Abstract
BACKGROUND: Automated urine sediment analysis has been developed to address the limitations of microscopic examination of dysmorphic red blood cells (RBCs). We evaluated the urinary RBC distribution (URD) parameter of a recently launched automated urinary flow cytometry analyzer, UF-5000 (Sysmex, Kobe, Japan), to differentiate glomerular hematuria (GH) from non-GH (NGH). METHODS: Samples submitted for urine sediment analysis from patients with hematuria (>20 RBCs/μL) were divided into derivation (N=156; 101 GH, 55 NGH) and validation cohorts (N=107; 60 GH, 47 NGH). The clinical diagnosis of GH or NGH was established based on clinical data review. Differences in UF-5000 parameters (URD, small RBC, lysed RBC, RBC-P70FSC, RBC-SF-FSC-W, mean forward-scattered light, and mean side-scattered light) between GH and NGH, and areas under the ROC curves (AUC) were analyzed in the derivation cohort. The derived ideal cut-off value was evaluated in the validation cohort. We applied the Kitasato criteria to compare the diagnostic performance. RESULTS: URD (%), differed significantly between GH and NGH (P<0.001) in the two cohorts. The AUC of URD was 0.814 and 0.806 in the derivation and validation cohorts, respectively. Using a cut-off of >20.1%, the sensitivity was 99.0%/89.4% and the specificity was 50.9%/63.3% in the derivation/validation cohort. When the Kitasato criteria were applied, the sensitivity and specificity were 80.2% and 52.7%, respectively. CONCLUSIONS: URD is a rapid, objective, and quantitative measure that can be used to differentiate GH and NGH.