Comparison of red blood cell volume distribution curves and phase-contrast microscopy in localization of the origin of hematuria

Abstract

Objectives To determine the ability, reliability, and accuracy of urinary red blood cell volume distribution curves obtained with the use of an autoanalyzer to identify the origin of isolated microscopic hematuria and compare the results with those obtained with phase-contrast microscopy (PCM). Methods A prospective evaluation was performed in 45 patients with glomerular or nonglomerular microhematuria detected by urinalysis, PCM, radiologic evaluation, endoscopy, and, sometimes, renal biopsy. Urine samples were analyzed in an electronic particle-size analyzer, and the tests were repeated to assess reliability. The kappa correlation coefficient was used to assess reliability and to compare the results with the final diagnosis and with those obtained with PCM. Results Of the 28 patients who had a single definite cause of hematuria, 16 had glomerular bleeding and 12 nonglomerular bleeding. The origin of hematuria was correctly identified by the autoanalyzer in 60.7% of cases. A statistically significant correlation was found with the final diagnosis (kappa = 0.433, P = 0.048). The reliability was excellent (kappa = 0.917, P <0.0001). Of 16 patients with glomerulonephritis, 10 (62.5%) were correctly identified by PCM and 14 (87.5%) by the autoanalyzer. In 12 patients with nonglomerular bleeding, PCM was accurate in 7 (58%) and the autoanalyzer in 3 (25%). The results were statistically correlated with the findings of PCM (kappa = 0.327, P <0.00001). Conclusions The use of an autoanalyzer is easy, reproducible, and noninvasive. It provides reliable information to orient the diagnosis toward glomerular or nonglomerular bleeding. It is as accurate as PCM for screening for the source of hematuria.