Résumé
We evaluated assays of the same fresh blood samples with six different types of reference automated hematology analyzers developed by the following manufacturers: Beckman Coulter, Sysmex, Bayer, Abbott, Nihon Kohden and Horiba. Fresh whole blood samples treated with dipotassium ethylenediaminetetraacetic acid (EDTA K2) were collected from three healthy adult volunteers. The complete blood counts (CBC) including red blood cell count (RBC), hemoglobin (Hgb), hematocrit (Hct), mean corpuscular volume (MCV), white blood cell count (WBC), platelet count (Plt), reticulocyte percentage (Ret) and leukocyte differential counts including % neutrophils (Neu), % lymphocytes (Lym) and % monocytes (Mon) were surveyed with a reference automated hematology analyzer from each manufacturer. The process from sampling to analysis was performed according to procedures in hospital clinical laboratories. RBC, Hgb, Hct and MCV exhibited allowable differences within 5% of mean value among all instruments. Large differences greater than 10% of mean value in WBC, Neu and Lym between Horiba and other manufacturers, and in Plt between Nihon Kohden and other manufacturers, were observed. Ret and Mon exhibited large differences over 10% of mean value among almost all of the instruments tested. This survey suggests that all parameters exhibiting differences greater than 10% of mean value among instruments should be improved for clinical use to ensure good external quality control in blood cell counting and leukocyte differential counting using automated instruments.