| Background:Nowadays hematology analysis has been the general test item in clinic. Automatically 5-differential hematology analyzer provides accurate results for clinic as the auxiliary indication of disease. Meanwhile, hematology analyzer provided limited information about the morphology of WBC, RBC and platelet, so it should be essential to take microscopic examination for the smears of peripheral blood cells in clinic. Recently, BCM-2 automated blood cell morphological differential system is a fully automatic hematology analyzer, and it can test blood smears automatically. It is not clear whether it could replace the manual examination of microscopy and whether there would be any difference from manual performances.Objective:Observe the results on degree of accuracy and correlation after BCM-2 automatic system analysis and manual correction. Evaluate the results on correlation of BCM-2 automatic system analysis, manual correction and XE-2100 automatic hematology analyzer. Evaluate the results on correlation of BCM-2 automatic system analysis, manual correction and microscopic examination. Compare the time spent on BCM system analysis and manual microscopic examination to evaluate the work efficiency improvement.Methods:425 Outpatient and hospitalized cases were collected randomly by clinical laboratory for outpatients in Qilu Hospital of Shandong University from April 2014 to September 2014. Blood was treated with EDTA K2 to be anticoagulated and analyzed by XE-2100 hematology analyzer.Then to smear automatically by sp-1000i, an automated slide-making staining system. After stain was used automatically. White blood cells were differentiated and analyzed in a term of percentage from the prepared smears by BCM-2 system. Finally,3 experienced chief technicians with intermediate professional title corrected the results obtained from BCM-2 system manually. Meanwhile, all the blood smears were examined manually to differentiate and count. Correlation analysis was used for the results from different methods for WBC differential counts. Peripheral blood cells were analyzed by BCM-2 automated morphology differential system automatically and corrected manually to evaluate the correlation. After that, the results of XE-2100 system, BCM-2 system (after corrected manually) and manual examination of microscopy were compared and correlation was analyzed. Using manual examination of microscopy as a gold stand, it was analyzed whether the results obtained by BCM-2 system and XE-2100 system were consistent. The results obtained by BCM-2 system and XE-2100 system on morphology and counts of RBC and platelet were compared to evaluate the capacity of BCM-2. At last, it was evaluated the efficiency of BCM-2 and whether it could replace manual examination of microscopy depending on workflow as above by 3 technicians with different working experiences.Data processing and statistical methods:The data was analyzed by Excel(Microsoft Office 2010, Microsoft Inc,Redmond,us) and SPSS (SPSS18.0,SPSS Inc,IBM,Chicago,US).Values of measurement data are the mean±SD. The correlation between results was determined by linear regression analysis and Kappa test. Instrument analysis ability of cell identification was evaluated by the coincidence rate (with double side test). The criterion for significance was P< 0.05.Results:1. The total precision of BCM-2 automatic analyzer reached 94.9%. It was more accurate for most matured cells (band neutrophils, segmented neutrophils and lymphocytes), the highest accuracy of which was lymphocytes (98.9%), followed by segmented neutrophils (95.1%) and monocytes (92.8%), eosinophils (92.2%), and band neutrophils (83.7%). But the accuracy of basophils, or juvenile cells (blast stage, early stage, middle stage and late stage) was unsatisfactory, which was 55.6% and 44.4%, respectively.2.The correlation coefficients on band neutrophil, segmented neutrophil, monocyte, lymphocyte and eosinophil differentials were 0.732,0.965,0.974,0.905 and 0.952, respectively. It demonstrated well relation between before and after correction for all kinds of cells by BCM-2 system. However, the correlation coefficients on basophil and juvenile cell differentials were lower, which were 0.581 and 0.038 only.3. After automatic differentials by BCM-2 system, to XE-2100 system, the correlation coefficients on WBC differentials (neutrophils, lymphocytes, monocytes, eosinophils and basophils) were 0.93,0.904,0.467,0.852 and 0.219, respectively. And the correlation coefficients were 0.929,0.907,0.476,0.869 and 0.378, respectively, after correction. Performances on differential and count could only be alarmed rather than taken for juvenile cells and abnormal lymphocytes by XE-2100 system, the results were analyzed with juvenile cells and abnormal lymphocytes depending on the results of microscopic examination as the standard (>3%and>5%)). Kappa consistency analysis showed that the consistency between BCM-2 and XE-2100 automatic analyzer for juvenile cells and abnormal lymphocytes was unsatisfactory, which Kappa value was 0.005 (P=0.883) and 0.064 (P=0.055).4. After automatic differentials by BCM-2 system, to microscopic examination on WBC differentials (band neutrophils, segmented neutrophils, lymphocytes, monocytes, eosinophils and basophils), the best correlation coefficient was 0.91 and 0.938 to segmented neutrophils and lymphocytes, respectively, followed by the correlation coefficient to band neutrophils, eosinophils and monocytes (0.746,0.812 and 0.623, respectively). The worst correlation coefficient was 0.355 to basophils. The correlation coefficient between the results of post-manual correction and microscopic examination for segmented neutrophils was 0.937, while it was 0.935 for lymphocytes, 0.795 for band neutrophils,0.688 for monocytes,0.822 for eosinophils, and 0.52 for basophils. Since the juvenile cells were less, the correlation coefficient was unsatisfactory, which was 0.104,0.061,0.005 and 0.037 for promyelocytes, myelocytes, metamyelocytes and blast, respectively. After manual correction, the correlation coefficient was 0.187,0.114,0.01 and 0.076. It was not analyzed in most cases, as quantity of nucleated red blood cell was less and the objective was the patient with hemophilia or from the department of paediatrics.5. It was better for BCM-2 system to observe the morphology of RBC and platelet than XE-2100 automatic analyzer, which was applied on counts only. There was no obvious difference between BCM-2 and XE-2100 system. The morphology was consistent to microscopic examination. For the morphology of platelet, several large platelets and platelet aggregation could be observed by BCM-2 system or microscopy, while there was no sensitive morphology provided by XE-2100 system. The correlation coefficient was 0.731 between BCM-2 system and XE-2100 system for platelet counts, while it reached 0.942 after correction, and it was no significant difference from microscopic examination.6. The time spent on instrument analysis and manually microscopic examination was compared.10 cases were selected for smear. It took 62s for BCM-2 system to reset, 24s to put the rack with smear to the specific location and 22s to take smear upon the microscope stage. When counting WBC up to 200, the time spent from blood smear location to instrument resetting until analysis accomplishment was illustrated in table The average time for one sample scanning was 44.7±6.45s, for oil lens dripping was 13s, for reading was 162.2±10.72s, and for slides return was 18s.3 experienced chief technicians with intermediate professional title took microscopic examination. The work seniority was 2 years,15 years and 25 years. The time spent on microscopic examination was 299.8±92.47s,273.3±61.5s and 251.4±64.86s, respectively, and the average time was 274.8±60.84s. For the technician with short working life, BCM-2 system could save time to improve efficiency, while it cost the technician with long working life similar time on examination by either BCM-2 or microscopy.Conclusions and suggestions:BCM-2 as a new kind of hematology identification device, The correlation coefficient of post-manual correction against XE-2100, a 5-differential hematology analyzer and microscopy for manual examination was satisfactory. It provided accurate and reliable database for clinical studies. Meanwhile, it could reduce working intensity and hours to improve efficiency. |
PDF Full Text Request