To the Editor.—Urinalysis is a useful tool in diagnosis, screening, and monitoring the progress of diseases.1 However, laboratory practice is inconsistent in the use of urine confirmatory tests for the presence of ketones and bilirubin in urine. Thus, we recently conducted a study and intended to standardize this issue.
Urine ketone level was measured using a nitroprusside reaction and bilirubin was measured by the diazo method on an Aution Max AX-4280 automated urine chemistry analyzer (Iris Diagnostics, Chatsworth, California). Urine ketones were confirmed with Acetest and bilirubin was confirmed with Ictotest (Bayer Corp, Elkhart, Indiana). Of 492 urine samples analyzed for ketones on the AX-4280, 42 (8.54%) were positive for ketones. Acetest confirmed 35 of the 42 ketone-positive samples as ketone positive (Table). Most of the discrepancies between urinalysis and Acetest results were seen at 10 mg/dL on the AX-4280. Of 4270 urine samples analyzed for bilirubin on the AX-4280, 108 samples were positive for bilirubin (2.53%) and 107 of those samples were confirmed to be bilirubin positive by Ictotest (Table).
A survey was also conducted showing that 13 of the 25 responding laboratories were still doing confirmatory testing for bilirubin by Ictotest (52%). Similarly, 2 of 8 responding laboratories were still using Acetest to confirm urine ketones.
The confirmatory testing was historically performed to confirm results obtained from the urine dipstick. Substantial improvements in the sensitivity and specificity of automated semiquantitative analyzers have dramatically reduced the necessity of confirmatory testing. We believe that there is no rationale for continuing these confirmatory tests, for the following reasons:
Currently there is no guideline on whether confirmatory testing needs to be done for urine ketones or bilirubin. The Clinical and Laboratory Standards Institute has stated that “many of the historical confirmatory chemical urinalysis tests such as the sulfosalicylic acid (SSA) test for protein, the tablet test for ketones, and the tablet test for bilirubin may not be relevant to current laboratory practice.” 2 The College of American Pathologists does not require these confirmations and suggests that users follow the manufacturer's recommendations.
Our survey findings suggest that most laboratories have discontinued or will soon discontinue these confirmatory tests.
Our data show that confirmatory tests results agree well with those from automated urinalysis with exception of the reading of trace ketonuria by Acetest.3 Acetest itself is less sensitive and more subjective compared with automated urinalysis using the AX-4280.
Detection of urine ketones by confirmatory testing does not change clinical practice or help in clinical decision making. The National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines 2011 state that “urine ketone measurements should not be used to diagnose or monitor the course of diabetic ketoacidosis.” 4
In summary, using the same or a similar qualitative methodology to confirm another qualitative method duplicates work, increases costs, and may provide misleading information. Thus, such confirmatory tests should be discontinued. Confirmatory and complementary tests such as serum bilirubin and ketones (in particular β-hydroxybutyrate) should be ordered when clinically warranted.
The authors have no relevant financial interest in the products or companies described in this article.