RESUMO
This is the second report of evaluation on the difference of blood cell counting among different automated blood cell counters in Japan. We tested reference blood cell counters of 6 different companies: Coulter, Sysmex, Bayer-Sankyo, Nihon Kohden, Horiba and Dainabot. Forty ml of whole blood were taken from 3 healthy persons and EDTA-2K anticoagulated blood samples (Sample 1, 2 and 3) were sent to each company to determine blood cell counts with a reference automated counter. As a result, the following items showed more than 10% difference among makers: RBC between Dainabot and Horiba in Sample 3, hematocrit values between Coulter and Dainabot in Sample 2, WBC between Nihon Kohden and each of three makers (Sysmex, Horiba and Dainabot) in all 3 samples and that between Bayer-Sankyo and each of two makers (Sysmex and Horiba) in Sample 3 and platelet count between Dainabot and each of 3 makers (Bayer-Sankyo, Nihon Kohden and Horiba) in all 3 samples. The following items showed difference between 5 and 10%: MCV between Coulter and each of two makers (Bayer-Sankyo and Horiba), WBC between each of two makers (Coulter and Nihon Kohden) and each of other 4 makers, and platelet count between each of two makers (Nihon Kohden and Horiba) and each of 3 makers (Coulter, Sysmex and Bayer-Sankyo). Recently Japanese Committee for Clinical Laboratory Standards proposed minimum clinical allowance of blood cell count as follows: hemoglobin 3%, RBC 4%, MCV 4%, WBC 7% and platelet count 10%. It is suggested that all of the items showing the difference more than above allowance among makers should be improved for clinical use to have good external quality control in blood cell counting by automated instruments.
