Aims/hypothesis: The aim of this study was to compare glycaemic control and maternal-fetal outcomes in women with type 1 diabetes managed on insulin pumps compared with multiple daily injections of insulin (MDI).
Methods: In a retrospective study, glycaemic control and outcomes of 387 consecutive pregnancies in women with type 1 diabetes who attended specialised clinics at three centres 2006-2010 were assessed.
Results: Women using insulin pumps (129/387) were older and had a longer duration of diabetes, more retinopathy, smoked less in pregnancy, and had more preconception care (p < 0.01 for each). Among 113 pregnancies >20 weeks' gestation in women on insulin pumps and 218 in women on MDI, there was a significant difference in HbA1c in the first trimester (mean HbA1c 6.90 ± 0.71% (52 ± 7.8 mmol/mol) vs 7.60 ± 1.38% (60 ± 15.1 mmol/mol), p < 0.001), which persisted until the third trimester (mean HbA1c 6.49 ± 0.52% (47 ± 5.7 mmol/mol) vs 6.81 ± 0.85% (51 ± 9.3 mmol/mol), p = 0.002). Rates of diabetic ketoacidosis were similar in women on insulin pumps vs MDI (1.8% vs 3.0%, p = 0.72). Despite lower HbA1c, women on insulin pumps did not have an increased incidence of severe hypoglycaemia (8.0% vs 7.6%, p = 0.90) or more weight gain (16.3 ± 8.7 vs 15.2 ± 6.2 kg, p = 0.18). More large-for-gestational-age infants in the pump group (55.0% vs 39.2%, p = 0.007) may have resulted from confounding by parity.
Conclusions/interpretation: In this large multicentre study, women using insulin pumps in pregnancy had lower HbA1c without increased risk of severe hypoglycaemia or diabetic ketoacidosis but no improvement in other pregnancy outcomes. This information can help inform care providers and patients about the glycaemic effectiveness and safety of insulin pumps in pregnancy.