Діабет у світі

Can an AI Chatbot Revolutionize Insulin Dosing?

Can an AI Chatbot Revolutionize Insulin Dosing?

Artificial intelligence has shown various applications in the treatment of diabetes, from scheduling consultations with a doctor to automated insulin delivery and detecting diabetic retinopathy.

A dedicated AI chatbot can suggest insulin dosing optimization for patients with type 2 diabetes, promoting better treatment. Initiation of insulin therapy can be difficult, requiring regular visits to healthcare professionals for dose adjustments. As a result, many patients do not receive the correct dose and are unable to achieve glycemic control because of this.
Researchers at Stanford University examined whether an AI voice chatbot could help patients with type 2 diabetes start basal insulin therapy and adjust doses. In the study, 32 participants were randomly assigned to receive standard assistance or voice AI. Members of the AI group received special AI voice software for insulin dosing using Amazon Alexa.

The study focused on the time to optimal insulin dose, which was defined as the number of days it took participants to reach a target of 3-day mean fasting blood glucose.

What were the results?

Participants who received AI-based care were able to optimize insulin doses faster, take insulin more consistently, and ultimately had better glycemic control. These participants also experienced a reduction in diabetes distress.

Here are some of the key findings from the study:

  • Participants in the AI-care group achieved optimal insulin doses at just 15 days, whereas those receiving usual care achieved ideal insulin doses at nearly two months.
  • Roughly 81% of participants using the AI chatbot achieved glycemic control (measured by fasting blood glucose levels), compared to 25% of those in the standard care group.
  • There was a significant decrease in diabetes distress among those using the AI chatbot compared to the standard care group.


The researchers described several limitations of this study. For example, most of the blood sugar data was obtained by the participants themselves, rather than being collected by continuous glucose monitoring (CGM) systems or healthcare professionals.

The study was short, lasting only two months, and included only 32 adults with type 2 diabetes. Because the study was very short, glycemic control was measured using fasting blood glucose rather than glycated hemoglobin.

Additionally, only one AI chatbot was examined in this study and was not compared to other virtual assistants. More research is needed to confirm the findings in a larger and more diverse patient population.

The results suggest that AI virtual assistants can benefit people with diabetes who are starting insulin therapy and need to optimize their dose. In addition to improving insulin doses, the chatbot has also led to positive health outcomes.

According to diaTribe Can an AI Chatbot Revolutionize Insulin Dosing?
EN/Наукові матеріали та дослідження про діабет