Breton et al [52] |
12 adults, randomized crossover trial, 24-hour closed-loop experiments each with exercise
Exercise detection using HRa
Meal bolus manually calculated
|
Time in euglycemiab for APc with HR and without HR overall 81% vs 75%, exercise 91% vs 85%, and overnight 89% vs 84%
Using HR resulted in fewer hypoglycemic events during exercise (0 vs 2)
|
DeBoer et al [51] |
18 adolescents, randomized crossover trial, 24-hour closed-loop experiments each with exercise
Exercise detection using HR
Meal bolus manually calculated
|
Time in euglycemia for AP with HR and without HR overall 77% vs 74%, exercise 96% vs 87%, and overnight 92% vs 84%
Small reduction in hypoglycemic events (0.39 HR-informed AP vs 0.50 without HR)
|
Jacobs et al [44] |
21 adults, randomized crossover trial, 22-hour experiments each with exercise
Exercise-detection algorithm triggered manually
|
Time in euglycemia with exercise detection 67%, without exercise detection 72%, and SAPd 68%
Time in hypoglycemia (<3.9 mmol/L) 0.3%, 3.1%, and 0.8%, respectively
Time in hyperglycemia (<10 mmol/L) 32%, 25%, and 31%, respectively
|
Castle et al [45] |
20 adults, randomized crossover trial, 4-day experiments each with exercise
Exercise-detection algorithm triggered using wearable sensor in SHe and DHf controllers
|
Time in euglycemia overall SH 74.3%, DH 72%, PLGSg 65.2%, and current care 63.1%
Time in hypoglycemia 2.8%, 1.3%, 2%, and 3.1%, respectively
|
Turksoy et al [36,39] |
3 young adults, seven 32- or 60-hour closed-loop experiments with exercise
Additional signals integrated continuously
|
|
Turksoy et al [37] |
3 young adults, 70-hour closed-loop experiments with exercise
Additional signals integrated continuously
|
|
Turksoy et al [40] |
9 young adults, 2-day closed-loop experiments with exercise
Additional signals integrated continuously
|
|
Turksoy et al [41] |
10 young adults, eighteen 60-hour closed-loop experiments with exercise
Additional signals integrated continuously, with submodules
|
|