Table 2.
Accuracy analysis using receiver operating characteristic (ROC) curves for quantitative variables: immunoglobulin G (IgG) index, kappa () index; and accuracy analysis with crosstabs for categorical variables (oligoclonal IgG bands (OCGBs) and Barkhof-Tintoré criteria (BTC)).
| index | OCGB | IgG index | BT≥3 | BTC+ index | BTC+ OCGB | |
|---|---|---|---|---|---|---|
| Sensitivity (95% CI) | 93.1 (77.2–99.2) | 96.5 (82.2–99.9) | 75.9 (56.5–89.7) | 72.4 (52.8–87.3) | 69.0 (49.2–84.7) | 72.4 (52.8–87.3) |
| Specificity (95% CI) | 95.7 (88.0–99.1) | 98.6 (92.3–99.9) | 94.3 (86–98.4) | 100 (94.9–100) | 100 (94.9–100) | 100 (94.9–100) |
| PPV (95% CI) | 90.0 (73.1–98.0) | 96.6 (82.2–99.9) | 84.6 (65.1–95.6) | 100 (83.9–100) | 100 (83.2–100) | 100 (83.9–100) |
| NPV (95% CI) | 97.1 (89.9–99.6) | 98.6 (92.3–99.9) | 90.4 (81.2–96.1) | 89.7 (80.8–95.5) | 88.6 (79.5–94.7) | 89.7 (80.8–95.5) |
| AUC | 0.971 | – | 0.886 | – | – | – |
| Cut-off | 10.62 | – | 0.56 | – | – | – |
AUC: area under the curve; CI: confidence interval. Results for combined test using BTC plus OCGB, and BTC plus index are shown.