1987-LB: Strategies for Setting Carbohydrate Ratios in Difficult-to-Learn Cases during MiniMed 770G Automated Insulin Delivery System Implementation

Introduction and Objective: This study explored a strategy for setting carbohydrate ratios for patients who struggle with precise carbohydrate counting (PCC) when using an automated insulin delivery (AID) system. Methods: This is a prospective study of 14 Japanese patients with type 1 diabetes who were newly introduced to the MiniMed 770G. Patients were divided into two groups: an intuitive carbohydrate counting (ICC) group, which consisted of patients who had difficulty learning PCC and were instructed to use the simple carbohydrate counting (CC) method that we developed and a PCC group. The ICC method quantifies the carbohydrate content of the staple food by setting the carbohydrate-to-insulin ratio on the basis of the amount of bolus insulin administered, with the carbohydrate content of a standard staple food being 100 g. Glycated hemoglobin (HbA1c) levels and ambulatory glucose profile reports were compared over 48 weeks. Owing to the small sample size, Bayesian statistics were applied for analysis. Results: At 48 weeks, both groups exhibited improvements in HbA1c levels. In the ICC group, the mean HbA1c decreased from 8.30 ± 0.63% to 6.50 ± 0.41% (ΔHbA1c of -1.68 ± 1.28%). In the PCC group, the mean HbA1c decreased from 7.81 ± 0.47% to 6.98 ± 0.89% (ΔHbA1c of -0.95 ± 0.93%). Bayesian analysis revealed that the estimated difference in ΔHbA1c between the ICC and PCC groups was -0.74 (95% credible interval: -1.23 to 0.66).The time in range (TIR) increased, while the time above range (TAR) decreased in both groups (ΔTIR: 16.8±8.38% vs. 12.1±8.68%; ΔTAR: -18.0±8.21% vs. -9.50±8.80%). Conclusion: The introduction of an ICC method for patients who struggle with PCC resulted in glycemic management comparable to that achieved by those using PCC. This suggests that the ICC and AID system may improve glycemic management in patients with difficult glycemic management who have difficulty with accurate CC. Disclosure K. Hirayama: None. M. Koshizaka: None. E. Hirayama: None. T. Sakai: None. S. Hayashi: None. K. Sakurai: None. R. Ishibashi: Research Support; Astellas Pharma Inc, Taiho Pharmaceutical Co. Ltd.