Measures of Blood Hemoglobin and Hematocrit During Cardiac Surgery: Comparison of Three Point-of-Care Devices

Objective: The primary objective was to compare I-Stat, HemoCue, and RapidLab in measurements of the hemoglobin concentration during cardiac surgeries using cardiopulmonary bypass. Design: Prospective analysis. Setting: Single-center, academic, tertiary care cardiovascular center. Participants: Thirty-four consecutive patients undergoing cardiac surgery requiring cardiopulmonary bypass. Interventions: Blood samples have been collected intraoperatively, and the hemoglobin concentration in each sample was measured, or calculated, simultaneously by the 3 point-of-care devices, HemoCue, RapidLab, and I-Stat. Measurements and Main Results: Correlation coefficients from the regression analysis for HemoCue versus I-Stat, RapidLab versus HemoCue, and RapidLab versus I-Stat were 0.89, 0.96, and 0.88, respectively. Results of the Bland-Altman analysis of the hemoglobin concentration measurements for each device against one another (Fig 1) were as follows: RapidLab versus I-Stat (bias 0.42; 95% confidence interval [CI], -1.05 to 1.89), I-Stat versus HemoCue (bias 0.23; 95% CI, -1.14 to 1.59), and RapidLab versus HemoCue (bias 0.65; 95% CI, -0.17 to 1.47). It appears that I-Stat slightly underestimated the concentration of hemoglobin when compared with both RapidLab and HemoCue. The results of Bland-Altman analysis of each device to a mean Z value (Fig 2) were as follows: RapidLab versus Z (bias 0.36; 95% CI, -0.29 to 1.01), I-Stat versus Z (bias -0.07; CI -0.97 to 0.84), and HemoCue versus Z (bias -0.29; 95% CI, -0.86 to 0.28). Based on the 174 paired samples used for the Pearson moment analysis, the R-2 values for I-Stat versus HemoCue, I-Stat versus RapidLab, and RapidLab versus HemoCue were 0.79, 0.80, and 0.87, respectively Conclusions: These data support the interchangeability of these 3 devices for the intermittent intraoperative point-of-care assessment of hemoglobin concentrations in cardiac surgery patients. It is important, however, to consider the possible pitfalls associated with each device when making a clinical decision to transfuse (C) 2017 Elsevier Inc. All rights reserved.