Table 3. Mortality prediction based on the plasma levels of the biomarkers and on the severity scores according to ROC curve analysis.
| Prediction for mortality | CIRP level | APACHE II score | SOFA score | Lactate level | Serum creatinine | PCT level | WBC count | N% |
|---|---|---|---|---|---|---|---|---|
| Optimal cutoff-value | 1.49 ng/mL | 26.00 | 9.50 | 1.75 mmol/L | 124.10 mmol/L | 3.25 ng/mL | 15.83×109/L | 93.34% |
| Sensitivity, (%) (95%CI) | 96.77 (83.30–99.92) | 51.61 (33.06–69.85) | 51.61 (33.06–69.85) | 61.29 (42.19–78.15) | 51.61 (33.06–69.85) | 48.39 (30.15–66.94) | 45.16 (27.32–63.97) | 29.03 (14.22–48.04) |
| Specificity, (%) (95%CI) | 42.11 (26.31–59.18) | 86.84 (71.91–95.59) | 89.47 (75.20–97.06) | 76.32 (59.76–88.56) | 78.95 (62.68–90.45) | 63.16 (45.99–78.19) | 73.68 (56.90–86.60) | 81.58 (65.67–92.26) |
| Positive likelihood ratio | 1.67 | 3.92 | 4.90 | 2.59 | 2.45 | 1.31 | 1.72 | 1.58 |
| Negative likelihood ratio | 0.08 | 0.56 | 0.54 | 0.51 | 0.61 | 0.82 | 0.74 | 0.87 |
| Positive predictive value (%) | 57.69 | 76.19 | 80.00 | 67.86 | 66.67 | 51.72 | 58.33 | 56.25 |
| Negative predictive value (%) | 94.12 | 68.75 | 69.39 | 70.73 | 66.67 | 60.00 | 62.22 | 58.49 |
| AUC | 0.674 | 0.747 | 0.743 | 0.717 | 0.646 | 0.506 | 0.526 | 0.483 |
| p- value | 0.013 | 0.000 b | 0.001 | 0.002 | 0.039 | 0.938 | 0.708 | 0.814 |
The optimal cutoff values for each plasma biomarker level and the severity scores are presented. A p-value < 0.05 was considered to be statistically significant.
CIRP = cold-inducible RNA-binding protein, APACHE II = Acute Physiology and Chronic Health Evaluation II, SOFA = Sequential Organ Failure Assessment score, PCT = procalcitonin, WBC = white blood cell, N% = neutrophil ratio.
ROC = receiver operating characteristic, AUC = area under the ROC curve, CI = confidence interval.
b p = 0.000437.