The road to personalized management of type 1 diabetes with a focus on closed-loop devices
Insulin therapy is of vital importance for people with type 1 diabetes (T1D). Attaining near-normal blood glucose values reduces the risk of diabetic complications, but a regimen of intensive insulin administration is needed. Any current exogenous insulin therapy is however associated with risk of developing hypoglycemia often leading to fear of hypoglycemia and thus a barrier to achieve optimal diabetes control. The monitoring of glucose values is essential in order to minimize the risk of hypoglycemia and hyperglycemia by balancing the insulin dose with current and predicted glucose values. Finding the right balance can be very challenging since it is influenced by factors such as meals, physical activity (PA), stress, and illness. New technologies have been developed in the past years to support people with T1D in self-monitoring, optimizing insulin therapy and lessen the burden. One of these new technologies is closed-loop insulin delivery system. Here, a continuous glucose monitor (CGM) is connected to an insulin pump which is controlled by an algorithm. As a result, the insulin pump will infuse insulin automatically through feedback of current and predicted sensor glucose values. However, even though such new advancements in technology have been proven to be beneficial in randomized controlled trials, there are still barriers to their adoption in daily living. A first barrier is the lack of integration of PA in the closed-loop control algorithm, which can result in an inadequate feedback response during and after moments of PA or the need to manually manipulate the settings of the device. Therefore, the main objective of this first subproject is the development of artificial intelligence enhanced algorithms which predict blood glucose levels and give insights in the relation between blood glucose variation and PA by incorporating the effect of PA on blood glucose. This algorithm might eventually be integrated in future closed-loop algorithms. This project is in collaboration with data scientists from UCLL. A second barrier is the lack of reimbursement for these expensive devices. This is partly the consequence of limited real-world data about the use of closed-loop devices in the daily lives of people with T1D. Belgium has a unique reimbursement system which allows the early implementation of such closed-loop systems in the daily lives of a limited number of people with type 1 diabetes. Within the framework of this “experimental garden”, the main objective of this second subproject is evaluating the real-world impact of commercially-available closed-loop devices on glycemic control and patient-reported outcomes. The important scientific data that comes out of well-designed prospective real-world studies might also be instrumental in convincing payers to make this type of technology accessible to a wider population. This project will be conducted as a multicenter observational study.