1000-P: A Feasibility Study Assessing the Accuracy of a Simplified Meal-Time Bolus Calculation Option

Carbohydrate counting is a significant source of burden for individuals using insulin to mitigate postprandial glucose excursion. To address this, a meal simplification (MS) algorithm was developed that allows one to choose a ready and user-specific small, medium, large, extra-large (S/M/L/XL) carb-entry size for meal-time boluses. Subjects with T1D (N=12, 15-76 yrs) underwent a 4-week learning phase where traditional carb-counting and entry were carried out via the MinMed™ 670G system. At the end of the learning phase, the MS algorithm used machine-learning techniques and insulin pump data to identify clusters of meals (a set of S/M/L/XL meal sizes along with a corresponding centroid carb value) on which a meal bolus would be calculated. Immediately after, there was a 2-week study phase where subjects announced carbs based on their personalized S/M/L/XL meal sizes. Postprandial SG levels after S/M/L/XL announcements did not differ from those after traditional carb-counting (Table) and suggest that the personalized S/M/L/XL option is a valid alternative for carbohydrate counting. It may alleviate some of the burden associated with meal-time carb-counting, and potentially improve meal-time bolusing compliance.