Carbohydrate Counting in the Real World: Mealtime Insulin Dosing Behaviors and Smartphone App Use in Adults with Type 1 Diabetes

OBJECTIVE: Strategies for prandial insulin dosing are not well characterized but have important implications for diabetes care in real-world practice. Our objective for this study was to characterize mealtime insulin dosing strategies, evaluate discrepancies between patient-reported and provider-documented dosing approaches, and assess smartphone app use for carbohydrate counting among adults with type 1 diabetes (T1D). RESEARCH DESIGN AND METHODS: Adults with type 1 diabetes (T1D) with an ICD diagnosis code for T1D and an active patient portal account within an academic health system were invited to complete a survey assessing mealtime insulin dosing behaviors and smartphone application (app) use for carbohydrate counting. Responses were linked to electronic health record data. RESULTS: Among 903 participants, mean age was 45 years and mean diabetes duration was 23 years; 53% were women, 67% were White non-Hispanic, 95% used CGM, and 58% used automated insulin delivery. Overall, 54% reported carbohydrate counting, 39% used experience-based estimation, and 2% used fixed dosing; A1c did not differ by dosing strategy. Among participants with provider-documented carb counting, 36% reported experience-based estimation. App use was infrequent, with 51% reporting never using apps; common barriers were memorization of commonly eaten foods and perceived time burden. CONCLUSIONS: Many adults with T1D rely on heuristic insulin dosing approaches that prioritize practicality over maximal precision and were not associated with worse glycemic outcomes, underscoring the need for practical, sustainable self-management strategies.