Albuminuria and proteinuria in hospitalized patients as measured by quantitative and dipstick methods

We tested patients' urines for albumin, protein, and creatinine by quantita tive and dipstick methods. The concentrations of these analytes were establ ished by quantitative, cuvet-based chemistry methods that we assumed gave t he "correct" values. There was good to excellent agreement of the dipstick results with the quantitative methods for the above three analytes. We foun d many patients who excreted pathological amounts of albumin and/or protein who did not have a diagnosis of kidney disease or other likely causes of p roteinuria, suggesting that albuminuria and/or proteinuria were underdiagno sed in our group of patients. Those with cardiovascular disease, kidney dis ease, or diabetes showed the greatest predictive value of a positive test f or albumin or protein by dipstick. Dipstick testing for albumin, protein, a nd creatinine had good or excellent agreement with quantitative methods. Th e dipstick tests were easy to use, simple, and low in cost, and can serve w ell for point-of-care testing. (C) 2001 Wiley-Liss, Inc.