Comparison of Manual Platelet Estimates and Automated Platelet Count at Tikur Anbessa Specialized Hospital, Addis Ababa

No Thumbnail Available



Journal Title

Journal ISSN

Volume Title


Addis Ababa University


Background: The accuracy of automated platelet counts can be compromised when measuring severely thrombocytopenic samples. This is especially of concern because current clinical guidelines lowered the prophylactic platelet transfusion threshold to 10 × 109/L for patients without additional risk factors. This count is a threshold for background count of many hematological analyzers, thus making validation of the automated platelet count a crucial step. Objective: To compare manual platelet estimates and automated platelet count by Sysmex KX 21N at Tikur Anbessa Hospital (TASH). Methods: A hospital based cross-sectional comparative study was conducted in Tikur Anbessa Specialized Hospital from January to March 2017 on 320 blood samples. Platelet count was enumerated by sysmex KX 21 and by manual method. Number of platelets/1000 RBC in PBS was multiplied by automated RBC count in 106/uL to get an estimate of platelet count in 103/uL and the average number of platelet/10 oil emersion field (OIF) was multiplied by 20,000. The data were entered and analyzed using SPSS version 20 and interpreted accordingly. Statistical significance was determined at 95% confidence interval. Pearson correlation coefficient was used to determine correlation between the methods. Bland Altman plot was used to assess agreement between tests. Figures and tables were used for the description of the data. Result: For normal platelet count the Paired t-test showed non-significant difference between automated and PLT estimated from average platelets/10 OIF (p>0.05), were positively correlated (r=0.994) and in the limit of agreement range (95%). For severe thrombocytopenic samples the two platelet estimation methods did not show significant difference when tested by paired t-test (p>0.05), were positively correlated (r=0.869) and in the limit of agreement range (97.5%). Conclusions: The mean PLT count by automated and PLT estimated from PLT:RBC ratio for normal count did not show significant difference so the two methods could give the same PLT count result. The mean PLT count for severe thrombocytopenic patient samples did not show significant difference when analyzed by the two manual platelet estimation methods. Key words: Peripheral blood smear, Platelet count, Platelet: red blood cell ratio, platelet estimate



Peripheral blood smear, Platelet count