Improving clinical performance of urine sediment analysis by implementation of intelligent verification criteria

Abstract

Objectives Urinary test strip and sediment analysis integrated with intelligent verification criteria can help to select samples that need manual review. This study aimed to evaluate the improvement in the diagnostic performance of combined urinary test strip and urinary sediment analysis using intelligent verification criteria on the latest generation automated test strip and urinary fluoresce flow cytometry (UFFC) analysers. Methods Urine test strip and sediment analysis were performed using the Sysmex UC-3500 and UF-5000 (Kobe, Japan) on 828 urinary samples at the clinical laboratory of the Ghent University Hospital. The results were compared to manual microscopy using phase-contrast microscopy as a reference. After the application of the intelligent verification criteria, we determined whether the diagnostic performance of urine sediment analysis could be improved. Results Application of intelligent verification criteria resulted in an increase in specificity from 88.5 to 96.8% and from 88.2 to 94.9% for red blood cells and white blood cells, respectively. Implementing review rules for renal tubular epithelial cells and pathological casts increased the specificity from 66.7 to 74.2% and from 96.2 to 100.0%, respectively; and improved the diagnostic performance of urinary crystals and atypical cells. Conclusions The implementation of review rules improved the diagnostic performance of UFFC, thereby increasing the reliability and quality of urine sediment results.

Analytical performance evaluation of two automated urine sediment analysers using two levels of commercially available quality control material

“…2021, Oyaert et al . 2022) Quality control material level 1, the normal control with fewer urine constituents, generated fewer errors and more true values; though, a couple of instruments reported repeated erroneous results and the need for maintenance or replacement was identified. RBCs and WBCs were well identified, with low % inaccuracy.…”

“…Inaccurate identification of crystalluria and bacteriuria is common and clinical practices need standard operating procedures in place to ensure accurate crystal and bacterial identification to avoid incorrect treatment of patients. (O'Neil et al 2013, Hernandez et al 2019, Monteiro et al 2021, Oyaert et al 2022) Quality control material level 1, the normal control with fewer urine constituents, generated fewer errors and more true values; though, a couple of instruments reported repeated erroneous results and the need for maintenance or replacement was identified. RBCs and WBCs were well identified, with low % inaccuracy.…”