Spending on Insulin by U.S. Payers and Patients From 2008 to 2017

Insulin prices have risen rapidly over time, making costs a major public health concern. Understanding annual spending net of discounts—rather than the list price, which does not reflect amount paid—may inform strategies to contain insulin-related spending. We examined trends in U.S. per-beneficiary spending on insulin from 2008 to 2017, including pre- and post-discount expenditures by payer type and out-of-pocket costs.

We analyzed data from the Medical Expenditure Panel Survey (MEPS), a nationally representative survey of the U.S. population (1). We estimated per-beneficiary spending on insulin, including only individuals with a single insurance plan, stratified by major payers: commercial, Medicare, and Medicaid. We applied previously published discount percentages on insulin (derived using gross and net prices from SSR Health’s prescription drug price database), stratified by Medicaid and non-Medicaid, to list prices (2,3). To improve precision, we pooled data in 2-year intervals. Costs were inflation adjusted to 2017 dollars (4). All analyses were conducted in R, accounting for complex survey design. t tests were used to compare 2008–2009 with 2016–2017 spending and test significance of trends; α of 0.05 was used for statistical significance.

Our unweighted study sample included 2,392 individuals on commercial insurance, 1,946 on Medicare, and 1,398 on Medicaid. Among commercial beneficiaries, mean age was 51 years, 50% were men, 77% were White individuals, 17% were Black individuals, and 5% were members of other races. Among Medicare beneficiaries, mean age was 67 years, 51% were men, 73% were White individuals, 21% were Black individuals, and 6% were members of other races. Among Medicaid beneficiaries, mean age was 49 years, 38% were men, 70% were White individuals, 24% were Black individuals, and 6% were members of other races.

Before discounts, per-beneficiary spending on insulin by commercial payers increased by 184.1% over the study period, from mean (SE) of $2,098 (195) in 2008–2009 to $5,962 (654) in 2016–2017 (P < 0.001). This increase was not significant after accounting for discounts ($1,762 [164] to $2,251 [246], P = 0.09) (Fig. 1A). Among Medicare beneficiaries, per-beneficiary spending before discounts increased by 275.6%, from $1,669 (236) to $6,267 (798) (P < 0.001); the increase was smaller but still significant after discounts ($1,405 [197] to $2,406 [302], P = 0.01) (Fig. 1B). Medicaid per-beneficiary spending before discounts increased by 141.3%, from $1,703 (342) to $4,108 (679) (P < 0.001), but demonstrated a nonsignificant decline after discounts ($748 [153] to $452 [64], P = 0.07) (Fig. 1C).

Out-of-pocket spending per beneficiary did not change significantly among individuals with commercial insurance ($807 [104] to $842 [140], P = 0.84) or Medicare ($806 [91] to $622 [66], P = 0.11), but it decreased by 57.4% ($701 [140] to $299 [65], P = 0.01) for individuals with Medicaid (Fig. 1A–C).

Per-beneficiary spending on insulin increased substantially for Medicare and commercial payers from 2008 to 2017. While rising rebates offset a considerable portion of increases, Medicare net spending per patient still rose 8 times faster than general inflation from the 2008–2009 period to the 2016–2017 period (4). In contrast, per-beneficiary spending for Medicaid net of discounts remained stable (and possibly declined), while out-of-pocket spending decreased significantly.

Our findings illuminate the marked difference between discounts available to Medicaid and non-Medicaid payers. Because Medicaid obtains obligatory discounts through regulation, it benefits from larger price reductions for branded drugs than commercial and Medicare plans that negotiate discounts on a case-by-case basis (5). Medicaid’s success at containing drug prices is likely also explained by increasing inflation rebates that penalize steep increases in list prices for insulins (5). Expanding these inflation rebates to Medicare and commercial insurance, as is being currently considered, could help contain spending. We find that while insulin list prices have soared, insured patients have, on average, been insulated from rising out-of-pocket costs. However, our results should not be interpreted as saying that all patients are fully protected from rising costs: increased expenditures may be passed on through increased insurance premiums or tax contributions, and some patients may experience financial toxicity related to clinical needs or economic circumstances.

Survey data may be affected by response bias. Our discount estimates rely on prior analyses that did not distinguish among non-Medicaid payers (2) and therefore may not fully capture differences in discount rates between Medicare and commercial payers. Limitations of available data precluded analyses of per-beneficiary spending by payer for individuals with multiple insurance plans. Our results are not generalizable to uninsured patients. Future studies should examine whether after-discount costs differ among demographic subgroups.

In conclusion, insulin-related spending increased for Medicare and commercial insurance between 2008 and 2017 but declined for Medicaid. Insured patients have, on average, been insulated from rising out-of-pocket costs. Further policy interventions are needed to control total insulin spending for Medicare and commercial beneficiaries.