This study evaluated the TensorTip Combo glucose meter (CoG, CNOGA Medical, Cesarea, Israel), which measures blood glucose invasively (INV) and tissue glucose non-invasively (NI). The optical non-invasive device module is individually calibrated over a period of 3-4 days.
Standardized meal experiments were performed at baseline and endpoint (after three months of home use). During the meal experiments, glucose was assessed at 11 time-points resulting in 2729 comparator readings vs. the YSI Stat2300 reference method. A subgroup of patients used two devices analysis of precision. The statistical analysis included parameters of glycemic control (HbA1c, hypoglycemia etc.) and system accuracy (mean absolute relative difference; MARD and consensus error grid analysis; CEG).
The study was performed with 88 patients (male/female: 43/45, type1/type2: 24/64, HbA1c: 7.4±1.0%; Ethnicities: 40 Caucasian, 19 African-American, 12 Hispanic, 14 Asian). The observed glucose range was 53.5-399 mg/dL. A positive NI-MARD and a negative NI-MARD of 5.1% and 11.5% were observed, respectively (total NI-MARD: 16.6%). 99.3 % of the NI-data-points were seen in zones A+B of the CEG. No differences were seen in the meal test results between baseline and endpoint. The frequency of use of the NI module increased significantly (baseline NI/INV measurement ratio: 1.04 vs. 2.62 at endpoint, p<0.001). HbA1c remained stable (endpoint: 7.5±1.1%, n.s.), while frequency of hypoglycemia (invasive readings <70 mg/dL) was substantially reduced by 54% (p<0.01).
Using the CoG for 3 months at home resulted in an increased measurement frequency, and in a major improvement of glycemic control in patients with both types of diabetes.