Skip to main content
. Author manuscript; available in PMC: 2016 Sep 1.
Published in final edited form as: Glob Heart. 2015 Sep;10(3):173–180. doi: 10.1016/j.gheart.2015.07.003

Table 3. Results of analyses of the association of candidate biomarkers with mortality among patients with advanced Chagas cardiomopathy (Stage D) using receiver operating characteristic curves and Cox regression models.

Crude model Adjusted model2
Biomarker AUC1 Cut-off1 (pg/ml) Status Median (IQR) biomarker level in pg/ml N (%) above cutoff P HR1 (95% CI) P HR (95% CI)
BNP3 0.61 184.1 Alive 109 (9-263) 11 (37%) 0.039 2.8 (1.1-7.2) 0.025 3.1 (1.2-8.4)
Died 267 (37-347) 14 (70%)
NTproBNP3 0.66 765.1 Alive 425 (288-740) 5 (16.7 %) 0.002 4.1 (1.6-10.0) 0.002 4.4 (1.8-11.0)
Died 777 (423-1310) 11 (55%)
CKMB3 0.56 8300 Alive 3994 (3447-7487) 5 (17%) 0.005 3.5 (1.6-8.5) 0.009 3.3 (1.3-8.0)
Died 7534 (2960-11,438) 10 (50%)
MMP-23 0.72 103980 Alive 94,680 (75,549-111,884) 9 (30%) 0.008 4.0 (1.4-11.2) 0.007 4.2 (1.5-11.8)
Died 126,828 (96,148-164,256) 14 (74 %)4
1

Area under the curve (AUC), cutoffs and hazard ratios (HRs) based on receiver operating characteristic (ROC) curves. Hazard ratios were only calculated for those with AUC greater than 0.55.

2

Model adjusted for age and sex.

3

BNP, Brain Natriuretic Peptide; NTproBNP, N-terminal brain natriuretic peptide; CKMB, Creatine kinase-MB; MMP2, Matrix metalloproteinase 2.

4

MMP2 data were missing for one individual who died