Objectives: The rapid increase in insulin prices over the last decade may negatively impact its use and adherence. This study aims to assess insulin adherence in individuals in the U.S. with type 1 (T1D) or type 2 diabetes (T2D) and newly-initiated insulin.

Methods: We identified individuals with diabetes from the MarketScan Commercial database based on ≥2 non-diagnostic medical (not laboratory or radiology) claims with diagnoses of diabetes on different days in 2018. Individuals were required to have filled at least one insulin prescription between July 1, 2018 and December 31, 2018, with the first observed insulin fill date defined as the index date. The study period for each individual ranges from the start of the pre-index 6 months (pre-period) through the end of the post-index 1 year (post-period). We excluded individuals who filled insulin prescriptions during the pre-period, were not continuously enrolled or had diagnoses of pregnancy. We measured adherence during the post-period by the proportion of days covered (PDC) adjusted for overlapping insulin prescriptions at person level.

Results: Among 16,165 included individuals, 2,315 had T1D (46.2% men) and 13,850 had T2D (52.3% men), with mean age of 33.8 and 51.0 years, respectively. Overall, 38.3% of the 16,165 individuals (47.6% of individuals with T1D and 36.7% of individuals with T2D) were adherent to insulin therapy (PDC ≥80%). Among individuals with T1D, a higher proportion of enrollees in health maintenance organizations – HMOs (54.7%) were adherent to insulin than those in high-deductible health plans (47.0%) or preferred provider organizations (46.3%). We found no insulin adherence differences in individuals with T2D by health plan type.

Conclusions: We found that insulin adherence was higher for individuals with T1D in HMOs. Future studies will examine adherence for elderly Medicare Part D beneficiaries overall and for subgroups that used specific short-acting or long-acting insulins.