This week Medtronic released data from its 180,000 users of the Minimed 670G, touting impressive outcomes with the automatic, closed-loop system.
The MiniMed 670G, alternately called the artificial pancreas, uses a continuous glucose monitor (CGM) and an insulin pump in tandem. While users still need to input their own boluses for meals, the system will otherwise adjust the basal rate as it sees fit in order to keep blood sugar within a certain range. The promise is of a virtually hands-off system, automating the details of blood sugar management and freeing the user to go about their day (and night) with less stress and fuss.
Medtronic reports that its users spent 71% of their time in the range of 70-180 mg/dL, across all age groups. The number is difficult to evaluate without a baseline to compare it to, as nobody has any idea how frequently its users were in the same range before they began using the system. The company further reports that users who engaged the advanced features of the 670G were in range about 10% more often than those that did not, and experienced significantly fewer hypo and hyperglycemic events. This is unsurprising, as it appears that the manual mode is at least sometimes triggered specifically by stubborn lows or highs which the system cannot solve without intervention on the part of the wearer.
With the rise of the CGM, time in range (as with standard deviation) has become an increasingly important goal for people with Type 1 diabetes. There is still a paucity of study on these statistics, and it may be many years before we have academically rigorous benchmarks for time in range the way we do today for A1C. Is 71% good? Compared to what? But in a world where a significant majority of people with T1D are not meeting the American Diabetes Association benchmarks, the system is surely an improvement for tens of thousands of happy Medtronic customers.
The attraction of a hands-off system is obvious and the product has been well-received. An early user interviewed in this space claimed: “There is no way I would ever go back to any system that isn’t dictating my blood sugars for me. It would be like going back to having a horse and carriage when you could drive a car.” Other companies have promised to bring their own systems to market soon, and it’s easy to imagine a future where such systems represent the standard of care.
The ASweetLife team is a bit doubtful, however, about the characterization of 70-180 mg/dL as an “optimal” glycemic range. The system has strictly limited options for glycemic targets – 120 mg/dL is the lowest currently available. Many adherents to a low-carbohydrate approach find it possible to spend the vast majority of their day in an even tighter range, clustered around normal or near-normal glucose levels, and as such would likely find that the 670G system would actually make their control worse. And any evaluation that counts an 83 mg/dL and a 173 mg/dL as equally good has obvious flaws to contend with. Still, there is no question that the technology itself has marvelous potential, and we are hopeful that future models will allow for more aggressive automatic glucose management.
This release undoubtedly just scratches the surface of the data. We would be particularly interested to see more detail on the improvement of outcomes in adolescents, typically the age group that struggles the most with diabetes management – and are curious how well the system works with alternative diets, intensive exercise, and other variables. The data from 180,000 CGMs is a potential treasure trove.