Inability of community-based laboratories to identify pathological casts in urine samples

To determine the accuracy of microscopic examination of urine from patients with known renal disease, performed by community-based laboratories.

Twenty-six urine specimens from 7 children with Alport syndrome, lupus nephritis, and IgA nephropathy were simultaneously submitted to 4 investigative sites: 2 community-based medical laboratories (laboratories A and B), and the offices of 2 nephrologists (nephrologists C and D). Participants A, B, and C were unaware of the nature of this investigation and blinded to the diagnosis associated with each specimen.

Twenty-six specimens from 7 children were analyzed. Pathological casts were identified in the 26 submitted specimens in the following order: 1 (4%) by laboratory A, 2 (8%) by laboratory B, 20 (77%) by nephrologist C, and 26 (100%) by nephrologist D. Four-way and 3-way (A, B, C) comparisons using the chi 2 test are significant at the P < .001 level.

The 2 community-based medical laboratories participating in this study did not accurately identify pathological casts in urine specimens routinely submitted to them. In addition to raising quality assurance issues, these misleading reports may result in inaccurate diagnosis and unnecessary laboratory and diagnostic investigations.