One hundred years after the discovery of insulin, technological advancements herald the dawn of a new era of automated insulin delivery systems for managing Type 1 diabetes (T1D).
Continuous glucose monitoring (CGM) devices with improved accuracy with reduced/no finger pricks are fully subsidised (federal initiative) and available to children and young adults up to the age of 21 years. These devices are inserted subcutaneously, last for 7-14 days and provide 5-minute glucose readings, with data available in real-time to the user to help make treatment decisions.
Continuous subcutaneous insulin infusion therapy involves constant delivery of rapid acting insulin through an insulin pump. This includes basal insulin as background continuous insulin and bolus insulin delivered with meals and corrections. There is private health insurance coverage.
Insulin pump settings for basal and bolus insulin are programmed into the pump, based on the individual’s total daily insulin requirement and/or weight. These require to be adjusted with regular review of glucose levels as insulin requirement is variable, subject to diet, exercise, stress, illness etc. Stand-alone pumps require the user to make diabetes-related decisions and have limited ability to adjust to individual needs.
CGM and/or insulin pumps have improved metabolic control, reduced glucose fluctuations, reduced time spent in hypoglycaemia and improved quality of life. In spite of these advancements, only a minority of children and adults achieve the recommended HbA1c target of <7%, highlighting the need for newer therapies and interventions.
Automated systems
Although CGM and insulin pump therapy were initially designed as stand-alone systems, further advancements incorporated algorithms to utilise CGM information (sensor glucose) to inform insulin delivery, paving the start of automated insulin delivery systems (Figure 1).
The first step was directed to minimise hypoglycaemia and algorithms were incorporated to suspend basal insulin delivery with hypoglycemia.
With Low Glucose Suspend, basal insulin delivery is suspended when sensor detects hypoglycaemia. The next step to prevent hypoglycaemia with Predictive Low Glucose Suspend switches off basal insulin delivery with prediction of hypoglycaemia. There are two systems currently available with predictive suspend: PLGM system (Medtronic 640G pump and its sensor) and Basal IQ (Tandem t:slim pump with Dexcom G6 sensor).
Further advancements increase automation, in the pursuit of a fully automated closed-loop system, also called artificial pancreas as it attempts to replicate the physiological insulin delivery dependent on ambient glucose levels in individuals.
Currently, available systems are semi-automated and referred to as hybrid closed-loop systems. Basal insulin delivery is dependent on sensor glucose levels and is therefore automated. However, the user manually boluses for meals and corrections. The Medtronic 670G was the first FDA/TGA approved system commercially available and has improved glycemic control.
Increasing trust in automation and user experiences has led to further advancements in algorithms with automated correction boluses and additional target set-points with improved user functionality. Closed-loop systems such as Control-IQ (Tandem t:slim with Dexcom G6) and the advanced algorithm in Medtronic 780G deliver auto-corrections and permit remote monitoring of glucose levels, an important feature for caregivers of children with T1D. These systems are anticipated to be available soon for clinical management.
There are several automated insulin delivery platforms utilising various combinations of insulin pumps, CGM systems and algorithms. Few are commercialised, with others in various stages of development. Despite the promising efficacy, the latency in the roll-out of these systems, the limited capacity for algorithm individualisation and the expense have also led to a community of ‘do it yourself’ systems to improve glycaemic control.
In the pursuit of a fully automated system, studies are looking into incorporating physiologic data (food and physical activity) into the algorithms to provide more accurate predictions and to fine-tune insulin delivery. Automated multihormone closed-loop systems are also in development which deliver glucagon in conjunction with insulin to mitigate the risk of hypoglycaemia and increase the aggressiveness of insulin therapy.