January 23 ,2026

Leads:

Carol Levy – Mt. Sinai (Adult)

Emily Coppedge – Cornell (Peds

Participants: Alexandra Sawyer, Don Buckingham, Christine Byer-Mendoza, Amit Lahoti, Arati Nagarkar, Amanda Perkins, Siham Accacha, Erin Cobry, Abha Choudhary, Hailee Delsart, Trevon Wright

Meeting Agenda:

• Welcome
• Presentation by, Alexandra Sawyer, MD, MPH  (Children’s Hospital Colorado & Barbara Davis Center)
• Topic: Meal Boluses for AID systems
• Discussion on current best practices recently implemented

Introduction

• Alex from the Barbara Davis Center, University of Colorado, is presenting on simple meal bolusing on automated insulin delivery.
• The presentation will cover the project, literature review, challenges with meal boluses, and other studies on simple meal bolusing.

Burden of Meal Bolusing

• Meal bolusing is a significant burden for patients with type 1 diabetes.
• A 2019 international survey of 1400 adults and 350 parents of children with type 1 diabetes showed that most believe accurate mealtime dosing is important, but fewer are confident in their ability to do so.
• 94-96% believe it’s very important.
• 35% of adults and 40% of parents felt very confident.
• Only 16% of physicians felt that their patients were very confident in figuring out how much insulin to give at a meal.
• Major challenges include needing to give more insulin after eating and not being sure how to dose accurately.
• Pre-meal insulin dosing has a negative impact on life, mood, social life, and relationships.
• 19% of adults and 44% of parents chose not to eat out at least once a week due to uncertainty about insulin needs.

Prescribing Practices and Carbohydrate Counting

• A survey of Pediatric Endocrine Society members in 2021 revealed considerations for prescribing insulin pumps.
• Most required a minimum number of blood sugar checks or CGM use.
• 25% had an A1C requirement, most commonly less than 10%.
• 80% felt that the ability to demonstrate carbohydrate counting was important.
• Adolescents often struggle with carbohydrate counting.
• A study showed only 23% of adolescents could accurately carb count within 10 grams.
• Another study showed only about 34% could accurately count carbs within 10 grams in four out of six common meals.
• The question arose whether carbohydrate counting is necessary on automated insulin delivery (AID) systems.

Research on Simplified Meal Bolusing

• Petrovsky conducted a study on simplified meal amount versus precise carbohydrate counting in adolescents using 780G.
• Participants were randomized to precise carb counting or fixed options (small, regular, large) based on a food diary.
• At 12 weeks, the carb counting group had significantly better time in range (80.3% vs. 73.5%), but both were above 70%.
• No difference in A1C, time below range, or time between 180 and 250.
• At one year, carb counting group still had significantly better time in range (80% vs. 73%), but no difference in A1C.
• A trial with 30 adults using a research AID system compared carb counting to four categories (very low, low, medium, large).
• Time in range was 70.5% with fixed options vs. 74% with carb counting, not demonstrating non-inferiority.
• The SMASH study, a randomized control trial with 46 adolescents using CAM APS FX, compared carb counting to a simple meal bolus period.
• Simple meal boluses were personalized based on a three-day diet history.
• The study demonstrated non-inferiority, with time in range of 70.7% during carb counting and 69.9% during simple meal bolus.
• A study in 14 adults with type 1 using Medtronic 770G compared carb counting to entering one preset amount or multiple preset amounts.
• Entering one same amount for every meal, the time in range was 75.4% compared to 77.7% with precise carb counting.
• In the multiple preset group, the time in range was significantly higher than with precise carb counting (80.5% vs. 77.7%).
• A study on simplified meal bolus strategies with Control IQ in adults with type 2 diabetes showed significant improvements in A1C.
• Participants were categorized into carbohydrate counting, preset carbohydrate amount, and fixed insulin dosing.
• Improvements in A1C were similar across all carbohydrate meal bolusing strategies.
• More people switched to simplified meal bolus strategies over time.

Fellowship Trial: Simple Meal Bolusing vs. Carb Counting

• A randomized crossover trial evaluated simple meal bolusing versus carbohydrate counting in adolescents on hybrid closed loop systems.
• The objective was to evaluate the effectiveness of a simple meal bolus strategy in achieving glycemic control and reducing mealtime burden.
• Participants aged 14-26 already using an AID system (excluding iLet) were enrolled.
• Data was gathered for two weeks at baseline.
• Participants were randomized to either simple meal bolus followed by carb counting, or vice versa, each for four weeks with a one-week washout.
• The simple meal bolus strategy involved entering 30 grams for a small meal, 60 grams for a medium, or 90 grams for a large, regardless of individual diet history.
• Carb ratios were standardized to 450 divided by the total daily dose.
• Participants were instructed not to carb count to decide which meal size to enter.
• Before the carb counting period, participants received basic re-education on carbohydrate counting and reading meal labels.
• The main hypothesis was that time in range would not be inferior by more than 5% during the simple meal bolus period compared to the carb counting period.
• 31 participants completed the trial.
• Mean age was 17. Average type 1 diabetes duration was 8.3 years.
• Approximately 70% were on Tandem Control IQ, 20% on Omnipod 5, and 13% on Medtronic 770G.
• At baseline, participants bolused 4.6 times per day.
• About 90% used an educated guess to determine bolus amount.
• Baseline time in range was 63.8%.
• Time in range during simple meal bolusing was 64.2% compared to 66% during carb counting, a difference of 1.7%.
• This demonstrated non-inferiority during the simple meal bolus period.
• No differences in hypoglycemia or time over 250.
• Participants bolused less frequently during the simple period (4.5 times per day) than during carb counting (5.1 times per day).
• No difference in average daily carbs entered or mean total daily dose of insulin.
• Surveys on burden, quality of life, and compensatory behaviors showed no significant differences except for how burdensome bolusing for meals felt.
• 50% said bolusing for meals never felt burdensome during the simple meal bolus period, compared to 10% during carb counting.
• 39% preferred the simple meal bolus strategy, 32% preferred carb counting, and 29% preferred what they were doing before the study (some form of simple guesstimation).
• The biggest complaint about the simple strategy was the lack of a lower number to enter for small snacks.
• Using a simple meal bolus strategy had a similar impact on glycemic control as precise carbohydrate counting.

Questions and Discussion

• Amanda asked if fixed dosing would cause people to enter carbs more often, since it’s the default for patients with A1Cs over 10 when starting pumps.
• It would be interesting to study this in patients less comfortable with carb counting or who enter carbs less than three times a day.
• Someone mentioned that in the T2 IQ meal challenge paper, some people never bolused, even with auto correction boluses.
• It’s clear people don’t have to be perfect with carb counting.
• Mindset has changed, and it’s important to consider the amount people entering and if they’re learning to associate certain entries with specific times of day.

Tools and Strategies

• Omnipod 5 has a custom food function that can be used for small, medium, or large food entries.
• Tandem provides PDFs with suggestions for small, medium, and large carb amounts (e.g., 30, 60, 90 grams) or fixed unit dosing.
• Medtronic has a preset bolus feature, but it can’t be used in auto mode.
• RDs often use worksheets with set numbers for small, medium, and large meals at different times.

Presentation: T1Dexchangemealbolusing

Recording: Hybrid Closed Loop Working Group-20260123_140428-Meeting Recording.mp4