Urinary total protein estimation--fact or fiction?

In two recent surveys of urinary total protein assays, 370 laboratories in the United Kingdom were requested to determine the protein content of a simulated 24-hour urine (a solution of sodium and potassium salts and urea, with no added protein) and of a 24-hour urine from a healthy individual. The nature of these specimens was not revealed and participants used their routine methods and calibrants. Quantitative results (range 0.005-12.23 g/l, median 0.03 g/l) were received from 31% of the participants for the salt solution and from 43% for the normal urine (range 0.01-2.96 g/l, median 0.05 g/l). Nonquantitative results, i.e. those given as less than a detection limit (range less than 0.005 to less than 0.5 g/l) were received from 29% of the participants for the salt solution and from 33% for the normal urine. Statements of 'nil', 'zero' or 'not detected' were received from the remainder. Further analysis of the results indicated that 29% of the laboratories reported, or did not unequivocally exclude, significant proteinuria in the salt solution, and 41% of the laboratories similarly did not exclude proteinuria in the normal urine. It is proposed, for both clinical and analytical reasons, that consideration be given to the discontinuation of urinary total protein estimation and that urinary albumin, supplemented where appropriate by other selected protein or enzyme measurements, be determined instead.