Skip to main content
. 2015 Oct 26;36(1):1–8. doi: 10.3343/alm.2016.36.1.1

Table 2. Receiver operating characteristics curve analysis results for each biomarker in the discrimination of septic patients from non-septic patients*.

Tests AUC (95% CI) Best cutoff Sensitivity (95% CI) Specificity (95% CI) NPV PPV Accuracy (95% CI)
LA 0.781 (0.714-0.840) > 2.0 mmol/L 50.5% (35.4-54.3%) 92.3% (75.7-97.9%) 80.8% 98.1% 55.2% (45.7-62.6%)
PCT 0.923 (0.873-0.957) > 1.38 ng/mL 71.1% (56.8-74.3%) 100.0% (87.1-100.0%) 68.7% 100.0% 74.4% (64.5-79.4%)
CRP 0.940 (0.894-0.970) > 5.25 mg/dL 77.5% (74.3-88.5%) 97.9% (88.9-99.6%) 50.5% 99.4% 81.3% (81.0-91.5%)
IPF 0.868 (0.810-0.914) > 3.1% 89.8% (84.9-99.8%) 53.2% (39.2-66.7%) 46.8% 89.8% 83.2% (78.8-90.0%)
IG 0.812 (0.747-0.867) > 0.4% 67.0% (56.8-74.3%) 74.5% (60.5-84.7%) 67.0% 92.3% 68.3% (60.9-75.3%)
IRF 0.658 (0.584-0.727) > 5.6% 77.6% (71.8-87.9%) 56.0% (37.1-73.3%) 75.9% 93.3% 75.1% (67.1-82.6%)

*The sensitivity, specificity, negative predictive value, positive predictive value, and accuracy of each biomarker were calculated from the application of predefined best cutoff values for each biomarker.

Abbreviations: LA, lactate; PCT, procalcitonin; CRP, C-reactive protein; IPF, immature platelet fraction; IG, immature granulocyte fraction; IRF, immature reticulocyte fraction; AUC, area under the curve; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value.