Current generation continuous glucose monitoring devices continue to improve in accuracy and performance, judging from the results from a comparative study
"Continuous glucose monitoring can improve glucose control when used as an adjunct to intermittent self-monitoring of blood glucose and is a critical component of bionic pancreas devices for automated glucose control," Dr. Steven J. Russell reported during the annual scientific sessions of the American Diabetes Association. "The effectiveness of continuous glucose monitoring in both applications is critically dependent on the accuracy and reliability of the data produced."
In a previous comparative trial with 2,360 reference values, Dr. Russell of the department of medicine at the Massachusetts General Hospital Diabetes Center in Boston and his associates compared three continuous glucose monitoring (CGM) devices: the FreeStyle Navigator (Abbott Diabetes Care; approved in March of 2008), the Seven Plus (DexCom; received premarket approval in November of 2010), and the Guardian (Medtronic; approved in July 2006), which were worn simultaneously by six subjects with type 1 diabetes (Diabetes Care 2013;36:251-9). The FreeStyle Navigator had the best overall accuracy, with an aggregate mean absolute relative difference (MARD) in plasma glucose, relative to venous levels, of all paired points of 11.8%. The Seven Plus and Guardian produced aggregate MARDs of all paired points of 16.5% and 20.3%, respectively.
For the current comparative study with 4,657 reference values, the researchers compared three CGM devices: the G4 Platinum (DexCom; approved in October of 2012), Enlite (Medtronic; not yet available in the United States), and the FreeStyle Navigator, which were worn simultaneously over a 2-day period by 23 volunteers participating in a trial of blood glucose control with a closed-loop artificial pancreas. Every 15 minutes, the researchers obtained venous glucose measurements and paired them in time with the CGM measurements.
Dr. Russell reported that the G4 Platinum and the FreeStyle Navigator had similar accuracy, with an aggregate MARD of 10.8% and 12.3%, respectively, while the aggregate MARD for the Enlite was 17.9%. Data reporting percentages, which the researchers described as a "measure of reliability," were 99.7% for the G4 Platinum, 99.7% for the FreeStyle Navigator, and 97.1% for the Enlite.
However, Dr. Russell noted, CGM calibration errors remain a problem, one that if not resolved will prevent closed-loop systems from use outside a controlled research setting. For example, when the Navigator calibrated at a time when blood glucose was rising sharply, it overestimated glucose for the rest of the day. On the other hand, calibration corrected low readings by the Navigator in another instance. Compared with the earlier comparison study, "the G4 Platinum accuracy is markedly improved over the Seven Plus, and slightly better than the Navigator, and markedly better than the Enlite." However, for all the sensors, "calibration remains an important and, I think, preventable source of error."
The study was supported by the National Institutes of Health, and several research foundations. Technical and material support came from DexCom, Tandem Diabetes, SweetSpot Diabetes, International Biomedical, Hospira, Abbott, Insulet, and Medtronic. Dr. Russell has received research support from Abbott, Dexcom, Insulet, International Biomedical, and Medtronic, and MiniMed.