Table 7.
Parameters of unconditional logistic regression model for differentiating different disease status of rheumatoid arthritis (RA) in cohort 3.
 |
Discovery set | External validation set | ||||
|---|---|---|---|---|---|---|
| AUC | Sensitivity (%) | Specificity (%) | 95% confidence interval | Cut-off value | Predictive value (%) | |
| Moderate to high disease activity vs. Remission to low disease activity | 0.935 | 95.16 | 77.83 | 0.911–0.960 | 0.29 | 92.66 |
| Low disease activity vs. Remission | 0.873 | 90.54 | 70 | 0.811–0.936 | 0.58 | 93.55 |
| Remission vs. Healthy subjects | 0.905 | 70.88 | 98.65 | 0.871–0.940 | 0.84 | 96.20 |
| Moderate to high disease activity vs. Low disease activity | 0.872 | 92 | 75.11 | 0.830–0.915 | 0.23 | 88.89 |
| Low to high disease activity vs. Remission | 0.959 | 97.30 | 85.24 | 0.940–0.977 | 0.16 | 91.74 |