Related Experiment Video
Updated: Aug 6, 2026
Evaluation of a Point-of-Care Testing Analyzer for Measuring Peripheral Blood Leukocytes
Published on: March 22, 2022
Analytical agreement and clinical interchangeability of routine complete blood count parameters between the Atellica
Quang Tung Nguyen1,2, Van Tuan Pham2, Chi Thanh Nguyen2,3
1Department of Hematology, Hanoi Medical University, Hanoi, Vietnam.
Background:
Analytical consistency in complete blood count (CBC) testing is essential when introducing new hematology analyzers into routine clinical use. This study evaluated the agreement between the Atellica HEMA 580 and the Sysmex XN-1000 using a standardized method-comparison approach.
Methods:
A total of 100 residual K₂EDTA whole-blood samples were analyzed on both systems between April and June 2025. Method comparison followed CLSI EP09c guidelines, including Passing-Bablok regression, Spearman correlation, bias estimation at predefined medical-decision levels, and Bland-Altman analysis.
Results:
Strong analytical agreement was observed for most parameters, with correlation coefficients exceeding 0.98 for key indices including RBC (r = 0.999), HGB (r = 0.998), WBC (r = 0.996), and PLT-I (r = 0.995), and regression slopes close to unity. Bias estimates at medical-decision levels were within allowable limits for the majority of parameters. WBC at 0.5 × 109/L and PLT-I at 100 × 109/L achieved Outcome A, indicating optimal agreement at clinically critical thresholds. In contrast, MCV showed proportional deviation with bias exceeding allowable limits at both 80 fL and 100 fL, and RDW-CV demonstrated lower correlation (r = 0.848) and greater dispersion. Bland-Altman analysis confirmed small mean bias for most parameters, with narrow limits of agreement for RBC (-1.48%, LoA - 4.19 to 1.36) and HGB (1.54%, LoA - 1.37 to 4.19), while wider limits were observed for WBC (-12.62 to 1.70) and PLT-I (-16.55 to 9.51). RDW-CV exhibited the broadest limits of agreement (-17.76 to 10.38). Among differential parameters, lymphocyte count showed optimal agreement (bias -1.12%, Outcome A), while other indices remained within acceptable ranges.
Conclusion:
The Atellica HEMA 580 shows acceptable analytical agreement with the Sysmex XN-1000 for RBC, HGB, WBC, PLT-I, and the evaluated neutrophil and lymphocyte differential parameters, and can be applied in clinical laboratory practice. In contrast, MCV and RDW-CV demonstrated analyzer-dependent differences that warrant cautious interpretation when comparing results across platforms or during longitudinal assessment.
