Use of a Commercially Available Automated Insulin Delivery System for the Management of Type 1 Diabetes in Pregnancy

In pregnancies with type 1 diabetes (T1D), maintaining glycemic control is critical for improving outcomes but challenging to achieve due to the physiologic demands of pregnancy. In nonpregnant adults with T1D, automated insulin delivery (AID) systems increase time in range (TIR), decrease hypoglycemia, and improve hemoglobin A_1c levels when compared with other methods of insulin delivery. ¹ In the pregnant population, there is a dearth of rigorous research regarding commercially available AID systems, and none have US Food and Drug Administration approval for use in pregnancy. Despite this, a growing number of patients opt for off-label use of AID systems during pregnancy. This retrospective descriptive cohort study was conducted to evaluate maternal, neonatal, and glycemic outcomes in gravidas with T1D who utilized the Omnipod 5 (Insulet Corp, Acton, Massachusetts), the newest commercially available AID system.

All patients with T1D who used this specific AID system for the duration of pregnancy and who delivered between February 2023 and February 2024 at University of California (UC) Los Angeles, UC San Diego, and UC San Francisco were included. The study was approved by a multi-institutional review board. Maternal and neonatal outcomes were abstracted from the electronic medical record, and glycemic data were derived from ambulatory glucose profiles at 12, 16, 20, 24, 28, and 32 weeks, with target range set to 63 to 140 mg/dL or 70 to 140 mg/dL. Proportions and medians were used to describe the cohort.

A total of 17 patients were included, with outcomes as summarized in Table 1. There were no episodes of diabetic ketoacidosis (DKA) or significant maternal hypoglycemia. The most common maternal complication was cesarean delivery (59%), followed by preeclampsia (35%). The most common neonatal complication was large for gestational age (LGA, 47%), followed by neonatal intensive care unit (ICU) admission (35%) and preterm birth (29%). Median TIR was 56% to 65%, and average glucose was 126 to 138 mg/dL throughout gestation.

Table 1.

Clinical and Glycemic Outcomes Among Pregnant Individuals With Type 1 Diabetes Using Omnipod 5 Automated Insulin Delivery System Through Pregnancy.

Clinical outcomes
Diabetic ketoacidosis	0 (0%)
Significant maternal hypoglycemia a	0 (0%)
Preeclampsia	6 (35%)
Cesarean delivery	10 (59%)
Obstetric anal sphincter injury	2 (12%)
Gestational age at delivery (weeks)	38.3 (36.6-39.0)
Preterm birth <37 weeks	5 (29%)
Birth weight (g)	3535 (3205-4285)
Birth weight percentile	90 (71-97)
Large for gestational age c	8 (47%)
Shoulder dystocia	0 (0%)
Neonatal ICU admission	6 (35%)
Severe neonatal hypoglycemia d	5 (29%)
Severe neonatal hyperbilirubinemia e	2 (12%)
Glycemic outcomes
Time below, in, and above range b (%)
12 weeks	5/65/31
16 weeks	6/65/29
20 weeks	4/58/38
24 weeks	4/56/39
28 weeks	7/58/38
32 weeks	6/59/35
Average glucose (mg/dL)
12 weeks	127
16 weeks	126
20 weeks	136
24 weeks	138
28 weeks	133
32 weeks	131

For clinical outcomes, data presented as N (%) for categorical variables and median (IQR) for continuous variables. For glycemic outcomes, data presented as median.

^aSignificant maternal hypoglycemia defined as hypoglycemic event requiring administration of glucagon and/or provision of resuscitative care from another individual.

^bRange set as 63 to 140 mg/dL or 70 to 140 mg/dL.

^cLarge for gestational age defined as birthweight greater than 90th percentile for gestational age.

^dSevere neonatal hypoglycemia defined as hypoglycemia requiring use of IV dextrose.

^eSevere neonatal hyperbilirubinemia defined as hyperbilirubinemia requiring use of phototherapy and/or exchange transfusion.

These outcomes are generally comparable to those found in other reports of gravidas with T1D,^2-5 suggesting that glycemic control and pregnancy complications with the newest commercially available AID system may be no better and no worse than that achieved with other methods of insulin delivery.

As AID systems grow in popularity, a two-pronged approach will be required to respond to the expansion and advancement of these systems in pregnancy. First, clinicians should understand the unique aspects of each AID system and use shared decision-making to support pregnant patients who are interested in AID systems, after adequate counseling regarding off-label use, the absence of rigorous clinical data in the pregnant population, and the possible risks and benefits. Second, researchers should examine the safety and efficacy of these systems in large, nationally representative populations of gravidas with T1D. Consideration should be given to the limitations imposed by the fact that the lowest glucose target available in current commercial AID systems remains higher than the pregnancy fasting target <95 mg/dL.

The generalizability of these findings is limited by the small sample size drawn from tertiary referral centers with expertise in advanced diabetes technologies. The study is also limited by the absence of a control group for direct comparison of outcomes between this AID system and other methods of insulin delivery. Because this AID system was introduced recently, provider and patient experience with this system is still evolving. Despite these limitations, this study provides important preliminary data regarding real-world outcomes with the use of this new method of insulin delivery in pregnancy.