Glucose monitoring at the arm: risky delays of hypoglycemia and hyperglycemia detection.

OBJECTIVE We have examined whether rapid changes in blood glucose (BG) result in clinically relevant differences between capillary BG values measured at the forearm and the fingertip and whether local rubbing of the skin before blood sampling can diminish such differences. RESEARCH DESIGN AND METHODS Capillary BG samples were collected every 15 min for 3-5 h from the fingertip and the forearm of 17 insulin-treated diabetic patients and analyzed with different glucose monitors (FreeStyle, One Touch Ultra, and Soft-Sense). In a subgroup of patients (n = 8), local rubbing of the forearm skin was performed before blood sampling. A rapid increase in BG was induced by oral administration of glucose, and subsequently, a rapid decrease in glucose was induced by intravenous administration of insulin. RESULTS In the fasting state, the BG values at the fingertip and at the forearm were similar (7.8 +/- 2.4 vs. 7.2 +/- 2.3 mmol/l, P = 0.06). However, during rapid increase in glucose, BG values at the fingertip were consistently higher than at the forearm (maximal difference 4.6 +/- 1.2 mmol/l, P < 0.001). During rapid decrease in glucose, lower BG values were recorded at the fingertip (maximal difference to forearm 5.0 +/- 1.0 mmol/l, P < 0.001). At the forearm, BG was delayed by a median of 35 min (P < 0.01) in relation to the fingertip. Rubbing of forearm skin decreased the observed differences but with a large intraindividual and interindividual variability. There were no obvious device-specific differences. CONCLUSIONS To avoid risky delays of hyperglycemia and hypoglycemia detection, BG monitoring at the arm should be limited to situations in which ongoing rapid changes in BG can be excluded.