Lisanne de Koster

182 chapter 2 Table 38. Hypothetical best-case scenario estimated pooled test performance of Afirma® GEC in indeterminate thyroid nodules reference ni TP FP FN TN sensitivity (95% CI), % specificity (95% CI), % LR+ (95% CI) LR- (95% CI) Alexander 2012 210 46 77 5 82 90 (79-97) 52 (44-60) 1.86 (1.55-2.24) 0.19 (0.08-0.44) Alexander 2014 326 73 66 1 169 99 (93-100) 72 (66-78) 3.51 (2.86-4.32) 0.02 (0.00-0.13) Harrell 2014 58 23 13 2 18 92 (74-99) 58 (39-75) 2.19 (1.43-3.37) 0.14 (0.04-0.54) Lastra 2014 132 36 26 0 70 100 (90-100) 73 (63-81) 3.61 (2.60-5.01) 0.02 (0.00-0.29) McIver 2014 72 17 27 1 15 94 (73-100) 36 (22-52) 1.47 (1.14-1.89) 0.16 (0.02-1.09) Sullivan 2014 14 3 4 0 6 100 (29-100) 60 (26-88) 2.14 (0.96-4.77) 0.21 (0.02-2.96) Brauner 2015 72 8 37 0 26 100 (63-100) 41 (29-54) 1.61 (1.24-2.09) 0.13 (0.01-2.01) Celik 2015 40 15 8 0 11 100 (78-100) 58 (33-80) 2.28 (1.36-3.82) 0.05 (0.00-0.85) Marti 2015 165 68 36 0 61 100 (95-100) 63 (52-72) 2.67 (2.06-3.45) 0.01 (0.00-0.18) Noureldine 2015 222 98 99 2 16 98 (93-100) 14 (8-22) 1.14 (1.05-1.23) 0.14 (0.03-0.61) Witt 2016 32 6 9 0 14 100 (54-100) 61 (39-80) 2.35 (1.37-4.01) 0.12 (0.01-1.74) Wu 2016 214 76 31 2 88 97 (91-100) 74 (65-82) 3.74 (2.76-5.07) 0.03 (0.01-0.14) Yang 2016 189 64 30 0 80 100 (94-100) 73 (63-81) 3.61 (2.67-4.89) 0.01 (0.00-0.17) POOLED RESULTS 1,746 533 463 13 656 98.2 (95.5-99.3) 56.5 (45.4-67.0) 2.26 (1.75-2.91) 0.03 (0.01-0.09) FN: false negative. FP: false positive. LR+: positive likelihood ratio. LR-: negative likelihood ratio. ni : number of included indeterminate nodules. TN: true negative. TP: true positive.

RkJQdWJsaXNoZXIy MTk4NDMw