Abstract
INTRODUCTION: Basal insulin injections have historically been administered via once-daily (OD) or twice-daily (BD) injections. Once-weekly (OW) basal insulin injections have recently been developed. This study aimed to quantify the relative importance of the administration frequency in basal insulin treatment preferences of people living with T2D in Canada, Spain, France, and Japan, using a discrete choice experiment (DCE). METHODS: Best-practice guidelines for patient preference studies were followed in a three-phase study design. Phases one (targeted literature review) and two (qualitative interviews) informed the development of an attributes and levels grid. Phase three consisted of pilot interviews to evaluate the feasibility of preference survey completion and DCE tasks among adults living with T2D across Canada, France, Spain, and Japan. Hierarchical Bayesian estimation was used to estimate part-worth utilities for attribute levels, then calculate the relative importance of each attribute among other attributes tested. RESULTS: The DCE survey was completed by N = 513 participants (aged 20-90; 54% male, 45% female; mean time since diagnosis: 11.6 years). Participants were split into three treatment groups: basal insulin and injectable glucagon-like peptide-1 receptor agonist (GLP-1 RA) naïve (n = 176), basal insulin naïve but with injectable GLP-1 RA experience (n = 176) and basal insulin experienced (n = 161). The administration frequency had a relative importance of 40% across the full sample, double that of any other treatment attribute tested in this study. A preference for OW administration was found relative to OD and BD. Findings were consistent across treatment groups and countries. CONCLUSIONS: This study demonstrated the value and importance of administration frequency in making choices for basal insulin treatments when glycemic control is held constant. Per the pre-specified conditions, participants expressed a preference for OW basal insulin, making considered trade-offs between treatment risks (e.g., risk of a severe hypoglycemic event) and convenience (e.g., frequency of administration).