Table 4.
Univariate logistic regression analyses of IL-6 signalling components as predictors of COVID-19 mortality based on proposed scores.
| Variable | Score | Risk factor cut-off characterization | Cut-off univariate logistic | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Cut-off | Sensitivity % | Specificity % | PPV % | NPV % | Accuracy % | OR | 95% CI | ||
| IL-6 | – | 27.40 | 72.97 | 58.82 | 79.41 | 50 | 68.52 | 3.86 | (1.64-9.06)** |
| IL-6+FME | 0-8 | 6 | 52.31 | 92.11 | 91.90 | 53.03 | 66.99 | 12.80 | (3.57-45.82)*** |
| IL-6+Lymphopenia+Creatinine | 0-16 | 8 | 70.31 | 73.68 | 81.82 | 59.57 | 71.57 | 6.63 | (2.70-13.30)*** |
| IL-6+FME+Lymphopenia+Creatinine | 0-22 | 8 | 85.25 | 72.22 | 83.88 | 74.29 | 80.41 | 15.02 | (5.44-41.500)*** |
| B/T+Lymphopenia+Creatinine | 0-18 | 9 | 51.56 | 97.30 | 97.06 | 53.73 | 68.32 | 38.32 | (4.95-166.67)*** |
NPV, Negative predictive value; PPV, Positive predictive value; B/T, Binary/ternary complex ratio; FME, Fold molar excess of sgp130 over sIL-6R; OR, Odds ratio. **p < 0.01; ***p < 0.001.