Quality Improvement of Urinalysis Results Based on Automatic Sediment Urinalysis and Urine Strip Results

BACKGROUND: Microscopic examinations are usually performed to confirm urine sediments in samples flagged in automated urinalysis. The aim of this study was to analyze the review rates and the difference in urinalysis results according to review rules. METHODS: A total of 1,408 urine samples submitted for health screening were collected. The urine chemistry test and urine sediment test were performed using EikenUS 3100 (Eiken Chemical Co. Ltd., Japan) and Sysmex UF-1000i (Sysmex Co., Japan), respectively. We assessed the rate of agreement between the 2 analyses and the kappa values for white blood cells (WBCs) and red blood cells (RBCs). Microscopic examinations were performed for all cases of discordant results between the urine strip and automated sediment analysis, some cases of concordant results, and cases of albuminuria. RESULTS: The review rate was 14.3%. Microscopic examinations were additionally performed on 77 samples (77/1,207, 6.4%) including 29 and 56 samples flagged for WBCs and RBCs, respectively. Based on the results of microscopic examination, the false-positive and the false-negative results of the urine chemistry test and automatic sediment analysis were corrected. Among concordant results between two methods, a clinically significant number of false-negatives were identified (6 results of WBC detection [6/125, 4.8%] and 4 of RBC detection [4/145, 2.8%]). Among the 22 unflagged cases of albuminuria, pathologic casts were detected in 21 cases (21/22, 95.5%). CONCLUSIONS: Microscopic examination based on the combined results of the two analyses improved the quality of the test.

Comparison of Analytical Performance between the Sysmex UF-100 flow cytometer and the Iris iQ200 Urine Microscopy System / 임상검사와정도관리

BACKGROUND: The UF-100 flow cytometer (Sysmex Co., Japan) and the Iris iQ200 (Iris Diagnostics, USA) are widely used for routine urinalysis in Korea. We compared the diagnostic accuracy of these two automated systems based on the microscopic finding, and evaluated the clinical performance of the automated systems. METHODS: A total of 323 fresh urine samples were selected and analyzed by conventional microscopy and the automation systems, the UF-100 and the iQ200. Quantification for RBCs, WBCs, and bacteria were also evaluated using both automated systems. RESULTS: For 158 of urine sample classified as normal urines, the agreement rate for the UF-100 and the iQ200 was 84.8% (N=134) and 89.9% (N=142), respectively. For 165 of urine samples classified as abnormal urines, the agreement rate for the UF-100 and the iQ200 was 90.9% (N=150) and 81.8% (N=135), respectively. The UF-100 showed a good linearity in the quantitative measurements of RBCs and WBCs. For both systems, false-negative value for WBCs and bacteria were about 30% in abnormal urines. Both systems showed inaccurate results for pathologic casts and bacteria. CONCLUSIONS: We compared the microscopic finding and the primary results of automated systems without user reclassification, and the agreement rate was about 85%. The agreement rate will be improved by deliberating "Review" comments of the instruments.