HOW TO SUCCESSFULLY ON-BOARD PATIENTS WITH TYPE 1 DIABETES FROM MULTIPLE DAILY INJECTION TO HYBRID CLOSED LOOP SYSTEM?
Abstract
Background and Aims
The aim of this study was to evaluate the effectiveness of a 10-day structured protocol to on-board patients from Multiple Daily Injection (MDI) to the MiniMed 670G system in achieving glycaemic control.
Methods
An open single-arm, single-center, clinical investigation in children aged 7 to 18 years on MDI following a structured protocol: 2 days - selection and introduction; 5 days - Minimed 670G system training (2-hours session in 5 consecutive days with groups of 3 to 5 patients and families); 3 - days in Manual Mode; 84 days in Auto Mode, cumulating in 10 days from MDI to Auto Mode activation of the MiniMed 670G system.
Results
30 children (age 10.24 ± 2.6 years) were enrolled in the study and completed the planned 84 days on Auto Mode. The participants used the sensor for a median of 89% of the time and spent a median of 92 % in Auto Mode. The mean HbA1c decreased from 8.2±1.4% at baseline to 6.7±0.5% at the end of the study phase (p=0.017). Time in Range (70-180mg/dL) increased from 46.9±18.5% at baseline to 75.6±6.9% in Auto Mode (p<0.001). This was achieved while spending 2.8% below 70 mg/dL and without any severe hypoglycemia nor DKA.
Conclusions
Patients with T1D on MDI therapy on-boarded to the MiniMed 670G system, using a structured 10-day protocol, while significantly improved their glycemic control in a safe manner.
REAL WORLD IMPROVEMENTS IN HYPOGLYCEMIA IN AN INSULIN-DEPENDENT COHORT PRE AND POST TANDEM BASAL-IQ TECHNOLOGY REMOTE SOFTWARE UPDATE
Abstract
Background and Aims
Software updatable insulin pumps, such as Tandem’s t:slim X2 pump, are available in the United States, enabling pump users to access new technology as soon as it is commercialized. Little to no quantitative data exists on the remote software update process, which allows for minimal interruption in therapy as compared to purchasing a new pump, nor on pre/post therapeutic outcomes from those who remotely update their pump. We examined real-world usage and impact of a software updatable predictive low-glucose suspend (PLGS) algorithm designed for maintenance of euglycemia and reduction of hypoglycemic events in people with insulin-dependent diabetes.
Methods
Approximately 14,000 Tandem pump users remotely updated their software to Basal-IQ® technology (PLGS) since its commercial release. We performed a retrospective analysis of users who uploaded at least 21 days of pre- and post- PLGS software update usage data to the Tandem t:connect® web application between August 28, 2018 and October 1, 2019. Insulin delivery and sensor-glucose concentrations were analyzed per recent international consensus guidelines. Time taken to perform the software update was also assessed.
Results
Median software update time was 321 seconds, or 5.35 minutes. Glycemic outcomes pre and post software update will be presented.
Conclusions
Introduction of a software updatable PLGS algorithm was easily performed remotely by Tandem t:slim X2™ insulin pump users and resulted in effective and sustained reduction of hypoglycemia.
HCL USE DECREASES SIGNIFICANTLY ACROSS TIME IN YOUTH WITH T1D
Abstract
Background and Aims
The aim of this study was to describe HCL use in a clinical sample of youth using the 670G and identify an HCL use threshold associated with meeting A1c target of 7.5% at 12 months
Methods
Youth starting the 670G HCL system for T1D care participated in a 12-month observational study. Data on HCL use and glycemic outcomes were obtained from pump downloads and chart review during routine clinic visits in the first 12 months after starting HCL.
Results
Ninety-three youth (14.7±3.7 yrs.;50% M;A1c 8.8±1.7%) with T1D for 6.8 (4.6,10.0) yrs. started HCL. HCL use was 69% 1 month after starting HCL, which declined to 52% at month 6 and 47% at month 12 (p<0.001). Sensor use declined from 78% at month 1 to 65% at month 6 and 55% at month 12 (p<0.001). Sensor TIR (70-180 mg/dL) declined from 61% at month 1, to 56% at month 6, and 54% at month 12 (p<0.05). Forty-six youth (49%) discontinued HCL within the first 12 months of use. ROC analysis identified 69% as the HCL use threshold associated with A1c ≤7.5% (AUC=0.76;specificity=63%;sensitivity= 84%).
Conclusions
HCL use ≥69% may be a clinical goal to increase likelihood of meeting A1c target of 7.5% in youth with T1D using 670G HCL. However, the decline in HCL and sensor use across time, and 50% HCL discontinuation rate, suggests youth struggle to sustain HCL use. Future developments in HCL technology should prioritize ease of HCL use to prevent discontinuation and ensure the benefits of HCL technology are realized in the real-world.
LOOP OBSERVATIONAL STUDY: EVALUATING DO-IT-YOURSELF (DIY) AUTOMATED INSULIN DELIVERY
- John Lum, United States of America
- Victoria Barnes-lomen, United States of America
- Ryan Bailey, United States of America
- Diana Naranjo, United States of America
- Korey Hood, United States of America
- Rayhan Lal, United States of America
- Daniel Desalvo, United States of America
- Jeremy Pettus, United States of America
- Peter Calhoun, United States of America
- Roy W. Beck, United States of America
Abstract
Background and Aims
Loop is a DIY app for automated insulin delivery using an iPhone and commercial continuous glucose monitor (CGM) and insulin pump. This study evaluates glycemic control, adverse event rates, and patient-reported outcomes (PROs) among individuals with type 1 diabetes (T1D) using Loop.
Methods
Two groups were enrolled in this ongoing observational study: participants using Loop at enrollment (Existing Users) and those planning to start Loop (New Users). Device data are collected via the Tidepool Mobile App, including available CGM data at enrollment and throughout follow-up. PRO data are collected at baseline, 3, 6, and 12 months. Adverse events (severe hypoglycemia, diabetic ketoacidosis, and hospitalizations) are monitored weekly.
Results
Among 875 participants, mean age was 27±17 years (range 1-77, 42% <18 years old), 55% were female, 93% were Caucasian, 85% of participants (households, if <18 years old) had a bachelor’s degree or higher, and median diabetes duration was 12 years (IQR: 5-24). Enrollment included 266 Existing Users (44% using Loop for >12 months) and 609 New Users. Existing Users mostly had Medtronic pumps (85%, versus 15% OmniPod), while New Users mostly used OmniPod (91%). Baseline glycemic metrics at enrollment are reported in the table; for Existing Users, CGM time-in-range 70-180 mg/dL was 81%, with median 0.44% time <54 mg/dL.
Conclusions
Participants using Loop prior to enrollment had high time-in-range at baseline with little hypoglycemia. Three-month follow-up will be completed in December 2019, and efficacy and safety data will be presented as well as PROs.
ACCESS ISSUES IN DIABETES TECHNOLOGY: THE CASE OF DO-IT-YOURSELF ARTIFICIAL PANCREAS SYSTEMS
Abstract
Background and Aims
Technology-enabled solutions have been discussed for many years as one of the most promising routes to improving clinical outcomes and reducing the burden associated with diabetes self-management. Growing impatient with both the direction and pace of industry-led innovation, people with diabetes (PwD) are increasingly turning to technology-enabled solutions which are off-label, user-driven and open-source, such as the “Do-it-Yourself Artificial Pancreas Systems” (DIYAPS).
Methods
Drawing on quantitative and qualitative data generated by the DIWHY survey which examines the motivations, barriers and retention factors of DIYAPS users (n=1058 from 34 countries), the aim of this study is to gain insights into the barriers and enablers to the wider adoption of DIYAPS.
Results
Study participants typically described two main challenges in their ’looping journey’: 1) accessing requisite components, such as CGM and compatible insulin pumps, and 2) technical challenges and perceptions associated with building and maintenance of a DIY closed-loop system. In both cases, the capacity to draw on social connections developed through participation in online communities as the Facebook group ‘Looped’ was key to overcoming many of these challenges. For example, many participants described how their (self-perceived) limited IT literacy was often overcome through the online advice and support of more ‘tech-savvy’ community members and how they often felt an enhanced sense of agency and empowerment after successfully closing the loop for themselves.
Conclusions
Social connectedness is a key driver in the uptake of DIYAPS solutions.
GLYCEMIC OUTCOMES OF USE OF CLC VS PLGS IN TYPE 1 DIABETES (T1D): A RANDOMIZED, CONTROLLED TRIAL
- Sue A. Brown, United States of America
- Dan Raghinaru, United States of America
- Bruce A. Buckingham, United States of America
- Yogish C. Kudva, United States of America
- Lori Laffel, United States of America
- Carol J. Levy, United States of America
- Jordan E. Pinsker, United States of America
- R. Paul Wadwa, United States of America
- John Lum, United States of America
- Roy W. Beck, United States of America
- Boris Kovatchev, United States of America
Abstract
Background and Aims
This study compares glycemic outcomes of a closed-loop control (CLC) and a predictive-low glucose suspend (PLGS) system that use the same insulin pump and CGM.
Methods
Participants with T1D used CLC for 6 months in a multicenter trial and then were randomly assigned to CLC (Control-IQ) or PLGS (Basal-IQ) for an additional 3 months. Primary outcome was time in range (TIR, 70-180mg/dL). Baseline comparison was the last 3 months of the preceding study.
Results
109 participants (mean age 33 years, glycated hemoglobin 7.1%) were randomized to CLC (N=54) or PLGS (N=55). The mean±SD TIR in the CLC group changed from 71±11% to 68±13% and decreased in the PLGS group from 70±14% to 60±17% from baseline to 13 weeks (CLC-PLGS treatment difference of 6% [95%CI: +4,+8;p<0.001]). When excluding a ~one-month suspension of CLC use due to a device error, TIR changed from 72±11% at baseline to 69±12% at 13 weeks for the CLC group. Hyperglycemia (time >180mg/dL) was lower in the CLC group (treatment difference of -6.0% [95%CI:-8.4,-3.7;p<0.001]). There was no significant difference in hypoglycemia (time <70mg/dL) between CLC and PLGS. Glycated hemoglobin was lower in the CLC group with mean difference of -0.34% (95%CI -0.57,-0.11;p=0.0035) with CLC group 7.05% to 7.18% and PLGS group 7.06% to 7.53% from baseline to 13 weeks.
Conclusions
Following 6 months of CLC, additional 3 months of closed-loop maintained TIR and HbA1c while on PLGS these metrics rebounded towards their pre-CLC values, with hypoglycemia remaining reduced with both CLC and PLGS.
IMPROVED GLYCAEMIC CONTROL AFTER TRANSITION TO THE HYBRID-CLOSED-LOOP (HCL) SYSTEM MINIMED 670G – REAL-WORLD EXPERIENCE OF A TERTIARY REFERRAL CENTRE
Abstract
Background and Aims
Improved glycaemia in patients using the hybrid-closed-loop system MiniMed 670G (MM670G) has been demonstrated in clinical trials, but only limited real-world data is available yet. The aim of the present study was to analyse glycaemic control after the transition to the new MM670G and to compare it to the preceding treatment.
Methods
This was an 8-month analysis of 25 patients. A structured training program was set up for the transition to MM670G. CGM data were downloaded from the proprietary manufacturer’s software, analyzed with the Glyculator-2 script and compared to pre-transition CGM-data (30-day periods) if available.
Results
Average operation time of CGM-sensor after transition to MM670G was 78%, and mean time in auto mode was 75%. HbA1c was 6.9±1.1%, mean glucose 8.2±0.8mmol/L, CV 31.4±5.8%, time in range (TIR, 3.9-10mmol/L) 76.1±11.6%, time > 10mmol/L 21.6±11.5% and time < 3.9mmol/L 2.3±2.2%, respectively. HbA1c significantly decreased using MM670G (7.4±1.5% vs 6.9±1.1%, p=0.002). Paired CGM data before and after transition to MM670G was available for 14 patients (all prior MM640G users). TIR was higher after change to MM670G (77.4±11.7% vs 70.0±14.5%, p=0.024), whereas time > 10.0mmol/L was lower (19.6±11.3% vs 26.4±15.3%, p=0.013). Coefficient of variation [CV] and interday-variability (MODD) were lower under MM670G compared to the preceding treatment (CV 32.7±6.2% vs 35.6±6.2%, p=0.013; MODD 2.7±0.8 vs 3.2±1.1mmol/L, p=0.04; respectively).
Conclusions
After switching to the new MM670G system, glycaemic control significantly improved in an already well-controlled cohort of diabetic patients. This improvement is reflected in particular by a higher time in range, decreased HbA1c and reduced glycemic variability.
EFFECTIVE TECHNOLOGY TRAINING IS A KEY COMPONENT TO CONTINUED USE OF MEDTRONIC 670G
Abstract
Background and Aims
Use of the first approved hybrid closed loop device in type 1 diabetes (T1D) is linked to improved glycemic outcomes. However, discontinuation of Automode is high due to patient reported usability challenges. Our aim is to assess the effect of patient training on Medtronic 670G use.
Methods
Data were obtained from a retrospective chart review of T1D pediatric patients who attended at least one training session for the 670G system between 2016-2018. Twelve-month average HbA1cs before and after the final training date were obtained in addition to ‘full’ (3 out of 3 sessions) or ‘partial’ (1-2 sessions) training. Data were analyzed with Fisher’s exact and paired Wilcoxon signed rank tests.
Results
Twenty-five subjects (mean±SD) 17.2±3.5 years old with T1D duration 9.1±4.7 years underwent training for pump, CGM, and/or Automode initiation; 13/25 (52%) completed full training; 9/25 completed two, and 3/25 only one. All subjects (100%) with full training and 33% with partial training were using Automode at the time of analysis (p=0.0005). There were no significant differences in average HbA1c from pre- to post-Automode between groups (p=0.8548).
Conclusions
Subjects with complete training were more likely to be using Automode. HbA1c did not differ between subjects who received full or partial training. This may be due to the small sample size and wide variability of HbA1c values. Future direction will include assessment of time in range, sensor wear, and user satisfaction to guide sustainable integration of diabetes education and technology training and its relationship to improved care.