The reliability of urinary albumin assay by immunonephelometry in the clinical practice. The critical role of medical surveillance on laboratory routine.

Measurement of the urinary albumin excretion rate (UAER) is essential for the early diagnosis and monitoring of diabetic nephropathy; immunonephelometry is a procedure used worldwide for routine screening of diabetic patients. Since we have met with occasional inconsistent values of UAER in serial urine collections, we searched for possible sources of analytic error. To assess the best working conditions of the instrument in use, the stability of urine samples during storage and the need for previous urine centrifugation, we assayed repeatedly the six automatically diluted points of the standard curve (55.6 to 1.7 mg/l), four control samples of human albumin in saline (100 to 1 mg/l) and 24-h urine collections from outpatient diabetic subjects. The last were also assayed with and without previous centrifugation, and both immediately after collection as well as after storage at -20 degrees C for 7, 42, 79, 97, 128 and 161 days. We concluded that: (1) pre-analytic centrifugation of urine samples in unnecessary; (2) the intra-assay coefficient of variation (CV) of the standard curve changed from 2.4% to 9.3% when moving from the highest to the lowest concentration; the inter-assay CV changed from 4.1% to 14.4%, respectively; (3) the intra-assay CV of the control samples (manually prepared) changed from 5.7% to 10.2% and the inter-assay CV from 7.7% to 22.9%; there was a constant and significant (P < 0.01) underestimation (from -9% to -30%) of the obtained values compared with the expected concentrations; (4) a progressive decrease in recovered albumin by multiple freezing and thawing of urine samples did occur, which became significant after 161 days of storage. In the BNA workbook (menu 7.1, assay protocols), a 7-day validity of the reference curve is reported. Moreover, to economize, pre-dilution cuvettes were often recycled in our hospital central laboratory. We observed that: the intra-assay CV for urine samples was 79.4% with recycled cuvettes and stored standard curve, 11.3% with new cuvettes and stored standard curve, 4.9% with both new cuvettes and newly performed standard curve; the inter-assay CV was 32.6%, 10.5% and 6.4%, respectively. These data emphasize, from the laboratory viewpoint, the need for both accurate calibration of BNA and use of native urines; in addition, they stress the importance of careful supervision of laboratory routine and interpreting analytic results in the clinical setting.