Table 3.
The results of ROC curves of training cohort and validation cohort in the Kazakhs.
| Variables | Cutoff | Sen (%) | Spe (%) | Youden's index | AUC (95% CI) | P value | |
|---|---|---|---|---|---|---|---|
| Model 1 | Lg(IL-6) | 1.703 | 47.79 | 67.09 | 0.149 | 0.576 (0.527-0.625) | P < 0.001 |
| Lg(APN) | 1.561 | 76.47 | 46.49 | 0.230 | 0.628 (0.582-0.675) | P < 0.001 | |
| Nomogram | -2.293 | 86.03 | 65.36 | 0.514 | 0.836 (0.802-0.869) | P < 0.001 | |
|
| |||||||
| Model 2 | Lg(IL-6) | 1.495 | 76.12 | 51.83 | 0.280 | 0.650 (0.582-0.719) | P < 0.001 |
| Lg(APN) | 1.470 | 73.13 | 53.90 | 0.270 | 0.663 (0.593-0.733) | P < 0.001 | |
| Nomogram | -2.307 | 88.06 | 68.58 | 0.567 | 0.812 (0.753-0.871) | P < 0.001 | |
Model 1, training cohort. Model 2, validation cohort. Sen: sensitivity; Spe: specificity.