Page 32 - Biomarkers for risk stratification and guidance in heart failure
P. 32
Multimarker risk score in emergency department dyspnea
*Variables entered into multivariable analysis (cutoff for entry p < 0.1). †Variables that remained significant in multivariable analysis including clinical risk factors (final clinical model). ‡Variables that remained significant in multivariable analysis including laboratory findings and were added stepwise to the final clinical model.
CI =confidence interval; NYHA = New York Heart Association functional class; OR = odds ratio. Other abbreviations as in Table 1.
plus hs-cTnT, hs-CRP, and Cys-C and reached a C-statistic of 0.86 (0.82-0.90) with 2 excellent calibration (Hosmer-Lemeshow p=0.78) (table 4) and significantly better
model perforance than single-marker models (NRI 13%, p=0.008 and IDI 5%,
p<0.001 compared with the model with clinical risk factors and hs-CRP alone,
p<0.001). Exclusion of patients with clinical signs of infection did not alter results (data not shown).
Figure 1. 90-day mortality rate by number of elevated biomarkers divided by (A) cause of dyspnea and (B) renal dysfunction.
ADHF=acute decompensated heart failure; Creat=creatinine. Cut-off for low/high creatinine was 1.6 mg/dl. Patient numbers per subgroup: ADHF n= 342, no ADHF n= 261, creat low n= 179, creat high n= 424. P-value for mortality rate by number of biomarkers elevated was <0.001 for all subgroups.
31
