Efficacy and Safety of Automated Insulin Delivery Systems in Patients with Type 1 Diabetes Mellitus: A Systematic Review and Meta-Analysis

Abstract

Backgruound: Automated insulin delivery (AID) systems studies are upsurging, half of which were published in the last 5 years. We aimed to evaluate the efficacy and safety of AID systems in patients with type 1 diabetes mellitus (T1DM).

Methods: We searched PubMed, Embase, Cochrane Library, Web of Science, and ClinicalTrials.gov until August 31, 2023. Randomized clinical trials that compared AID systems with other insulin-based treatments in patients with T1DM were considered eligible. Studies characteristics and glycemic metrics was extracted by three researchers independently.

Results: Sixty-five trials (3,623 patients) were included. The percentage of time in range (TIR) was 11.74% (95% confidence interval [CI], 9.37 to 14.12; P<0.001) higher with AID systems compared with control treatments. Patients on AID systems had more pronounced improvement of time below range when diabetes duration was more than 20 years (-1.80% vs. -0.86%, P=0.031) and baseline glycosylated hemoglobin lower than 7.5% (-1.93% vs. -0.87%, P=0.033). Dual-hormone full closed-loop systems revealed a greater improvement in TIR compared with hybrid closed-loop systems (-19.64% vs. -10.87%). Notably, glycemia risk index (GRI) (-3.74; 95% CI, -6.34 to -1.14; P<0.01) was also improved with AID therapy.

Conclusion: AID systems showed significant advantages compared to other insulin-based treatments in improving glucose control represented by TIR and GRI in patients with T1DM, with more favorable effect in euglycemia by dual-hormone full closedloop systems as well as less hypoglycemia for patients who are within target for glycemic control and have longer diabetes duration.

Keywords: Diabetes mellitus, type 1; Insulin infusion systems; Meta-analysis; Systematic review.

Fig. 1.

Forest plot for time in range comparing automated insulin delivery systems with other insulin-based treatment. SD, standard deviation; CI, confidence interval.

Fig. 2.

Prespecified subgroup analyses for primary and secondary outcomes by the type of automated insulin delivery systems (dual-hormone full closed-loop systems, single-hormone full closed-loop, or hybrid closed-loop systems). CI, confidence interval; HbA1c, glycosylated hemoglobin.