Use of commercially available automated insulin delivery systems in pregnant people with type 1 diabetes

Automated insulin delivery (AID) systems are first-line therapy for type 1 diabetes, but commercially available AIDs in the United States are not approved for pregnancy. We aimed to compare glycemic control achieved during pregnancy by people with type 1 diabetes using AIDs versus standard of care therapy (multiple daily injections and sensor augmented pump therapy).

This was a retrospective cohort study of people with type 1 diabetes who used a continuous glucose monitor (CGM) during pregnancy. The primary outcome was time in range (TIR); time below range (TBR), time above range, and glucose standard deviation were secondary outcomes. Outcomes were compared using analysis of covariance.

38 people were included: 21 treated with AIDs and 17 with standard of care. The mean antenatal TIR in the AID group was 68.5 ± 12.9 % compared to 55.7 ± 16.7 % (adjusted mean difference 7.9 %, 95 % CI: 0.8 to 15.0, p = 0.03). The AID group achieved a lower TBR (1.7 ± 1.3 % vs 3.0 ± 2.8 %, p = 0.03), and lower glucose standard deviation (37.2 ± 8.0 vs 45.5 ± 8.9, p = 0.02) than standard of care.

In this real-world study, off-label AID use during pregnancy improved TIR while decreasing TBR. While awaiting commercially available AIDs with pregnancy algorithms, standardized approaches to optimizing current systems are needed.