Evaluation of determination of glucose in urine with some commercially available dipsticks and tablets.

Four commercial products for urine glucose determination were evaluated and compared with a quantitative hexokinase procedure. We examined precision, sensitivity, and analytical recovery of glucose from glucose-supplemented urine samples and comparison of methods, using patients' samples. Only "Chemstrip uG" (Bio-Dynamics Inc.) could differentiate between 0.3 g/L (upper limit of normal) and 0.6 g/L urine glucose concentrations. "Tes-Tape" (Lilly) and "Diastix" (Ames) gave positive readings at 0.3 g/L; "Clinitest" (Ames) detected glucose only over 1 g/L. Analytical recovery of glucose was best, for all four products, between 1 and 2.5 g/L; Chemstrip uG was the most nearly accurate among the four. Between 5 and 20 g/L glucose concentrations, Tes-Tape, Diastix, and Clinitest tended to give falsely low results; the use of Chemstrip uG resulted in overestimates of concentration at 20 g of glucose per liter. Only Chemistrip uG and Clinitest (two-drop method) had linear ranges extending to 50 g/L; Chemstrip uG had better precision and accuracy at this concentration. Of the four products, Chemstrip uG had the lowest within-technologist and technologist-to-technologist random analytical errors. In method comparison on patients' samples, Chemstrip uG was significantly stronger in its association with the quantitative hexokinase method than was Diastix, Clinitest, or Tes-Tape.