Page 328 - Glucose Monitoring Devices

From prototype to product: the MiniMed 670G system 335

impacted by algorithm-driven insulin delivery. Furthermore, meal preparation in a
metabolic kitchen could ensure the nutritional content of meals was captured, or,
in some instances, the duration and intensity of exercise. A clear picture emerged
from these trials, namely that CL insulin delivery led to an increase in the percent
time participants remained in the target range, with exquisite glycemic control noted
in the overnight period. These trials involved a variety of patient populations,
including adults [38,41e54], children [41,55], adolescents [41,43e45,55e59],
and even pregnant women [60].

From transitional environments to tests at home
Thanks to the safety of these systems in inpatient research facilities, the next step
involved the simulation of a home environment. For children and adolescents, the
camp setting provided a perfect construct in which to assess the safety and efficacy
of automated insulin delivery [61e70], while in adults hotel-based studies were uti-
lized [70e73]. In addition, the same findings held truedthe devices were safe and
demonstrated an increased amount of time in the target range, paving the way for
outpatient home-based studies.
The research setting moved to free-living home assessments of devices. Some
studies were performed overnight only [74e83], while others opted for daytime
and overnight evaluations of the systems [80,81,84e91]. Furthermore, a wide
range of patient populations have been studied, including young children [90,92],
adolescents [69,74,76e78,80,84,89,91,93e95], and adults [74e76,78,80e85,87,88,
94,96e98]. The picture painted by each of these trials demonstrates the benefits of
closed-loop insulin delivery with increased time in range, especially in the overnight
period.

From prototype to product: the MiniMed 670G system
Studies of the algorithm driving insulin delivery in the 670G system have been con-
ducted for over 10 years’ time. Early assessment of the algorithm in 10 adult partic-
ipants in a research unit-based study demonstrated the feasibility of using this
approach with an increase in time in target range by 12% [39]. Yet, postprandial
hyperglycemia persisted, leading to the adoption of a hybrid approach that was first
tested by Weinzimer and colleagues [40].
Transitioning to outpatient studies, prototype equipment was first used in an
overnight camp study [64] as well as a 24-h outpatient home-based study with
remote monitoring of the system [84]. Both studies demonstrated an increase in
the time spent in the target range with the closed-loop controller. Subsequently,
the feasibility of a pump incorporating a closed-loop algorithm was first assessed
in a camp setting [65]. Although that study failed to demonstrate an improvement
in glycemic control, it supported further testing of the integrated platform. This