A comparison of four bedside methods of hemoglobin assessment during cardiac surgery.

The purpose of this study was to compare the accuracy of conductivity, adjusted conductivity, photometric, and centrifugation methods of measuring or estimating hemoglobin (Hb) with Coulter measured HB as the reference. These bedside methods were studied in 25 cardiac surgery patients during euvolemia and hemodilution and after salvaged autologous red blood cell transfusion. In vivo patient blood samples were obtained before induction, at the start of cardiopulmonary bypass (CPB), after CPB, and after blood transfusion. In 10 patients, blood was sampled in vitro from units of processed blood. Hb values were determined using conductivity by Stat-Crit, adjusted conductivity by Nova Stat Profile 9, bedside photometry by HemoCue, and centrifugation methods. The calculated bias values of Coulter test method Hb (mean +/- SD) for in vivo patient blood samples (n = 90) were: Stat-Crit = 0.6 +/- 0.8 g/dL; Nova Stat Profile 9 = -0.7 +/- 0.4 g/dL; HemoCue = -0.1 +/- 0.2 g/dL; and centrifuge = 0.1 +/- 0.5 g/dL (P < 0.0001). Hb bias values (g/dL) for in vitro samples (n = 10) obtained from processed blood were Stat-Crit = 5.1 +/- 0.6; Nova Stat Profile 9 = 3.0 +2- 0.6; HemoCue = 0.4 +/- 0.4; and centrifuge = 0.6 +/- 0.3 (P < 0.0001). Hb assessment by different test methods may be significantly affected during hemodilution and after blood transfusion. In vitro conditions exaggerated the inaccuracy of conductivity and adjusted conductivity Hb estimates. The rank order of closest approximation to the Coulter measurement for all in vivo blood samples was provided by bedside photometry, followed by centrifugation, adjusted conductivity, and uncorrected conductivity methods.