The Advice4U Study: Insulin Dose Optimization Using an Automated Artificial Intelligence-based Decision Support System in Youths with Type 1 Diabetes

רויטל נימרי ¹ Tadej Battelino ² Lori M Laffel ³ Robert H Slover ⁴ Desmond Schatz ⁵ Stuart A Weinzimer ⁶ Klemen Dovc ² Thomas Danne ⁷ Moshe Phillip ¹

¹The Jesse Z and Sara Lea Shafer Institute for Endocrinology and Diabetes, National Center for Childhood Diabetes, Schneider Children’s Medical Center of Israel, Petah Tikva, Israel
²Department of Endocrinology, Diabetes and Metabolic Diseases, UMC-University Children's Hospital Ljubljana, and Faculty of Medicine, University of Ljubljana
³Joslin Diabetes Center, One Joslin Place, Harvard Medical School, Boston, Massachusetts, Joslin Diabetes Center, One Joslin Place, Harvard Medical School, Boston, Massachusetts
⁴Barbara Davis Center for Childhood Diabetes, University of Colorado Anschutz Medical Campus, Aurora, CO, Barbara Davis Center for Childhood Diabetes, University of Colorado Anschutz Medical Campus, Aurora, CO
⁵Department of Pediatrics, College of Medicine, University of Florida, Gainesville, FL
⁶Pediatric Endocrinology & Diabetes, Yale School of Medicine, New Haven, Connecticut, Pediatric Endocrinology & Diabetes, Yale School of Medicine, New Haven, Connecticut
⁷Diabetes Center for Children and Adolescents, Children's Hospital AUF DER BULT, Hannover, Germany, Diabetes Center for Children and Adolescents, Children's Hospital AUF DER BULT, Hannover, Germany

BACKGROUND: Despite increasing adoption of insulin pump and continuous glucose monitoring devices, most people with type 1 diabetes do not achieve their glycemic goals. This could be related to a lack of expertise or inadequate clinicians’ time to analyze complex sensor-augmented pump data. We tested whether frequent insulin dose adjustments guided by an automated artificial intelligence-based decision support system (AI-DSS) is as effective and safe as those guided by physicians in controlling glucose levels. METHODS: ADVICE4U was a 6-month, multicenter, multinational, parallel, randomized controlled, non-inferiority trial in 108 participants with type 1 diabetes, aged 10-21 years and using insulin pump therapy. Participants were randomized 1:1 to receive remote insulin dose adjustment every 3 weeks guided by either an AI-DSS, (AI-DSS arm, n=54) or by physicians (physician arm, n=54). RESULTS: The percentage of time spent within target glucose range (70 to 180 mg/dl), in the AI-DSS arm was statistically non-inferior to the physician arm (50.2±11.1% vs. 51.6±11.3%, respectively, P<1e-7). The percentage of readings below 54 mg/dl within the AI-DSS arm was statistically non-inferior to the physician arm (1.3±1.4% vs. 1.0±0.9%, respectively, P<0.0001). Three severe adverse events related to diabetes were reported in the physician arm and none in the AI-DSS arm. CONCLUSIONS: Use of an automated tool for treatment optimizing was non-inferior to intensive insulin titration provided by physicians from specialized academic diabetes centers. The AI-DSS provides opportunities for a new modality for intensive insulin management necessary to improve glucose control in young people with type 1 diabetes.