates of missing histology and also include the thyroid nodules from all studies that had a reported index test result but no result for the reference test. Note that some studies – particularly retrospective ones – did not report these numbers or not separately for the indeterminate cytology categories, and only reported results for the included nodules that had both index test and reference standard available [69, 76, 107, 109, 116, 119, 202, 204, 217-219, 242, 260, 268, 352, 353]. This affected the estimations for a best-case and worst-case scenario. The results of the various diagnostics are exemplified in the following chapters.
RkJQdWJsaXNoZXIy MTk4NDMw