1 Division of Pediatric Endocrinology, University of Michigan Medical School , Ann Arbor, Michigan.
2 USF Diabetes Center, Morsani College of Medicine, University of South Florida , Tampa, Florida.
Diabetes Technol Ther. 2018 Nov;20(11):731-737. doi: 10.1089/dia.2018.0209. Epub 2018 Oct 6.
The Medtronic predictive low-glucose management (PLGM) algorithm automatically stops insulin delivery when sensor glucose (SG) is predicted to reach or fall below a preset low-glucose value within the next 30 min, and resumes delivery after hypoglycemia recovery. The present study evaluated the PLGM algorithm performance of the MiniMed™ 670G system SmartGuard™ "suspend before low" feature in children aged 7-13 years with type 1 diabetes (T1D).
Participants (N = 105, mean ± standard deviation of 10.8 ± 1.8 years) underwent an overnight in-clinic evaluation of the "suspend before low" feature with a preset low limit of 65 mg/dL. After exercise, frequent sample testing (FST) was conducted every 5 min if values were <70 mg/dL; every 15 min if 70-80 mg/dL; and every 30 min if >80 mg/dL. First-day performance of the Guardian™ Sensor 3 glucose sensor and continuous glucose monitoring system was also evaluated.
Activation of the "suspend before low" feature occurred in 79 of the 105 participants, 79.7% (63/79) did not result in SG falling below 65 mg/dL. Mean glucose at activation was 102 ± 19 mg/dL and the initial insulin suspension duration was 87.5 ± 32.7 min. Four hours after insulin resumption, mean reference glucose was 130 ± 42 mg/dL. Mean absolute relative difference between the FST reference glucose and SG values on the first day of sensor wear was 11.4%. For the 26 participants in whom the "suspend before low" feature did not activate, none involved a reference glucose value ≤65 mg/dL, suggesting that the PLGM algorithm performed as intended.
In children aged 7-13 years with T1D, the "suspend before low" feature of the MiniMed 670G system demonstrated a hypoglycemia prevention rate of nearly 80% after exercise and did not involve rebound hyperglycemia. There were no events of severe hypoglycemia during the evaluation.
美敦力的预测性低血糖管理（PLGM）算法会在传感器血糖（SG）预测在接下来 30 分钟内达到或低于预设低血糖值时自动停止胰岛素输送，并在低血糖恢复后恢复输送。本研究评估了 MiniMed™ 670G 系统 SmartGuard™“低血糖前暂停”功能在 7-13 岁 1 型糖尿病（T1D）儿童中的 PLGM 算法性能。
参与者（N=105，平均年龄为 10.8±1.8 岁）接受了一次过夜门诊评估，设定低血糖下限为 65mg/dL。运动后，如果值<70mg/dL，则每 5 分钟进行一次频繁样本测试（FST）；如果值为 70-80mg/dL，则每 15 分钟进行一次 FST；如果值>80mg/dL，则每 30 分钟进行一次 FST。还评估了 Guardian™传感器 3 血糖传感器和连续血糖监测系统的第一天性能。
在 105 名参与者中，有 79 名（79.7%，63/79）激活了“低血糖前暂停”功能，SG 没有降至 65mg/dL 以下。激活时的平均血糖为 102±19mg/dL，初始胰岛素暂停时间为 87.5±32.7 分钟。胰岛素恢复后 4 小时，平均参考血糖为 130±42mg/dL。第一天佩戴传感器时，FST 参考血糖和 SG 值之间的平均绝对相对差异为 11.4%。对于 26 名未激活“低血糖前暂停”功能的参与者，没有参考血糖值≤65mg/dL，这表明 PLGM 算法按预期运行。
在 7-13 岁的 T1D 儿童中，MiniMed 670G 系统的“低血糖前暂停”功能在运动后表现出近 80%的低血糖预防率，且不会出现反弹性高血糖。评估期间没有发生严重低血糖事件。
