Analytical relationships among Biosite, Bayer, and Roche methods for BNP and NT-proBNP

This study determined whether, for patient monitoring, it is feasible to convert B-type natriuretic peptide (BNP) results obtained using Triage (BNP, Biosite, San Diego, CA), Centaur (BNP, Bayer Diagnostics, Tarrytown, NY), and Elecsys 2010 (N-terminal proBNP; Roche, Indianapolis, IN) assays. Concordance between assays and effects of renal impairment also were assessed. Samples were primarily from emergency center patients. Biosite testing was performed immediately; Bayer and Roche testing was performed later on plasma stored frozen (-30 degrees C). Logistic regression relationships were as follows: Bayer = 0.57 Biosite + 23.1, n = 121, R2 = 0.85; Roche = 6.09 Biosite -220.4 + 1,131.6 (if female), n = 131, R2 = 0.57; and Roche = 15.34 Bayer + 2,400.8, n = 150, R2 = 0.23. An increased serum creatinine level (>/=2 mg/dL [>/=177 micromol/L]) influenced the Roche results. We conclude the following from this preliminary study: (1) Results from one method cannot be converted reliably to another using regression relationships. (2) When using manufacturers'cutoff values, concordance between assays was acceptable. (3) Renal impairment affected Roche results.