Improved Real-World Glycemic Control With Continuous Glucose Monitoring System Predictive Alerts

Abstract

Background: Most standalone real-time continuous glucose monitoring (RT-CGM) systems provide predictive low and high sensor glucose (SG) threshold alerts. The durations and risk of low and high SG excursions following Guardian™ Connect CGM system predictive threshold alerts were evaluated. Methods: Continuous glucose monitoring system data uploaded between January 2, 2017 and May 22, 2018 by 3133 individuals using multiple daily injections (MDIs) or continuous subcutaneous insulin infusion (CSII) therapy were deidentified and retrospectively analyzed. Glucose excursions were defined as SG values that went beyond a preset low or high SG threshold for ≥15 minutes. For a control group, thresholds were based on the median of the low SG threshold limit (70 mg/dL) and the high SG threshold limit (210 mg/dL) preset by all system users. During periods when alerts were not enabled, timestamps were identified when a predictive alert would have been triggered. The time before low horizon was 17.5 minutes and the time before high horizon was 15 minutes, of all users who enabled alerts. Excursions occurring after a low SG or high SG predictive alert were segmented into prevented, ≤20, 20-60, and >60 minutes. Results: Excursions were prevented after 59% and 39% of low and high SG predictive alerts, respectively. The risk of a low or high excursion occurring was 1.9 ( P < 0.001, 95% CI, 1.88-1.93) and 3.3 ( P < 0.001, 95% CI, 3.20-3.30) times greater, respectively, when alerts were not enabled. Conclusions: The predictive alerts of the RT-CGM system under study can help individuals living with diabetes prevent some real-world low and high SG excursions. This can be especially important for those unable to reach or maintain glycemic control with basic RT-CGM or CSII therapy.

Real-World Evidence and Glycemic Improvement Using Dexcom G6 Features

“…This suggests that this feature is especially useful for patients with limited vision; hands-free glucose monitoring may also be an attractive option for patients with limited dexterity. The beneficial effects of predictive alerts for hypoglycemia and hyperglycemia have been described for Medtronic's GuardianÔ Connect system, 12 and benefits of the G6 System's ULS alert have been described with respect to mitigation of hypoglycemia mitigation. 13 Alerts for hypoglycemia (whether existing or impending) may also reduce subsequent ''rebound'' hyperglycemia events, 14 perhaps because symptomatic hypoglycemia is a strong incentive for carbohydrate intake.…”

Hypoglycaemia detection and prediction techniques: A systematic review on the latest developments

“…(Medtronic) 52 and can reduce hypoglycaemic events by more than half. Connectivity of CGM systems, insulin pumps and activity wearables with smartphones gives the capacity to use multivariate algorithms and permits cloud-based complex calculations.…”

“…An observational study 51 showed that the activation of such predictive hypoglycaemia alarms reduced time below 3 mmol/L (54 mg/dl) by almost 40%. Another retrospective study 52 found that predictive alarms prevented 59% of low glycaemic values.…”

Advanced Diabetes Management Using Artificial Intelligence and Continuous Glucose Monitoring Sensors

“…The availability of dynamic information on glucose concentration from CGM sensors permits not only the detection of adverse events but also their prediction. The most recent CGM sensors adopt very simple prediction algorithms, e.g., linear extrapolation, to predict future interstitial glucose concentrations (e.g., 15–30 min in advance) in real-time and generate predictive alerts when a hypo/hyperglycemic event is foreseen [ 49 ]. If provided sufficiently in advance, predictive alerts allow the patient to take suitable countermeasures, e.g., preventive hypo-treatments or correction boluses, to mitigate, or even prevent, such critical episodes [ 50 ].…”