132 chapter 2 Best-case and worst-case scenarios Hypothetical best-case and worst-case scenarios were constructed to overcome the 28.9% missing histopathology in nodules with a conclusive RAS mutation analysis, including 57 RAS-positive nodules and 767 RAS-negative nodules. In the best-case scenario, the estimated pooled sensitivity, specificity, positive and negative LR would be 33.6% (95% CI: 23.5%-45.5%), 96.5% (95% CI: 94.7%- 97.8%), 9.70 (95% CI: 5.21-18.06) and 0.69 (95% CI: 0.58-0.82), respectively. In the worst-case scenario, the estimated pooled sensitivity, specificity, positive and negative LR would be 14.1% (95% CI: 8.0%-23.5%), 90.9% (95% CI: 82.6%-95.5%), 1.55 (95% CI: 0.61-3.95) and 0.95 (95% CI: 0.84-1.07), respectively (Figure 4). Therefore, whatever the results of the nodules without histopathology would be, the true specificity of RAS mutation analysis is most likely high, while its sensitivity remains inadequate. Figure 19. SROC curves of HRAS, KRAS, and NRAS mutation analysis Summary receiver operating characteristic plots showing sensitivity versus 1-specificity of HRAS (a), KRAS (b), and NRAS (c) mutation analysis in indeterminate thyroid nodules with available histopathology. HRAS AUC = 0.45 (95% CI: 0.41-0.50). KRAS AUC 0.99 (95% CI: 0.98-1.00). NRAS AUC = 0.53 (95% CI: 0.480.57). AUC, area under the curve; HSROC, hierarchical summary receiver operating characteristic. a. b. c.
RkJQdWJsaXNoZXIy MTk4NDMw