A-003 Point of Care BNP Values Obtained by the Abbott i-STAT are not Interchangeable With Laboratory Determined Values Using Beckman DXI 800 Analyzer

Abstract

Abstract Background B-type natriuretic peptide (BNP) is used as a biomarker of heart failure. In our hospital, point of care (POCT) BNP test is performed on whole blood using the i-STAT (Abbott, Abbott Park, IL) while in clinical laboratory, the DXI 800 analyzer is used (Beckman, Brea, CA) for determination of BNP in EDTA plasma. Sometimes clinicians interchangeably use POCT BNP with laboratory BNP test for monitoring heart failure patients, but values are sometimes significantly different causing diagnostic confusion. We compared POCT BNP values in 88 patients with laboratory-based BNP values analyzed between less than one hour to less than 12 h. In addition, 12 specimens were simultaneously analyzed for POCT BNP and lab-based BNP. No additional blood was drawn from patients and medical record was not accessed. The study had exempt status: by the IRB of University of Kansas Medical Center. Methods POCT BNP values were obtained using the i-STAT analyzer following manufacturer’s protocol. It is a two-site enzyme-linked immunosorbent assay (ELISA) with an analytical measurement range of 15–5000 pg/mL. In the laboratory BNP concentrations were obtained using the DXI 800 analyzer. It is also a sandwich immunoassay with the analytical measurement range of 5–5000 pg/mL. Results Overall comparison using lab-based BNP as x-axis and POCT BNP as y-axis produced the following regression equation; y = 1.48 x − 23.45 (n = 88, r = 0.96) indicating that although POCT BNP values correlated well with lab-based BNP values, POCT values showed an average 48% positive bias. When specimens analyzed within 4 h were compared, the regression equation was: y = 1.40 x + 74.12 (n = 20, r = 91) and when specimens analyzed within 1 h were compared, the regression equation was: y = 1.56 + 59.0 (n = 23, r = 0.97). In contrast, when 12 specimens were analyzed simultaneously using POCT and lab-based BNP, using x-axis as the lab-based BNP and y-axis as the POCT BNP, the following regression equation was observed: y = 1.83 x − 102.8 (n = 12, r = 0.99). Conclusion Higher positive bias in specimens analyzed simultaneously indicates true positive bias with POCT BNP compared to lab-based BNP because lab-based BNP in real patients were ordered later after POCT BNP. To our knowledge i-STAT BNP values were never compared with DXI 800 BNP values. We communicated with our clinicians not to use lab-based BNP test and POCT BNP in monitoring patients being treated for heart failure interchangeably. POCT BNP can be used for initial screening to identify potential patients with heart failure.