Medicare Part D insulin coverage: formulary strategies amid policy headwinds

Abstract

Introduction

Diabetes is a chronic condition impacting over one-quarter of Medicare beneficiaries,¹ with more than 3.3 million of those beneficiaries requiring insulin to manage their blood sugar levels.² Insulin has been notably expensive, particularly for Medicare beneficiaries with Part D drug coverage who, prior to 2025, entered certain phases of the benefit where cost-sharing was higher.^2,3 Insulin's necessity coupled with its financial burden^3,4 have made this drug class the target of many policies in recent years for Medicare beneficiaries, such as the Part D Senior Savings (PDSS) Model test implemented between 2021 and 2023 and the 2022 Inflation Reduction Act (IRA).

One group of policies has focused on lowering out-of-pocket costs. The Centers for Medicare and Medicaid Innovation's PDSS Model tested a $35 maximum copayment for insulin, which the IRA expanded in 2023 to all insulins and plans.^5,6 Typically, plans encourage the use of preferred drugs through favorable tier placement (a tier with lower cost-sharing) to increase volume in exchange for higher manufacturer rebates, or to encourage generics over brands. However, the IRA's $35 insulin copayment cap limits Part D plans' ability to steer beneficiaries toward preferred insulins using formulary tiering and cost-sharing strategies.

With less flexibility in cost-sharing, plans may move insulins into 1 tier, rather than having nonpreferred insulins in higher tiers. Consolidating insulins into fewer tiers may mean certain insulins move to a higher tier with higher cost-sharing, or the reverse. To maintain negotiating leverage over manufacturer rebates, plans could increase their reliance on utilization management (UM) tools (eg, quantity limits, prior authorization, and step therapy requirements). While helping plans manage costs, these strategies may impede beneficiary access to insulin, despite the copayment cap.

In addition to the copayment cap, the IRA introduced broader changes to Part D's standard benefit design to further reduce out-of-pocket costs that may also affect insulin coverage. These provisions included eliminating out-of-pocket spending in the catastrophic phase starting in 2024 and, in 2025, eliminating the coverage gap (also known as the doughnut hole) and capping out-of-pocket costs at $2000 annually.⁷ These changes shift more financial responsibility for the Part D benefit to plans, particularly in the catastrophic phase. Consequently, plans may implement strategies to slow beneficiaries' progression to the catastrophic phase. For example, they could use UM tools to slow the use of the most expensive insulins to reduce beneficiaries' spending.

This study describes how Part D plan formulary coverage for insulin has changed from 2020 to 2025. We examine several outcomes capturing elements of the strategic formulary design process. We separate results between Medicare Advantage plans that combine medical and drug coverage (Medicare Advantage Prescription Drug Plans [MAPDs]) and stand-alone Prescription Drug Plans (PDPs), because MAPDs cover both medical and drug benefits. The PDPs also do not have certain financial tools to offset formulary generosity that are particular to MAPDs; thus, formulary strategies may differ between these 2 plan types.

Data and methods

Data sources

We identified insulins using the publicly available insulin lists from the Pharmacy Quality Alliance's Star Ratings Measure for adherence to non-insulin diabetes drugs, which includes a list of insulins to exclude insulin users from this measure.⁸

We compiled publicly available Part D formulary files from 2020 to 2025, encompassing 1 year prior to and 3 years of PDSS and the IRA implementation. These files include information on covered drugs, cost sharing, tier placement, and any UM tools at the formulary and plan level.⁹ We counted unique insulins at the RxCUI (RxNorm Concept Unique Identifier) level, which groups all National Drug Codes for a given brand, strength, and dose form of an active ingredient, since this is the unit used in the formulary files. We use the term “drug” in the paper to refer to the RxCUI level.¹⁰

Formularies are often designed and implemented at a higher level than a plan, commonly the region or plan-sponsor (parent organization) level. We used the plan level as the unit of analysis because each formulary may be used by multiple plans in a given year and thus using formulary as the unit of analysis may underrepresent the market availability. We limited our analysis to general enrollment plans available to all beneficiaries in a county, excluding employer plans or other limited enrollment plans, such as special needs plans.

Outcomes

The $35 maximum copayment may be limiting plans' ability to steer beneficiaries toward preferred insulins on lower formulary tiers. We used descriptive methods to examine 2 outcomes related to beneficiary steering. The first is the percentage of insulins that a plan covered on each tier each year, which we anticipated will become more consolidated over time, although some plans may use copays lower than $35 to steer towards preferred insulins.

The second outcome is the average percentage of insulins with at least 1 UM tool applied per plan. With more limited ability to steer beneficiaries using cost sharing, plans may increase their use of UM tools to both steer patients and use as leverage with manufacturers in rebate negotiations. The UM tools included in the data are quantity limits (resulting in more frequent fills), prior authorization requirements (whether the insurer must approve coverage for the drug before it is dispensed), and step therapy requirements (requiring the patient to try another, usually lower-cost, drug before filling the drug in question).

We examined trends in UM tools by insulin type, which refer to the speed at which the insulin takes effect after administration, such as rapid or long-acting. Part D plans are required to cover at least 2 chemically distinct agents in a drug class.¹¹ Clinically, plans aim to cover some of each type, and their coverage decisions are likely made within each insulin type. As in prior studies,¹² we combined rapid/short and intermediate/long-acting insulins, resulting in the following insulin-type categories: (1) rapid/short, (2) intermediate/long, (3) mixes, (4) combinations, and (5) concentrates. Combinations, Xultophy (insulin degludec/liraglutide, Novo Nordisk) and Soliqua (insulin glargine/lixisenatide, Sanofi-Aventis), are long-acting insulins combined with a glucagon-like peptide-1 receptor agonist (GLP-1). Concentrates, Humulin R U-500, are concentrated rapid/short-acting insulins for patients who would otherwise require multiple daily insulin injections.

Results

The average number of insulins covered on MAPD formularies increased from 24 (SD = 5) in 2020 to 29 in 2025, peaking at 32 (SD = 8) in 2024 (Table 1). The PDPs covered fewer insulins, on average, but peaked in 2024 at 28 (SD = 9). The maximum number of covered insulins varied annually, ranging from 40 to 77, depending on plan type and year. Novo Nordisk's discontinuation of Levemir in the United States at the end of 2024 contributed, in part, to the decrease in total number of covered insulins for all plans in 2025.¹³

By 2025, most insulins were covered on Tier 3 in both MAPDs and PDPs (Table 1). For MAPDs, the average percentage of insulins covered on Tier 3 grew from 82% in 2020 to 92% in 2025, with the entire increase occurring between 2024 and 2025. For PDPs, the average percentage of insulins covered on Tier 3 increased from 87% to 95% over the study period, with the remainder only in Tier 5 in 2025. The MAPDs shifted insulins on lower tiers up and upper tiers down, while PDPs mainly moved insulins down to Tier 3. Moving insulins from Tiers 1 or 2 to Tier 3 may increase beneficiary cost sharing, but on average, copayments were likely to decrease since some insulins moved down tiers. For both MAPDs and PDPs, Humulin R U-500 was the main insulin covered on the specialty drug tier (commonly Tier 5), although there was some variation across plans (data not shown).

For MAPDs, the average percentage of insulins with a UM tool applied per plan decreased from 23% in 2020 to 19% in 2025 (Table 1). Among PDPs, the percentage of insulins with a UM tool remained relatively stable at approximately 12%–14%, with a temporary increase to 15% in 2024.

Use of UM tools was concentrated in 2 insulin types: combinations and concentrates. Over 90% of the insulins with quantity limits were combinations (Figure 1), and by 2025, 100% of these drugs faced quantity limits in PDPs. Prior authorization was most commonly required for concentrates, although the average percentage of concentrates requiring prior authorization per plan declined for both MAPDs and PDPs in 2025.

Figure 1.

Average percentage of covered insulins with UM tool per plan, combinations and concentrates: 2020–2025. The concentrates category includes insulin regular (Humulin R U-500). The combinations category contains insulin degludec/liraglutide (Xultophy) and insulin glargine/lixisenatide (Soliqua). Abbreviations: MAPD, Medicare Advantage Prescription Drug Plan; PDP, Prescription Drug Plan; UM, utilization management. Source: Authors' analysis of Part D plan formulary data.

Open in new tabDownload slide

Discussion

By 2025, both MAPDs and PDPs typically placed their insulins on Tier 3. This consolidation may suggest a strategic response to the IRA's $35 maximum copay and diminished incentives to place preferred insulins on lower formulary tiers. The PDSS likely did not impact tiering from 2021 to 2023 because participating plans could still exempt certain insulins from the model's $35 copay maximum. While the IRA's copay limit began in 2023, plans did not have time to adjust formularies for the 2023 plan year due to the IRA's passage late in 2022. As shown in Table 1, tiering changes were particularly pronounced for PDPs in 2024 and 2025, and for MAPDs in 2025.

The MAPDs apply UM tools to more insulins than PDPs, possibly because they cover more insulins on average, with quantity limits used most often. The MAPDs have specific mechanisms enabling them to offset the costs associated with larger formularies. For example, MAPDs bidding below their local benchmark for providing Parts A and B services can receive a portion of the difference as a rebate scaled according to their plan's quality rating (Star Rating).¹⁴ These rebates must be spent for the benefit of enrollees; plans often use rebate dollars to buy down Part D premiums that might be higher with larger formularies. An MAPD with a broader formulary could potentially increase or at least maintain enrollment in the competitive Medicare Advantage market. Additionally, MAPDs may emphasize larger formularies for chronic disease medications which could reduce downstream medical costs, since MAPDs are responsible for both medical and drug costs.^15,16

Both MAPDs and PDPs imposed quantity limits for the most expensive insulins—the combinations (Xultophy and Soliqua).¹⁷ Plans may be attempting to reduce the number of beneficiaries progressing to the catastrophic phase by requiring more frequent fills where beneficiaries incur cost sharing each time. However, a comprehensive review of coverage trends is needed due to the significant role that manufacturer rebates play in coverage decisions for this drug class. Manufacturer rebates, which are calculated as a percentage of list price, are more prevalent in drug classes with multiple branded and unbranded alternatives like insulin.^18,19 Given that insulin rebates and other discounts have been estimated to be as high as 84% of the list price (or negotiated price with the pharmacy in Part D),^20-22 plans might favor newer, higher-priced insulins to maximize rebate revenue, which they use to lower premiums for beneficiaries. Limiting spending on combination insulins would require increased coverage of unbranded and biosimilar insulins, which has remained low due to lower list prices and consequently lower manufacturer rebates relative to the reference products.¹² Monitoring changes in covered insulins will be important to ensure beneficiary access in the future.²³

This paper uses descriptive methods to examine formulary coverage for insulin and thus cannot causally distinguish changes in insulin coverage due to the specific policies mentioned herein vs other policies or pharmaceutical market dynamics that might affect insulin coverage. Furthermore, this study does not examine insulin utilization trends, which will also be important to monitor. Finally, we did not examine other drug classes, so it may be the case that plan sponsors are generally moving drugs to a single tier, and this is an important area for future study.

Conclusion

This study highlights the strategic adjustments insurers may be making in response to recent policy changes. Examining descriptive trends in insulin coverage in Part D during a period when multiple insulin-related policies took effect, we observed that plans are consolidating insulins into Tier 3 but are not broadly changing their UM tool strategies.

Formulary strategies may continue to evolve. In 2026, Part D plans must cover Fiasp and Novolog, as these insulins have been selected for the Medicare Price Negotiation program.²⁴ Consequently, future research should continue to monitor insulin and non-insulin diabetes medication coverage trends, and spending, as these dynamics will impact total spending for beneficiaries, plans, and taxpayers.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

Funding for this research was provided by gifts from RAND supporters and income from operations.

Data availability

Part D formulary data are publicly available at data.cms.gov.

Notes

Author notes