Background: We aimed to establish an equation for the estimation of the BNP concentration in plasma when only serum is available.
Methods: We enrolled 27 subjects aged at least 45 years, participating in a Portuguese cohort study. Blood samples were collected in plastic whole blood tubes, containing either ethylenediaminetetraacetic acid to obtain plasma or clot activator to obtain serum. The natural logarithm of serum BNP was calibrated with the natural logarithm of plasma BNP using a linear equation.
Results: The estimated regression parameters were 0.58 (95 % CI: 0.23 - 0.93) for beta0 and 1.01 (95 % CI: 0.90 - 1.11) for beta1. The absolute agreement between plasma BNP and that predicted by the equation according to the cut-off points 30 and 100 pg/mL were 96.3% (kappa = 0.92) and 96.3% (kappa = 0.91), respectively.
Conclusions: Serum samples cannot be used to estimate absolute plasma concentrations, but serum BNP values and the calibration equation can be used to classify correctly the individuals with the usual cut-offs.