Tsai JJ et al. (2005) Comparison and interpretation of urinalysis performed by a nephrologist versus a hospital-based clinical laboratory. Am J Kidney Dis 46: 820–829

A prospective study at the University of California Davis Medical Center has revealed worrying inaccuracies in urinalyses performed by the centers' affiliated clinical laboratory. The findings call into question the Clinical Laboratory Improvement Amendments of 1988, which recommend that certified technicians conduct laboratory tests rather than physicians.

Four sets of diagnoses were assigned by two blinded nephrologists (A and B) to each of 26 patients with acute renal failure, based on urinalyses by Nephrologist A and the clinical laboratory. Diagnoses were compared with those determined by consulting nephrologists with full access to patients' clinical history. The accuracy of Nephrologist A's diagnoses based directly on his own observation of patient urine (92.3%) was greater than his own or Nephrologist B's accuracy using laboratory-generated reports (23.1% and 19.2%, respectively; P <0.0001 for both). Renal tubular epithelial cells and casts, granular casts, hyaline casts and ACANTHOCYTES were detected more frequently by Nephrologist A than by laboratory staff, who probably misclassified some cells as squamous epithelia.

The researchers suggest that such errors are common in clinical laboratories, and call for a reversal of the recent trend towards reduced reliance on urinalyses performed by nephrologists. This study might also spur clinical laboratories to improve their practices by reducing delays between collection and analysis of urine, and to train personnel to correctly identify different cell types.