SAT093 Closed-Loop Insulin Delivery Versus Standard Care In Type 1 Diabetic Pediatric Patients: A Meta-analysis Of Randomized Controlled Trials

Disclosure: L.C. Hespanhol: None. V.C. Moreira: None. A.C. Silva: None. I.R. Marques: None. A. Godoi: None. A. Mahesh: None. C. Oommen: None. C.H. Silva: None. C. Gomes: None. I.A. Souza: None. I.A. Miyawaki: None. J.E. Loyola Júnior: None. J. De Sa: None. E.M. Padrao: None. Background: Treatment of Type I Diabetes Mellitus (T1DM) in the pediatric population is a clinical challenge. The first closed-loop system (CL) was approved for pediatric use in 2020 with the aim of improving therapeutic and safety outcomes for patients. Since then, many randomized clinical trials have been performed, yet, to our knowledge no meta-analysis evaluating the effect of long-term CL on glycemia in children and adolescents with T1DM was performed. Purpose: To compare the therapeutic efficacy and safety of long-term use of CL insulin delivery systems to standard care (SC) in pediatric patients with T1DM. Methods: PubMed, Cochrane and EMBASE were systematically searched in October 2022 to include randomized controlled studies (RCTs) comparing CL systems (automated pancreas, hybrid CL and advanced hybrid CL) versus SC (daily insulin injections, sensor augmented pump and continuous glucose monitoring) with at least 12 weeks of duration. Outcomes assessed were percentage time in range (% TIR) 70-180mg/dL, change in HbA1c and percentage time with hypoglycemia (glucose < 70mg/dL). Statistical analysis were performed with RevMan 5.4.1 and R software. Heterogeneity was assessed with I² statistics and random-risk effect was used if I2 > 50%. The protocol was registered in PROSPERO (ID: CRD42022366710). Results: A total of 7 randomized studies with 827 patients with T1DM were included, of whom 471 (56.7%) used automated devices and 356 (43.3%) continued their usual care. The %TIR was significantly higher in the CL systems when compared to SC (MD 8.70%; 95% CI 7.08 to 10.31; p < 0.001; I² = 0%). Similarly, CL systems showed a significantly higher HbA1c mean difference when compared to SC (MD -0.38; 95% CI -0.59 to -0.16; p < 0.001; I² = 0%). When evaluating percentage time with hypoglycemia, the time was non-significantly lower in the CL system group when compared to SC (MD -0.47%; 95% CI -1.06 to 0.13; p = 0.12; I² = 55%). Conclusion: In this meta-analysis of randomized trials in pediatric population with T1DM, the use of CL insulin devices was associated with a higher proportion of time in the optimal glucose range and greater decrease of HbA1c relative to standard care, regardless of CL system type. However, the use of CL insulin delivery was not associated with less time in hypoglycemia range. Therefore, CL systems seem to be safe and superior to standard care regarding glycemic control. Presentation: Saturday, June 17, 2023