| dc.description.abstract | Objective: To compare two platelet counting automated hematology analyzer
methods (PLT-F & PLT-O) in thrombocytopenic samples
Methods: This study is a diagnostic-cross sectional study at H. Adam Malik
General Hospital Medan, Indonesia (May – December 2022). We included 157
thrombocytopenic blood samples (K2-EDTA, thrombocyte count < 150.000 /μL)
from 114 patients. Platelet counts of optical-fluorescence flowcytometry / PLT-O
(Mindray BC-760, Shenzen, China) was compared to fluorescence flowcytometry
method / PLT-F (Sysmex XN-1000, Kobe, Japan). Bland-Altmann Plot and linear
regression test was used as analysis of method comparison (significant p < 0,05).
Results: Non-malignancy (60.5%) and hematologic disorders (62.3%) were the
most common etiology of thrombocytopenia. Median platelet count of PLT-F
(56,000 /μL [1,000 – 146,000]) and PLT-O (57,000 /μL [1,000 – 206,000]) were
not different (p = 0.963). Both platelet counting methods were in agreement
significantly and there was no proportion bias found in this study (Mean
difference -828.02; SD 11,939.98; Upper LoA 21,504.18; Lower LoA -23,160.20;
B -0.40; p = 0.05). Platelet counting flagging report of PLT-F and PLT-O were in
agreement significantly (p < 0.05) with minimal agreement in platelet clumping (κ
= 0,264), moderate agreement in abnormal platelet histogram and erythrocyte
microcytosis (κ = 0,651; κ = 0,793), and weak agreement in erythrocyte fragments
(κ = 0,483).
Conclusion: Both platelet counting automated hematology analyzer methods in
thrombocytopenic sample were significantly in agreement, however there was a
higher trends in platelet counts of PLT-O method compared to PLT-F method in
this study. | en_US |
