179 Diagnostic utility of molecular and imaging biomarkers 2 Table 37. Estimated pooled test performance of Afirma® GEC in Bethesda IV indeterminate thyroid nodules reference nH Bethesda IV TP FP FN TN malignant (TP+FN/ni) sensitivity (95% CI), % specificity (95% CI), % LR+ (95% CI) LR- (95% CI) Alexander 2012 81 18 31 2 30 25% 90 (68-99) 49 (36-62) 1.77 (1.33-2.36) 0.20 (0.05-0.78) Harrell 2014 9 3 5 1 0 44.4% 75 (19-99) 0 (0-52) 0.76 (0.41-1.42) 3.60 (0.18-70.34) Lastra 2014 32 11 19 0 2 34.4% 100 (72-100) 10 (1-30) 1.08 (0.89-1.31) 0.37 (0.02-7.03) McIver 2014 31 4 23 1 3 16.1% 80 (28-99) 12 (2-30) 0.90 (0.57-1.43) 1.73 (0.22-13.48) Brauner 2015 26 4 21 0 1 15.4% 100 (40-100) 5 (0-23) 0.96 (0.71-1.31) 1.53 (0.07-32.41) Celik 2015 14 6 6 0 2 42.9% 100 (54-100) 25 (3-65) 1.29 (0.82-2.03) 0.26 (0.01-4.54) Witt 2016 5 3 2 0 0 60.0% 100 (29-100) 0 (0-84) 1.05 (0.56-1.97) 0.75 (0.02-28.08) Yang 2016 12 6 3 0 3 50.0% 100 (54-100) 50 (12-88) 1.86 (0.86-4.01) 0.14 (0.01-2.28) POOLED RESULTS 210 55 110 4 41 28.1% 93.6 (82.0-97.9) 16.2 (6.8-34.0) 1.12 (0.94-1.33) 0.39 (0.10-1.49) FN: false negative. FP: false positive. LR+: positive likelihood ratio. LR-: negative likelihood ratio. nH: number of indeterminate thyroid nodules with conclusive index test and histopathological follow-up. TN: true negative. TP: true positive.
RkJQdWJsaXNoZXIy MTk4NDMw