Skip to main content
. 2015 Sep 19;34(4):440–449. doi: 10.2478/jomb-2014-0061

Table IV.

The results of ROC analysis for discriminating significant from subclinical coronary stenosis.

AUC (95% CI) p HL chi2
hs-CRP 0.66 (0.52–0.79) 0.031 14.35*
PTX3 0.67 (0.54–0.81) 0.017 8.41*
hsCRP and PTX3 0.71 (0.57–0.85) 0.004 11.09*
Model 2 0.68 (0.54–0.81) 0.020 14.03*
Model 2 and hsCRP 0.71 (0.58–0.84) 0.006 6.85*
Model 2 and PTX3 0.71 (0.58–0.85) 0.005 33.54
Model 2, hsCRP and PTX3 0.74 (0.60–0.87) 0.002 7.25*
Model 3 0.69 (0.56–0.82) 0.013 8.54*
Model 3 and hsCRP 0.761 (0.63–0.89) 0.001 12.20*
Model 3 and PTX3 0.74 (0.61–0.88) 0.002 21.73
Model 3, hsCRP and PTX3 0.772 (0.64–0.90) <0.001 13.29*

Hosmer-Lemeshow (HL) tests based on 10 groups.

*

Significant HL goodness of fit, p>0.05;

1

significantly different from model 2;

2

significantly different from model 3.