Table 2.
Receiver operating characteristic curve analysis to determine the optimum cutoff value of the PCT (ng/ml) for the prediction of BSI.
| AUROC (95% CI) | p | Cutoff | Sensitivity | Specificity | PPV | NPV | ||
|---|---|---|---|---|---|---|---|---|
| BSI | PCTmax | 0.68 (0.65, 0.70) | <0.001 | 8.24 | 0.60 | 0.66 | 0.15 | 0.82 |
| PCTgap | 0.67 (0.65, 0.69) | <0.001 | 8.10 | 0.57 | 0.67 | 0.25 | 0.89 | |
| Bacteremia (Gram–) | PCTmax | 0.73 (0.70, 0.76) | <0.001 | 8.06 | 0.72 | 0.64 | 0.15 | 0.96 |
| PCTgap | 0.72 (0.69, 0.75) | <0.001 | 8.08 | 0.69 | 0.66 | 0.16 | 0.96 | |
| Fungemia | PCTmax | 0.74 (0.68, 0.81) | <0.001 | 12.70 | 0.72 | 0.69 | 0.04 | 0.99 |
| PCTgap | 0.74 (0.68, 0.81) | <0.001 | 8.13 | 0.76 | 0.64 | 0.03 | 0.99 |
Gram–, Gram stain negative; PCT, procalcitonin; PPV, positive predictive value, NPV, negative predictive value; BSI, bloodstream infection; AUROC, area under the ROC curve.