Glucose Measurements Using Blood Extracted from the Forearm and the Finger

Using the FreeStyle blood glucose monitor that measures glucose in a 0.3 μl blood sample, blood from the fingertip and blood from the forearm were measured for glucose concentration. There was a difference in glucose concentration between these sites. A linear regression of glucose measurements from the arm versus those from the finger in blood samples from diabetic patients taken at random times during the day gave an intercept of 19.4 mg/dL, a slope of 0.913 and a correlation coefficient of 0.956. Further studies indicated that changes in blood glucose are first detected in the finger sample and the changes lag in the arm measurements. This difference can be significantly reduced by rubbing the test site on the forearm to increase blood perfusion. Using this technique the correlation of arm to finger was nearly ideal with an intercept of -0.5 mg/dL, a slope of 1.027, and a correlation coefficient of 0.971. The accuracy of FreeStyle was assessed using venous blood versus YSI. The regression statistics were: intercept = 7.1 mg/dL, slope = 0.923, and correlation coefficient = 0.992. A Clarke Error Grid analysis yielded 99.6% of the measurements in Zone A and 0.6% in Zone B. Using capillary blood from the finger the regression statistics versus YSI were: intercept = 6.6 mg/dL, slope = 0.934, and correlation coefficient = 0.982. Zone A of the Clarke Error Grid contained 98.3% of the measurements with the remaining 1.7% in the B Zone. When arm measurements were compared to YSI measurements of finger capillary blood the regression statistics were: intercept = 9.0 mg/dL, slope = 0.945, and correlation coefficient = 0.967. Zone A of the Error Grid contained 87.7% of the measurements, Zone B 11.4%, and Zone D 0.8%. There would be very little difference in therapeutic decisions when the arm is used as the lancing site rather than the finger.