STATSTRIP® Glucose Monitor Performs Accurately in Hyperglycemia Ranges
Hatch, Matthew   (University of Colorado at Denver and Health Sciences Center)
Mentor: Nadeau, Kristen J. (University of Colorado)
Background: Increased prevalence of type 2 diabetes has increased testing of ??-cell function in patients by gold standard hyperglycemic clamp. This test requires accurate measurement of glucose concentration frequently to determine necessary glucose infusion rate. Current glucose measurement is performed by the YSI 2300, which, despite its accuracy has several drawbacks. Study Goal: Statstrip (SS) is a handheld glucose monitor that requires less blood, is less expensive, faster, and relatively low maintenance. Previously validated in euglycemic ranges, we wanted to verify the SS accuracy in hyperglycemic concentrations. Methods: During each clamp, samples were drawn every 5 minutes during two stages: 200 mg/dl (n=414) and 450 mg/dl (n=332). Samples were tested at the bedside in duplicate on the YSI using serum plasma and in triplicate on the SS using whole blood. A subset of these samples (stage 1 n=71, stage 2 n=34) were tested against lab standard UV Hexokinase to verify which of the two methods was more accurate. Comparisons were made between the methods directly and between the methods and the lab standard individually at both stages. Results: SS performed with similar accuracy to the YSI showing a bias of -2.0 during stage 1 and a bias of -11.6 during stage 2. Both SS and YSI had negative biases compared to UV Hexokinase during stage 1 (-8.4, -3.4, respectively) and during stage 2 (-47.7, -25.7, respectively). SS and YSI had very tight correlation over both stages (r=0.95, r=0.92). Both methods had precise measurements, with paired readings having a mean difference of less than 1.5 mg/dl. When comparing the two values against each other, coefficient of individual agreement was above acceptable during both stages. Conclusion: With YSI as the gold standard during hyperglycemic clamps, StatStrip performs closely enough to be an acceptable replacement in these studies. Although a larger bias exists compared to UV Hexokinase, this bias is also present in the YSI and therefore is acceptable. The largest setback of SS is higher variance in duplicate readings. This can be corrected for in practice with a third reading if the first two are further apart than would be acceptable. StatStrip can be used in hyperglycemic ranges in place of YSI without significant loss in accuracy.