Urine Sediment Examination: Comparison Between Laboratory-Performed Versus Nephrologist-Performed Microscopy and Accuracy in Predicting Pathologic Diagnosis in Patients with Acute Kidney Injury

Abstract

Introduction Automated urine technology is becoming the standard for urinalysis microscopy. We sought to compare urine sediment analysis performed by a nephrologist with the analysis performed by the laboratory. When available, we also compared the suggested diagnosis per nephrologists' sediment analysis with the biopsy diagnosis.MethodsWe identified patients with AKI who had urine microscopy with sediment analysis performed by the laboratory (Laboratory-UrSA) and by a nephrologist (Nephrologist-UrSA) within 72 hours of each other. We collected data to determine the following: number of red blood cells (RBCs) and white blood cells (WBCs) per high-power field, presence and types of casts per low-power field, and presence of dysmorphic RBCs. We evaluated agreement between the Laboratory-UrSA and the Nephrologist-UrSA using cross-tabulation and the Kappa statistic. When available, we categorized the nephrologist sediment findings into four categories: (1) bland, (2) suggestive of acute tubular injury (ATI), (3) suggestive of glomerulonephritis (GN), and (4) suggestive of acute interstitial nephritis (AIN). In a group of patients with kidney biopsy within 30 days of the Nephrologist-UrSA, we assessed agreement between the nephrologist diagnosis and the biopsy diagnosis.ResultsWe included 387 patients with both Laboratory-UrSA and Nephrologist-UrSA. The agreement was moderate for the presence of RBCs (Kappa, 0.46; 95% CI, 0.37 to 0.55) and fair for WBCs (Kappa, 0.36; 95% CI, 0.27 to 0.45). There was no agreement for casts (Kappa, 0.026; 95% CI, -0.04 to 0.07). Eighteen dysmorphic RBCs were detected on Nephrologist-UrSA compared with zero on Laboratory-UrSA. Among the 33 patients with kidney biopsy, 100% ATI and 100% GN suggested per Nephrologist-UrSA were confirmed on the biopsy. Of the five patients with bland sediment on the Nephrologist-UrSA, 40% showed ATI pathologically while the other 60% demonstrated GN.ConclusionA nephrologist is more likely to recognize the presence of pathologic casts and dysmorphic RBCs. Correct identification of these casts carries important diagnostic and prognostic value when evaluating kidney disease.