NCT06411548

InsuLearn Feasibility With Type 1 Diabetes Patients Under MDI Therapy

Feasibility of a Bolus Calculator Without Carbohydrate Counting in Type 1 Diabetes Patients Under Multiple Daily Injections (MDI) Therapy: A Supervised Randomized Controlled Trial.

• Status: Completed
• Phase: N/A
• Study type: Interventional
• Enrollment: 7 (actual)
• Sponsor: University of Virginia · Academic / Other
• Sex: All
• Age: 18 Years
• Healthy volunteers: Not accepted

Summary

This is a 24-hour, randomized, crossover, single-center trial where participants are randomized to either start with the InsuLearn intervention or the usual care (UC) intervention. In the InsuLearn intervention, insulin doses are optimized using data collected in a 4-weeks at home data collection period.

Detailed description

Following recruitment and screening, 12 participants will be randomized to either the InsuLearn→UC or the UC→InsuLearn sequence. Participants will be trained in the use of a Continuous Glucose Monitor (CGM) (Dexcom G6 or Guardian™ 4), smart insulin pen (SIP, InPen™, Medtronic) and the InPen™ bolus calculator phone application. The InPen™ application will be set to the "meal estimation" mode where the user can select the meal category: if it is a breakfast/lunch/dinner/snack and if it has low/med/high carb amount. During the one-month data collection period, participants will be asked to use the InPen™ app to log and deliver their insulin doses. Before the hotel admission, InsuLearn will analyze the collected data to learn a new optimized insulin dose for each meal category. InsuLearn will generate a report of suggested new insulin doses that will be reviewed, approved, or changed by the study physician as needed. In the 24-h supervised interventions (hotel admission), the participant will consume three standardized meals, a med-carb lunch, a high-carb dinner, and low-carb breakfast. The insulin dose delivered in each admission will be either calculated by participants as per their UC or is as provided by the InsuLearn approved report. The two interventions are combined into a 2-day admission without a washout period.

Conditions

• Type 1 Diabetes Mellitus

Interventions

Type	Name	Description
DEVICE	InsuLearn	InsuLearn is an artificial intelligence (AI)-algorithm that analyses collected participants data: CGM, insulin, and meal categories to learn the optimal insulin dose. InsuLearn will run offline in a study laptop to generate the optimal insulin dose for each meal category. The algorithm will generate an ambulatory glucose profile (AGP) report with new settings for the InPen™ meal estimation bolus calculator. The study physician will review the report and the algorithm recommendations. The study physician can accept or modify the algorithm recommendations as per his judgment. The new settings will be used during the intervention admission.

Timeline

• Start date: 2025-03-27
• Primary completion: 2025-10-15
• Completion: 2025-10-17
• First posted: 2024-05-13
• Last updated: 2025-10-27

Locations

1 site across 1 country: United States

Regulatory

• FDA-regulated device study