Urine glucose testing in the critically ill: a comparison of two enzymatic test strips.

Abstract

The urine glucose concentration is commonly used to monitor indirectly the degree of hyperglycemia in critically ill patients and to adjust insulin dosage. Most commercially available urine glucose reagent test strips measure the urine glucose concentration from 0% to 2%. When the urine glucose is at the 2% level, the blood glucose concentration may vary over a wide range. We compared a new urine glucose test strip which measures the urine glucose concentration from 0% to 5% versus a conventional strip (0% to 2%) in the analysis of double-voided urine specimens from 285 patients with diabetes mellitus. Both types of test strips were insensitive in detecting hyperglycemia and showed a wide range of blood glucose values for each estimated urine glucose concentration. However, the new test strips which gave measurements at the 3% and 5% urine glucose concentrations allowed for more specificity (99%) in detecting blood glucose levels above 250 mg/dl. We conclude that: (a) test strips measuring from 0% to 5% are superior to conventional 0% to 2% test strips because the 3% and 5% urine glucose readings allow for a high level of specificity in detecting severe hyperglycemia (greater than 250 mg/dl); (b) urine glucose testing is insensitive and nonspecific in detecting hyperglycemia when urine glucose values are 2% or less.