Evaluation of Optical Platelet Counts in HORIBA Yumizen H2500 and Platelet Counts by Digital Morphology Platform in Cases of Thrombocytopenia with Platelet Interference Flag: Finding Solutions with Automation in High Volume Laboratory.

Platelet counts which are generated by impedance methods (PLT-I) in cases of thrombocytopenia can be unreliable due to interference of red cells. We evaluated utility of optical platelet counts on HORIBA Yumizen H2500 and platelet count by digital morphology platform CellaVision DM1200 with multivariate analysis for understanding causes in cases of thrombocytopenia with platelet interference flag. A total of 150 samples with platelet counts < 150 X 10⁹/L (PLT-I) were selected for this study. Platelet counts were then analyzed in parallel within 6 h of collection using optical technique (PLT-O) and also PLT-D (Digitalized Smear techniques). Platelet count on peripheral smear was also performed by two pathologists (MPC). Statistical analysis using Pearson's correlation, Bland-Altmann plot analysis and multivariate regression analysis was performed. The mean platelet count with PLT-I was 74.6 × 10⁹/L while this was significantly higher in PLT-O (108.9 × 10⁹/L). Pearson's correlation between PLT-O and MPC as well as PLT-D and MPC showed a good correlation between the two values, proving the utility of PLT-O over PLT-I in cases of RBC -platelet interference flags. Multivariate regression analysis demonstrated a positive effect of MPV and indicated that with increase in MPV the difference between PLT-I and PLT-O is significantly affected. PLT-O was found to give more accurate platelet count than PLT-I in cases of thrombocytopenia with platelet interference flag. Also, digital morphology is a useful alternative to manual platelet count estimation.