Cost-effectiveness of switching to insulin degludec from other basal insulins: evidence from Swedish real-world data.

OBJECTIVES

Health economic analysis from a healthcare and societal point of view was conducted to assess the cost-effectiveness of insulin degludec (IDeg) after switching from other basal insulins in people with type 1 diabetes.

MATERIAL AND METHODS

This was a prospective, open-label, single arm, observational follow-up from August 2013 to October 2015 of 476 consecutive patients at Danderyd Hospital (Stockholm, Sweden) who switched to IDeg from other basal insulins (99% basal insulin analogs). The IMS CORE Diabetes Model (CDM) was used to predict the cost-effectiveness of life-long treatment with IDeg vs. other basal insulins, based on a Swedish setting.

RESULTS

Mean (SD) duration of follow-up was 21.7 (6.0) weeks. Mean HbA decreased by 2.7 mmol/mol, mean basal insulin dose decreased by 13.1% (p < .0001), and mean bolus insulin dose decreased by 7.5% (p < .0001) after switching. Frequencies of non-severe daytime hypoglycemia and non-severe nocturnal hypoglycemia decreased by 12% (p = .0127) and 53% (p < .0001) respectively and severe hypoglycemia was reduced by 62% (p = .0225). The CDM predicted a gain in life expectancy of 0.33 years, a discounted gain in quality-adjusted life-years (QALYs) of 0.54, and lower estimated direct lifetime healthcare costs of SEK 22,757 for patients switching to IDeg. The incremental cost-effectiveness ratio (ICER) showed IDeg as dominant (i.e. higher effectiveness with a lower cost). Sensitivity analyses confirmed the results.

CONCLUSION

Based on this prospective, real-world, follow-up and using the CDM, it was estimated that switching to IDeg from other basal insulins translated into QALY gains including improved life expectancy and health-related quality of life, as well as dominant ICER, meaning cost-savings for the healthcare system. However, the study is limited by its observational design. Extrapolation into the future is only estimated since the actual treatment effect cannot be projected with certainty.