Accuracy and Safety of Dexcom G7 Continuous Glucose Monitoring in Adults with Diabetes

Abstract

Background: We evaluated the accuracy and safety of a seventh generation (G7) Dexcom continuous glucose monitor (CGM) during 10.5 days of use in adults with diabetes. Methods: Adults with either type 1 or type 2 diabetes (on intensive insulin therapy or not) participated at 12 investigational sites in the United States. In-clinic visits were conducted on days 1 or 2, 4 or 7, and on the second half of day 10 or the first half of day 11 for frequent comparisons with comparator blood glucose measurements obtained with the YSI 2300 Stat Plus glucose analyzer. Participants wore sensors concurrently on the upper arm and abdomen. Accuracy evaluation included the proportion of CGM values within 15% of comparator glucose levels >100 mg/dL or within 15 mg/dL of comparator levels ≤100 mg/dL (%15/15), along with the %20/20 and %30/30 agreement rates. The mean absolute relative difference (MARD) between temporally matched CGM and comparator values was also calculated. Results: Data from 316 participants (619 sensors, 77,774 matched pairs) were analyzed. For arm- and abdomen-placed sensors, overall MARDs were 8.2% and 9.1%, respectively. Overall %15/15, %20/20, and %30/30 agreement rates were 89.6%, 95.3%, and 98.8% for arm-placed sensors and were 85.5%, 93.2%, and 98.1% for abdomen-placed sensors. Across days of wear, glucose concentration ranges, and rates of change, %20/20 agreement rates varied by no more than 9% from the overall %20/20. No serious adverse events were reported. Conclusions: The G7 CGM provides accurate glucose readings with single-digit MARD with arm or abdomen placement in adults with diabetes. Clinicaltrials.gov: NCT04794478.

Keywords: Accuracy; Continuous glucose monitoring; Dexcom; G7; MARD.

Conflict of interest statement

S.K.G. reports consultant fees from Medtronic, Novo Nordisk, Zealand, Lifescan, Roche, and Lilly; and research grants through the University of Colorado from Lilly, Novo Nordisk, Medtronic, Dexcom, T1D Exchange, Helmsley Trust, NIDDK, and JDRF. M.K. reports research support from Medtronic, SARO, Pfizer, Dexcom, Abbott, Regenacy, Metacrine, Lumos, Ascendis, Sagimet, MannKind, Aeterna-Zentaris, Bio89, Gilead, Senseonics, KOWA, Allergan, NGM, Tolerion, Lilly, Vertex, and Diamyd; and has a consulting agreement with Quest Diagnostics.

K.C. receives research support provided to her institution from Dexcom, Abbott Labs, Abbott Diabetes, Medtronic, Eyenuk, Laxmi, Novo Nordisk, and Insulet and has received consulting fees from Dexcom. T.S.B. reports research support from Abbott Diabetes, Abbott Rapid Diagnostics, Biolinq, Capillary Biomedical, Dexcom, Eli Lilly, Kowa, Livongo, Mannkind, Medtronic, Novo Nordisk, REMD, Sanofi, Sanvita, Senseonics, Viacyte, vTv Therapeutics, and Zealand Pharma; consulting honoraria from Abbott, CeQur, Lifescan, Mannkind, Medtronic, Novo, and Sanofi; and speaker honoraria from BD, Medtronic, and Sanofi.

H.K.A. received research grants through the University of Colorado from Senseonics, Dexcom, Eli Lilly, REMD Biotherapeutics, IM Therapeutics, Institute for the Advancement of Food and Nutrition Sciences, and Mannkind; and honorarium from the American Diabetes Association for speaking and consulting. M.P.C., on behalf of Diablo Clinical Research, reports research support from Abbott Diabetes Care, Biolinq, Dexcom, Medtronic, SanVita, Senseonics, Abbott Point of Care, Ascensia, Novo Nordisk, Merck, Boehringer Ingelheim, and Helixmith. S.A.B. reports research support through the University of Virginia from Tandem Diabetes Care, Insulet, Tolerion, Roche Diagnostics, and Dexcom. J.B.W., A.K.B., J.L.R., and S.E.B are employees and stockholders of Dexcom, Inc.

Figures

FIG. 1.

G7 components (left to right) include an optional receiver and display unit, packaging, the wearable sensor/transmitter, and the applicator. The wearable measures 24 × 27.3 × 4.6 mm, roughly the size of three stacked quarters. G7, seventh generation.

FIG. 2.

Aggregated sensor MARD (%) of CGM-YSI histogram plot for 619 sensors placed on the arm or abdomen. CGM-YSI; MARD, mean absolute relative difference.