Cost-effectiveness of insulin aspart compared to human insulin in pregnant women with type 1 diabetes in the UK

Objectives: In women with type 1 diabetes, poor glycaemic control during pregnancy is associated with high risk of pre-term delivery, perinatal mortality and morbidity. This economic analysis utilises clinical effectiveness data from the Insulin Aspart Pregnancy Study Group Trial to assess costs and outcomes associated with insulin aspart (IAsp) and human insulin (HI) as part of a basal-bolus insulin regimen in pregnant women with type 1 diabetes in the UK. Research design and methods: Women with type 1 diabetes were enrolled if <= 10 weeks pregnant or planning to become pregnant, and had HbA(1c) <= 8% at confirmation of pregnancy. Subjects were randomised to treatment with IAsp or HI in a basal-bolus regimen with NPH insulin, with doses titrated according to American Diabetes Association guidelines. An effectiveness endpoint, retrospectively defined for this analysis, was the percentage of women with a live birth at term (>= 37 weeks' gestation). We considered costs of insulin, adverse events, delivery, and neonatal care for pre-term infants. Expected need for neonatal care was estimated from gestational age, using data from the literature and a large UK hospital. Costs were calculated from the perspective of the UK National Health Service. Results: A total of 322 pregnant women were enrolled in the study and the outcome of pregnancy was known for 302, 151 in each arm. More women experienced a live birth at term with IAsp (72.8%) than with HI (60.9%), difference 11.9% (95% CI 2.0%, 22.5%, p=0.028). Mean cost per woman was 3222 pound for IAsp and 3539 pound for HI, difference -318 pound ( 95% CI -1353 pound, 576; pound p = 0.49). Conclusions: Compared with HI, the use of IAsp in pregnant women with type 1 diabetes resulted in more live births at term, without increasing total costs of treatment. A prospectively defined study is required to confirm these conclusions.