A Comparative Study of Automated Urine Sediment Analyzer Using Flow Cytometry and Digital Image-base with Manual Microscopic Examination

Abstract

This study evaluated the performance of three automated urine analyzers, the flow cytometry-based  UF-1000i and UF-5000, and the digital image-based US-2000, with manual microscopic urine sediment examination.  A total of 735 routine urine samples were examined focusing on 4 parameters, red blood cell (RBC), white blood cell (WBC), squamous epithelial cell and bacteria. The results showed the kappa correlation between all automated urine sediment analyzers and manual microscopic sediment examination were poor to good ranging from 0.083 to 0.759. The kappa correlation is fair, moderate and good for squamous epithelial cell, RBC and WBC ranging from 0.296 to 0.364, 0.559 to 0.570 and 0.694 to 0.759, respectively. There were poor to good agreements for bacteria ranging from 0.083 to 0.716. The concordance rates were 89.52-91.16, 90.07-92.65, 88.71-91.02, and 51.43-82.99 for RBC, WBC, squamous epithelial cell and bacteria, respectively. The study indicates that the three automated urine analyzers can be used in the laboratory to reduce high workload. However manual microscopic sediment examination by experienced staff is still necessary to classify urine sediments for confirmation.