Skip to main content
. 2021 Feb 5;16(2):e0246027. doi: 10.1371/journal.pone.0246027

Table 2. Accuracy and cut points of individual biomarkers and combinations of biomarkers for diagnosis of SBI on admission to PICU.

Individual biomarker or biomarker combinations AUC (95% CI) Cut-point# Sensitivity 95% CIs Specificity 95% CIs PPV 95% CIs NPV 95% CIs LR+ 95% CIs LR- 95% CIs
PCT (ng/ml) * 83.73 (78.11, 89.35) 1.25 0.78 (0.69, 0.88) 0.80 (0.76, 0.84) 0.37 (0.26, 0.48) 0.96 (0.94, 0.98) 3.92 (3.17, 4.85) 0.27 (0.17, 0.42)
Urine NGAL (ng/ml) 56.72 (48.18, 65.25) 95.10 0.47 (0.34, 0.60) 0.67 (0.62, 0.71) 0.17 (0.07, 0.27) 0.90 (0.87, 0.93) 1.41 (1.03, 1.93) 0.79 (0.61, 1.03)
Plasma NGAL (ng/ml) 68.07 (60.09, 76.06) 194.00 0.61 (0.49, 0.73) 0.75 (0.69, 0.81) 0.38 (0.26, 0.50) 0.88 (0.84, 0.93) 2.44 (1.81, 3.30) 0.52 (0.37, 0.72)
Resistin (ng/ml) 80.17 (73.14, 87.20) 96.28 0.64 (0.51, 0.76) 0.85 (0.80, 0.90) 0.51 (0.38, 0.65) 0.90 (0.86, 0.94) 4.20 (2.90, 6.10) 0.43 (0.30, 0.61)
APTT slope 65.90 (58.89, 72.92) -0.03 0.46 (0.35, 0.57) 0.84 (0.81, 0.87) 0.33 (0.23, 0.43) 0.90 (0.87, 0.93) 2.86 (2.09, 3.92) 0.65 (0.53, 0.79)
APTT TR18 (%) 67.23 (60.29, 74.17) 99.40 0.51 (0.40, 0.62) 0.84 (0.81, 0.87) 0.35 (0.25, 0.45) 0.91 (0.88, 0.93) 3.13 (2.32, 4.22) 0.59 (0.47, 0.74)
CRP mg/l 89.76 (85.83, 93.69) 2.09 0.93 (0.87, 0.99) 0.76 (0.72, 0.79) 0.37 (0.25, 0.48) 0.99 (0.98, 1.00) 3.81 (3.21, 4.52) 0.09 (0.04, 0.22)
Model (1) PCT, Resistin, pNGAL, CRP 91.64 (87.58, 95.70) 0.12 0.92 (0.84, 1.00) 0.83 (0.77, 0.89) 0.57 (0.42, 0.73) 0.98 (0.95, 1.00) 5.42 (3.77, 7.79) 0.10 (0.03, 0.28)
Model (2) PCT, Resistin, APTT TR18, CRP 92.52 (88.22, 96.81) 0.08 0.97 (0.92, 1.00) 0.75 (0.67, 0.82) 0.52 (0.35, 0.68) 0.99 (0.97, 1.00) 3.83 (2.82, 5.19) 0.04 (0.01, 0.27)
Model (3) PCT, Resistin, CRP 92.36 (88.54, 96.17) 0.10 0.95 (0.88, 1.00) 0.79 (0.73, 0.86) 0.54 (0.38, 0.69) 0.98 (0.96, 1.00) 4.54 (3.30, 6.23) 0.06 (0.02, 0.25)
Model (4) PCT, CRP 92.66 (88.47, 96.85) 0.06 0.92 (0.85, 0.99) 0.86 (0.83 0.90) 0.50 (0.37, 0.62) 0.99 (0.97, 1.00) 6.67 (5.18, 8.58) 0.10 (0.04, 0.22)

*Pre-test probability is 12.89%.

#Cut-points were determined using the Youden Index, and probability cut-points were given for biomarker combinations from models (1)-(4), see below.

Prediction via logistic regression models:.

Model (1): p=2.68+6.31×103PCT+1.45×103Resistin+0.46×103pNGAL+14.79×103CRP.

Model (2): p=0.85+5.64×103PCT+1.73×103Resistin36.34×103APTT+20.10×103CRP.

Model (3): p=2.60+7.29×103PCT+1.73×103Resistin+15.16×103CRP.

Model (4): p=2.96+27.02×103PCT+18.49×103CRP.