Evaluation of the Analytical Performance of Hemoglobin Analysis on the Point-of-Care Testing Device HemoCue Hb 801

Abstract Introduction Evaluating the accuracy and imprecision of Point-of-Care Testing (POCT) devices is an important prerequisite for ensuring proper diagnosis. The objective of this study was to assess the reliability of POCT hemoglobin (Hb), conducting a method comparison study on HemoCue Hb 801 and Sysmex XN 9000 as reference. Methods Thirty EDTA whole blood samples previously tested for Hb on Sysmex XN 9000 were also analysed on HemoCue Hb 801 within the measurement range 6.4-21.9 g/dL. Data were assessed by linear regression and correlation with a 95% confidence interval (CI) for the slope and intercept, a paired Student’s t-test, bias calculations and a Bland-Altman plot. Twelve independent measurements on the same sample material determined the imprecision of HemoCue Hb 801. Results We found a strong correlation coefficient (r 2 ) of 0.992 between Hb measured on HemoCue Hb 801 and Sysmex XN 9000 with a slope of 0.984 (CI: 0.956 to 1.021) and an intercept of 0.181 (CI: -0.283 to 0.544). The paired Student’s t-test indicated that the mean Hb differences between the methods were not statistically significant (p=0.84). Bias (-0.01 g/dL) did not deviate from the allowed maximum absolute bias (±0.35 g/dL). The values were within the 95% limits of agreement (±1.40 g/dL) for the differences. The coefficient of variation was 1.0%. Conclusion There was a strong correlation and good agreement between Hb measured on HemoCue Hb 801 and Sysmex XN 9000, as well as an acceptable bias and imprecision, supporting a good accuracy.