Analytical and Clinical Performance of the Atellica <scp>HEMA</scp> 580 Haematology Analyzer: A Comparison With the <scp>ADVIA</scp> 2120i

INTRODUCTION: The Atellica HEMA 580 is a novel haematology analyser providing eight-part leukocyte differential and optical platelet (PLT) counting. We compared its analytical and clinical performance with that of the ADVIA 2120i. METHODS: A total of 796 residual whole blood specimens were analysed. Carryover, linearity and precision were assessed. Inter-instrument comparison was performed for complete blood counts and five-part differential (5-Diff) parameters using Passing-Bablok regression and Bland-Altman analyses. Additional regression against manual microscopy was conducted for flagged samples, including PLT, reticulocytes and common 5-Diff parameters. RESULTS: ≥ 0.99). Precision was within acceptable limits; coefficient of variant was < 5% for most parameters, except optical PLT, lymphocytes, monocytes, eosinophils and nucleated red blood cells under specific conditions. Inter-instrument comparisons revealed remarkably high correlations (r ≥ 0.9) for most parameters, whereas mean corpuscular haemoglobin concentration (MCHC) (r = 0.429) and basophils (r = 0.627) showed weaker concordance. Passing-Bablok regression indicated minimal systematic or proportional bias for most parameters. Bland-Altman analysis showed that > 95% of results were within the 95% limits of agreement, except white blood cells, mean corpuscular volume and PLT (94.7%-94.8%). Comparison with microscopy confirmed good concordance for PLT and reticulocytes; however, lymphocytes and monocytes had proportional and systematic biases. CONCLUSION: The Atellica HEMA 580 demonstrated reliable analytical performance and strong agreement with the ADVIA 2120i. This supports its suitability as a routine haematology analyser, providing accurate and clinically comparable results.