Urinary Red Blood Cell Distribution for Detection of Glomerular Hematuria: A Biopsy-Based Comparison with Direct Microscopy

Background: Differentiating glomerular from non-glomerular hematuria remains challenging. Automated urine analyzers with red blood cell (RBC) discrimination algorithms may provide an adjunct to conventional urine microscopy. We evaluated the diagnostic performance of the UF-5000 analyzer for detecting glomerular hematuria (GH). Methods: In this cross-sectional diagnostic study, patients with hematuria were classified into a GH group with biopsy-proven glomerular disease and a non-glomerular hematuria (NGH) group with non-glomerular causes confirmed by clinical, radiological, and/or pathological evaluation. Urine specimens were analyzed by a technician, a nephrology fellow, and the UF-5000 analyzer. Diagnostic performance was assessed using receiver operating characteristic (ROC) curve analysis and multivariable logistic regression. Results: A total of 201 patients were included, comprising 101 in the GH group and 100 in the NGH group. Urinary RBC distribution (URD) from UF-5000 analyzer was significantly higher in the GH group (66.3 ± 23.6 vs. 12.6 ± 19.8%; p < 0.0001). URD demonstrated the highest diagnostic performance among evaluated UF-5000 parameters, with an area under the ROC curve (AUC) of 0.946 (95% CI 0.914–0.979). The optimal Youden-derived URD cutoff ≥ 25% yielded 91.1% sensitivity and 89.0% specificity for identifying GH. Compared with microscopy, URD showed higher sensitivity than technician-performed microscopy (91.1% vs. 60.4%; p < 0.001) and similar performance to fellow-performed microscopy (91.1% vs. 83.2%; p = 0.237). Performance remained consistent across RBC count strata and major glomerular disease subtypes. Conclusions: URD was strongly associated with GH and demonstrated good discriminatory performance in this biopsy-based cohort. Automated urinary RBC analysis may complement conventional urine microscopy in hematuria evaluation.