Cost-effectiveness of insulin analogues for diabetes mellitus

Abstract

Background: Insulin analogues may be associated with fewer episodes of hypoglycemia than conventional insulins. However, they are costly alternatives. We compared the cost-effectiveness of insulin analogues and conventional insulins used to treat type 1 and type 2 diabetes mellitus in adults.

Methods: We conducted a cost-effectiveness evaluation of insulin analogues versus conventional insulins using the Center for Outcomes Research Diabetes Model. We compared rapid-acting analogues (insulin aspart and insulin lispro) with regular human insulin, and long-acting analogues (insulin glargine and insulin detemir) with neutral protamine Hagedorn insulin. We derived clinical information for the comparisons from meta-analyses of randomized controlled trials. We obtained cost and utility estimates from published sources. We performed sensitivity analyses to test the robustness of our results.

Results: For type 1 diabetes, insulin aspart was more effective and less costly than regular human insulin. Insulin lispro was associated with an incremental cost of Can$28,996 per quality-adjusted life-year. The incremental cost per quality-adjusted life-year was Can$87,932 for insulin glargine and Can$387,729 for insulin detemir, compared with neutral protamine Hagedorn insulin. For type 2 diabetes, insulin aspart was associated with an incremental cost of Can$22,488 per quality-adjusted life-year compared with regular human insulin. For insulin lispro, the incremental cost was Can$130,865. Compared with neutral protamine Hagedorn insulin, insulin detemir was less effective and more costly. Insulin glargine was associated with an incremental cost of Can$642,994 per quality-adjusted life-year. The model was sensitive to changes in the effect size of hemoglobin A(1c) and to decrements applied to utility scores when fear of hypoglycemia was included as a factor.

Interpretation: The cost-effectiveness of insulin analogues depends on the type of insulin analogue and whether the patient receiving the treatment has type 1 or type 2 diabetes. With the exception of rapid-acting insulin analogues in type 1 diabetes, routine use of insulin analogues, especially long-acting analogues in type 2 diabetes, is unlikely to represent an efficient use of finite health care resources.

Figure 1: Schematic representation of the Center for Outcomes Research Diabetes Model (adapted and simplified from the original schematic by Palmer and colleagues6,7). The cycle length is 1 day for mild to moderate hypoglycemia, 3 months for severe hypoglycemia and 1 month for ulcer submodels. All other submodels have a cycle length of 1 year.

Figure 2: Cost-effectiveness acceptability curves of insulin analogues for the treatment of type 1 and type 2 diabetes mellitus in adults, by treatment comparison. Costs are in 2007 Canadian dollars.