Automated insulin delivery in pregnant women with type 1 diabetes mellitus: a systematic review and meta-analysis

AIMS: To assess the efficacy and safety of automated insulin delivery (AID) systems compared to standard care in managing glycaemic control during pregnancy in women with Type 1 Diabetes Mellitus (T1DM). METHODS: We searched MEDLINE, Cochrane Library, registries and conference abstracts up to June 2024 for randomized controlled trials (RCTs) and observational studies comparing AID to standard care in pregnant women with T1DM. We conducted random effects meta-analyses for % of 24-h time in range of 63-140 mg/dL (TIR), time in hyperglycaemia (> 140 mg/dl and > 180 mg/dL), hypoglycaemia (< 63 mg/dl and < 54 mg/dL), total insulin dose (units/kg/day), glycemic variability (%), changes in HbA1c (%), maternal and fetal outcomes. RESULTS: Thirteen studies (450 participants) were included. AID significantly increased TIR (Mean difference, MD 7.01%, 95% CI 3.72-10.30) and reduced time in hyperglycaemia > 140 mg/dL and > 180 mg/dL (MD - 5.09%, 95% CI - 9.41 to - 0.78 and MD - 2.44%, 95% CI - 4.69 to - 0.20, respectively). Additionally, glycaemic variability was significantly reduced (MD - 1.66%, 95% CI - 2.73 to - 0.58). Other outcomes did not differ significantly. CONCLUSION: AID systems effectively improve glycaemic control during pregnancy in women with T1DM by increasing TIR and reducing hyperglycaemia without any observed adverse short-term effects on maternal and fetal outcomes.