Page 333 - Glucose Monitoring Devices
P. 333

340    CHAPTER 16 The dawn of automated insulin delivery




                            Other challenges that remain are not unique to closed-loop insulin delivery.
                         Whenever insulin is infused in the subcutaneous tissue there is a risk of infusion
                         set failures, which can lead to DKA if they go unrecognized. Algorithms that can
                         detect and alert users of infusion set failures would benefit all patients on insulin
                         pump therapy. Small studies investigating methods to tackle these issues have
                         been conducted [115e117].
                            For hybrid closed-loop systems to be adopted in toddlers and children with T1D
                         on very low insulin doses, there may be a need for diluted insulin that will allow for
                         fine-tuned insulin dose titrations in this extremely insulin sensitive population [118].
                         Previous studies using diluted insulin in hybrid closed-loop systems show promising
                         results, with reduced rates of hypoglycemia, a trend toward reduced glycemic
                         variability, and reduced interindividual variability in time to peak insulin action
                         with diluted insulin [119,120]. Yet, a more recent study could not corroborate this
                         in an outpatient trial where participants were randomized diluted or standard
                         concentration insulin [90]. Importantly, dosing frequency in that trial was in
                         15 min intervals, not the 5 min intervals that some systems use.
                            The duration that sensors and infusion sets can be worn can also pose additional
                         hassle to those with T1D. Extended wear of both components would decrease the
                         frequency of diabetes-related interruptions that occur. Finally, the person with
                         diabetes must not be forgotten in these technological advancements. Indeed, to
                         have a device integrate seamlessly into one’s life would require systems to allow
                         users to push as few buttons as possible, use their existing technology like their
                         cell phones to administer bolus insulin, and the physical footprint of these devices
                         should be minimized.


                         A bright future
                         Although tremendous progress has been made in the creation of closed-loop systems
                         and commercial availability of the first-generation hybrid closed-loop was a major
                         step, much remains to be done. Numerous academics, industry partners, and even
                         online communities seek to create more sophisticated automation of insulin delivery
                         (Table 16.1). Indeed, recent data from a randomized control trial of 14e71-year-old
                         participants using the Tandem Control IQ system showed that time in the range
                         between 70 and 180 mg/dL increased by 2.6 h (70% vs. 59%). Users were extremely
                         satisfied with the system based on technology acceptance questionnaires, had
                         minimal fingerstick (<0.5 per/day) throughout the study as the system uses the
                         Dexcom G6, and used automated insulin delivery 92% of the time. FDA approval
                         of the device is being sought.
                            Although commercial approval of new hybrid closed-loop systems is sought, a
                         movement by patients has created a Do it Yourself (DIY) diabetes community. Start-
                         ing with the creation of The Nightscout Project, when persons with diabetes and
                         their loved ones created methods to allow remote monitoring of sensor glucose
                         data, a strong online group has emerged. Moving from the viewing of sensor data
                         to creating insulin delivery algorithms and providing instructions to others on how
                         to create such DIY closed-loop systems has exponentially increased the community
   328   329   330   331   332   333   334   335   336   337   338