Introduction
Modern haematology analysers provide an easy approach for enumerating the cellular components of blood in automated manner with improved turn-around-time. In our haematology laboratory, we have existing full blood count analyser, Sysmex XN-1500. We recently bought a backup full blood count analyser from Abbott Diagnostic, Cell Dyn Emerald 22 AL that enumerates cellular blood components using relatively similar measuring principles. As part of the method verification procedure required by ISO15189, we compared our newly purchased Cell Dyn Emerald 22 AL full blood count analyser to the existing verified Sysmex XN-1500 full blood count analyser. We further assessed the precision of both full blood count analysers. The residual EDTA blood samples were analysed using Sysmex XN-1500 (Sysmex Company Limited, Germany) as the reference analysers and CD Emerald 22 AL (Abbott, Abbott Park, Illinois, U.S.A) as newly purchased analysers.
Method and Results
We processed 120 residual blood samples for comparison study to improve the confidence interval in our statistical inference and incorporate the effects of unexpected interfering factors as reported in our previous study. Intra-run precision (repeatability) and inter-run precision (intermediate precision) were carried out using patient EDTA blood samples with sufficient volume to accommodate the runs. Using Pearson Product-Moment Correlation test, our data showed strong linear relationship between the two methods except for MCHC and PDW which generated moderate correlation coefficient (0.61, CI: 0.48–0.71) and weak correlation coefficient (0.228, CI: 0.469 − 0.395) respectively. The correlation results were supported by regression analysis scatterplot showing a weak linear relationship between the two analysers for only MCHC and PDW. Following linear relationship and regression analysis, we performed Bland-Altman analysis to determine agreement between the two analysers in a form of bias. The bias of all the tested full blood count parameters met the clinically acceptable bias range based on the 95% CI except for MCV, MCH, MCHC and MPV as they fall outside the clinically acceptable bias range.
Conclusion
From this study, we can confirm that Sysmex XN-1500 and Cell-Dyn Emerald 22 AL are comparable and interchangeable for all the clinically relevant full blood count parameters. However, MCV and HCHC as well as PLT derivatives (MPV and PDW) results should be treated with caution, preferably using a locally established equipment specific reference interval.