Background: The plasma concentration of B-type natriuretic peptide (BNP) in outpatients is hard to interpret because of the lack of knowledge of the natural within-person variation of BNP. In this study, we estimated the safety range of within-person variation in the plasma concentration of BNP in outpatients with stable heart failure.
Methods: In a prospective historical cohort study, 6 consecutive measurements of the plasma concentration of BNP were made at 4-week intervals in 131 consecutive patients with strictly stable heart failure. The reference change values at the 95% CI and the 95% limits of agreement of the peptide concentration were calculated with a log-normal approach, and the results were back-transformed to a normal scale.
Results: The within-person distribution of BNP was right-skewed, and a Gaussian distribution was achieved by log transformation. In all of 15 combinations of paired measurements randomly selected from 6 measurements, the correlations were significant (P < .001) with correlation coefficients ranging from 0.615 to 0.835. The up and down reference change values at the 95% confidence level for all measurements were 240.4% and -70.6% of the geometric means, respectively, and the median values of the upper and lower 95% limits of agreements were 219.7% and -71.8% of the geometric mean, respectively.
Conclusion: The plasma concentration of BNP may triple or fall by one third without a change in the status of heart failure. In monitoring of patients with heart failure, BNP should be interpreted in the context of the skewed within-person distribution.