Evaluation of Automated Haematology Analyser Performance Against Manual Methods in a District Hospital Laboratory in Kwara State, North Central Nigeria

Automated haematology analysers have become standard in clinical laboratories for full blood count determination, but their performance characteristics require validation against manual methods in local operational contexts to ensure reliability. This study evaluated the performance of an automated haematology analyser against manual microscopic methods at General Hospital Ilorin, Kwara State, North Central Nigeria. A method comparison study was conducted using 200 EDTA blood samples from haematology and general outpatient clinic patients. The Sysmex XP-100 analyser was evaluated for haemoglobin, white blood cell count, differential leucocyte count, and platelet count against manual haemoglobin by Drabkin's method, chamber counting for WBC and platelets, and microscopic differential on Giemsa-stained blood films. Bland-Altman analysis, Pearson correlation, and Passing-Bablok regression were applied. Strong correlation was found for haemoglobin and WBC with r values of 0.98 and 0.97. Platelet concordance was good at r of 0.94, though the analyser underestimated platelet counts in samples with platelet clumping. Differential count correlation was moderate for eosinophils and monocytes. The analyser flagged abnormal morphology requiring manual review in 18.5% of samples. The study establishes the Sysmex XP-100 as reliable for routine blood count parameters but recommends mandatory microscopic review of all flagged samples and regular quality control calibration to maintain analytical accuracy at district laboratory level. Keywords: automated haematology analyser, manual methods, method comparison, blood count, Kwara State