PERFORMANCE EVALUATION OF NUCLEATED RED BLOOD CELL (NRBC) COUNT USING A FULLY AUTOMATED HAEMATOLOGY ANALYZER VERSUS MANUAL COUNTING

Objective: To evaluate the performance of Nucleated RBC (NRBC) Count using a fully automated haematology analyzer versus manual counting. Study Design: Cross-Sectional Study. Place and Duration of Study: Department of Hematology, Armed Forces Institute of Pathology, from Sep 2019-Jun 2020. Methodology: Routine fresh whole blood samples were run on Sysmex XN-3000 automated haematology analyzer and 384 samples with results of ≥0.1% Nucleated red blood cells were included in this study. Manual NRBC counting was carried out twice on Leishman-stained peripheral blood smears from all 384 samples. Comparison between manual and automated nucleated red blood cell counting methods was statistically analyzed through linear regression analysis & coefficient correlation. The degree of agreement between two methods was analyzed through Bland-Altman plot. Finally, concordance between the two methods was also analyzed at 5 different ranges of nucleated red blood cells. Results: Linear regression analysis revealed a (r2) value of 0.97. Regression equation was calculated as XN = 0.76MC ± 1.28, with 95% limits of agreement between ± 40.42% and -24.47%. A mean bias of 7.97% was demonstrated through Bland-Altman plot. Concordance analysis revealed a concordance rate of 93.74% (360/384). Nucleated red blood cell counting between two methods were more concordant when nucleated red blood cell counts were <200%. Conclusion: Nucleated red blood cells counting by XN-3000 automated hematology analyzer is statistically comparable to manual nucleated red blood cell counting. We suggest that automated counting can be adopted in routine hematology laboratory as a replacement of manual NRBC counting.