Table 2.
ROC characteristics
| CTC | Cut-Off | Sensitivity | Specificity | Accuracy | AUC 1 |
|---|---|---|---|---|---|
| iCTC | −25% 2 | 79.7% | 36.4% | 62.0% | |
| J (−98.2%) 3 | 43.8% | 88.6% | 62.0% | 0.657 | |
| OfSn (−58.6%) 4 | 68.8% | 50.0% | 61.1% | ||
| aCTC | −25% 2 | 72.3% | 54.5% | 65.0% | |
| J (−50%) 3 | 70.8% | 64.6% | 68.3% | 0.651 | |
| OfSn (−9.8%) 4 | 73.9% | 54.6% | 66.0% | ||
| iCTC + aCTC | −25% 2 | 79.7% | 38.6% | 63.0% | |
| J (−66.7%) 3 | 64.1% | 63.6% | 63.9% | 0.686 | |
| OfSn (−50.3%) 4 | 71.9% | 50.0% | 63.0% |
1 AUC: area under the curve; 2 −25%: 25% reduction cut-off; 3 J: Youden J point with the percent reduction corresponding to that J point represented in the parentheses; 4 OfSn: cut-off optimized for sensitivity with the corresponding percent reduction represented in the parentheses.