Automated insulin delivery (AID) systems, including the artificial pancreas developed by the University of Virginia Health, hold promise for improving the management of type 1 diabetes (T1D). According to a comprehensive review published by the University of Virginia Health System, these technologies have already aided millions in monitoring blood glucose levels and enhancing their healthcare experience. Despite their benefits, the review highlights several limitations that need addressing before full automation can be realized.
Dr. Sue Brown, an endocrinologist and researcher at the University of Virginia Center for Diabetes and Technology, emphasized the challenges still faced by users. “These automated insulin delivery devices have significantly helped people with type 1 diabetes manage their blood sugars, and yet challenges still remain,” she stated. Current systems often require users to manually input carbohydrate information before meals, which hinders their effectiveness. This hybrid model does not account for the dose-dependent effects of protein and fat on insulin needs, leading to potential mismanagement of blood sugar levels.
Identifying Key Limitations in Current Systems
One significant limitation of existing AID systems is their requirement for users to provide carbohydrate counts prior to meals. This step complicates the process and can lead to inaccuracies in insulin delivery. The review suggests that future advancements should focus on simplifying carbohydrate entry, developing ultra-rapid-acting insulin formulations, and creating meal detection algorithms that integrate meal composition models.
Another noted challenge is the activity/exercise mode present in many AID devices. Currently, users must activate this function well in advance of physical activity. Clinical guidelines recommend temporarily raising insulin targets before exercise; however, many individuals with T1D do not consistently apply these strategies. This gap is particularly concerning given the risk of activity-induced hyperglycemia, especially following intense exercise or competition. Although some proof-of-concept devices can adapt insulin delivery based on exercise type, improving glucose time-in-range outcomes during and after activity remains a critical challenge.
The Role of Artificial Intelligence in Future Developments
Accessibility and user-friendliness are also significant barriers to widespread adoption of AID technologies. Dr. Marc Breton, an associate professor of research at the UVA Center for Diabetes Technology, pointed out that despite the revolutionary impact of AID, most patients using insulin do not have access to these advanced systems. “Simplifying the use of these systems will greatly improve access,” he remarked. The potential integration of artificial intelligence and data science into these devices could streamline insulin monitoring and delivery, making them more user-friendly.
The researchers concluded that while AID technology has made substantial strides, ongoing efforts are essential to address the limitations identified in their review. By focusing on these areas, the hope is to enhance the quality of life for individuals living with type 1 diabetes and to facilitate better management of their condition.
The findings were published in the journal J Diabetes Sci Technol on August 19, 2025, and further insights can be accessed through the University of Virginia Health System website. As the field continues to evolve, the integration of innovative approaches and technologies could pave the way for a future where managing type 1 diabetes is significantly more straightforward and effective.
