897-P: Real-World Response to Real-Time Continuous Glucose Monitoring by Caregivers of Children with Type 1 Diabetes

Abstract

The use of real-time continuous glucose monitoring (CGM) is on the rise in children with type 1 diabetes (T1D), but little is known about how caregivers respond to high and low sensor glucose (SG) values in a real-world setting. We previously reported results from a cross-sectional survey, distributed throughout the United States, of 169 caregivers of children with T1D using CGM. In the current study, 21 of these participants, who were also using insulin pumps, voluntarily contributed their device data anonymously through the Tidepool platform. Up to 90 days of CGM and pump data per participant were analyzed to determine how caregivers respond to SG levels below and above target range. The children had a mean age of 8.8 ± 3.0 years, were 60% female, were 15% ethnic minority or mixed ethnicity, and had an average hemoglobin A1c of 7.2% ± 0.8%. For these participants in a 90-day period, there were a total of 453,908 CGM values, or a mean of 7205 values per 30 days per participant (mean 2263 to 14,323 values per 30 days for each participant). The average SG was 171.7 ± 29.9 mg/dL, with 57.8% of sensor values between 70-180 mg/dL. The percentage of values in hypoglycemic range (<70 mg/dL) was 2.9% and the percentage in hyperglycemic range (>180 mg/dL) was 39.3%. Per 30 days, participants had means of 26.6 ± 30.3 hypoglycemic and 73.7 ± 28.6 hyperglycemic events lasting at least 15 minutes. Glucose values were rarely entered into pumps within 5 minutes of having SG values outside of target range for at least 15 minutes and were entered for 2.7% ± 5.3% of hypoglycemic and 3.0% ± 3.5% of hyperglycemic episodes. In response to high SG lasting at least 15 minutes, an insulin bolus was given 4.7% ± 4.0% of the time within 5 minutes and 59.3% ± 36.8% of the time within 60 minutes. The variable response to hypo- and hyperglycemia suggests further interventions that decrease the need for caregivers to perform active tasks in response to CGM data may help ease management burden and improve time in target range. Disclosure H. Howell: None. B.J. Yeung: None. A. Nguyen: None. J.C. Wong: Consultant; Self; pH Pharma. Research Support; Self; Dexcom, Inc. Funding Dexcom, Inc. (IIS-2017-048)