RESUMO
BACKGROUND: BNP is a sensitive and widely used biomarker for an early diagnosis of heart failure. Currently, most commercial BNP detection products use EDTA plasma samples. The aim of this study was to evaluate the clinical performance of the BNP test by using whole blood samples compared to plasma samples, and to evaluate the effect of the anticoagulant type on the BNP test result. METHODS: In total, 106 patients with different BNP levels from the Dahua Hospital volunteered for this study. Clinically homogenous samples, including EDTA anticoagulant plasma, EDTA whole blood, and heparin anticoagulant plasma, were collected and analyzed by using i-Reader S automatic immuno-analyzer and its supporting reagent kits. Pearson's correlation and weighted least squares linear regression analysis, Bland-Altman plotting, and Kappa test were used for statistical analysis. RESULTS: Correlation analysis showed that BNP concentrations, measured from EDTA anticoagulated plasma samples, had a good linear regression relationship with BNP from whole blood samples, with a slope of 0.9477, r = 0.9978, p < 0.05. A similar correlation was observed between EDTA anticoagulated plasma samples and heparin anticoagulant plasma, with a slope of 0.8413, r = 0.9793, p < 0.05. The BNP concentration measured from the heparin plasma samples were lower than of the EDTA plasma samples. Bland-Altman analysis for assessing BNP concentration agreement showed there was no outlier ratio between EDTA whole blood and EDTA plasma within the range of the detection system, as well as no outlier between EDTA anticoagulated and heparin anticoagulant plasma. Kappa coefficient of BNP concentration between homologous EDTA anticoagulated and heparin anticoagulant plasma was 0.8553 (p < 0.001), and for EDTA anticoagulated plasma and homologous whole blood it was 0.8941 (p < 0.001). CONCLUSIONS: The diagnostic performance of EDTA anticoagulated whole blood samples did not differ significantly from EDTA anticoagulated plasma samples for the BNP test. This study showed no big significant difference between EDTA anticoagulated and heparin anticoagulated plasma measurements within 2 hours. The type of anticoagulant should be carefully chosen when performing the BNP test if BNP samples were in vitro for a long time.