29 January 2026
Keeping watch on Medicare: 2026 maximum fair prices for ten selected drugs
With appreciation for feedback from Mark Fendrick, Director of the Value-Based Insurance Design (V-BID) Center at the University of Michigan; Daneen Sekoni, Vice President of Policy and Advocacy at the Cancer Support Community; and Adina Lasser, Director of Public Policy and Government Relations at the Alliance for Aging Research.
Summary
The Inflation Reduction Act (IRA) changed how Medicare covers and reimburses drugs. For the first time, the federal government has set prices for ten drugs selected through the Medicare Drug Price Negotiation (MDPN) Program, which went into effect on January 1, 2026.
Below, we evaluate the proportion of beneficiaries taking a selected drug and the characteristics of their plans. In addition, we assess formulary coverage in Medicare Part D plans for the selected drugs and their therapeutic alternatives. We also consider how access to affordable medicines may change as a result of the maximum fair price (MFP). Other elements of the IRA, such as the total out-of-pocket (OOP) cap on drug expenses or the $35 monthly limit on insulin, affect beneficiary expenses but are outside the scope of this analysis, which focuses on the MDPN and drug plan premiums.
We find that beneficiary cost sharing for medicines with an MFP will decline in 2026 relative to 2021 (prior to the IRA) in a plan that uses co-insurance rather than co-pay. However, for certain drugs, a plan with a co-pay charges less per prescription than a co-insurance plan for the MFP drug. Further, we find that formulary coverage for therapeutic alternatives to the MDPN-selected drugs is declining and that cost sharing for those treatments is increasing, including medicines for rare diseases. We estimate that a minority of beneficiaries (11 percent) enrolled in Medicare Part D who do not receive a low-income subsidy and are enrolled in a plan that uses co-insurance for non-specialty drugs will spend less per prescription for the ten selected drugs as a result of the MFP. Of that 11 percent, 52 percent are enrolled in a stand-alone drug plan (PDP) in which premiums have increased over the study period from 2021 (pre-IRA) to 2026, which may offset savings from the MDPN.
As such, policymakers may consider the implications of the MDPN for beneficiaries' costs and for federal spending and reflect on modifications to the IRA to help prevent the loss of affordable access to medicines in Medicare.
Figure 1: Proportion of beneficiaries taking a drug subject to the MDPN in 2026 and plan characteristics
Background
The IRA directs the federal government to set the prices of selected medicines in Medicare Part D (i.e., self-administered drugs such as pills), later expanding to Medicare Part B (i.e., physician-administered drugs like intravenous infusions), through the MDPN. Once the MFP price is in effect, that drug must be covered on Medicare Part D plan formularies. Products are removed from having an MFP if they have bona fide marketing from a generic or biosimilar competitor, which has occurred for three of the first ten selected drugs in 2027.[1]
A formulary is the mechanism drug plans use to determine which drugs are covered and how beneficiaries’ cost sharing is structured. In Medicare Part D, “specialty” drugs are allowed to be placed on a specialty tier, typically tier 4 or higher. This means that patients are charged a co-insurance, or a percentage of the non-discounted price of the drug, typically at 25–30 percent. This is a less predictable cost as it is a percentage of the price rather than a fixed amount.[2] In 2021, lower-cost drugs that are not considered “specialty” were more typically placed on a co-pay tier, with a fixed cost of $45–$60 in lower formulary tiers 1–3. However, co-insurance as a cost-sharing structure for non-specialty drugs has increased over the study period since 2021, particularly as it relates to PDPs, as compared to the integrated medical benefit and drug plans (MA-PD).[3], [4], [5]
Beneficiaries can gain more affordable access to their medicines through the MDPN. First, their drug may become available on their plan formulary if the drug was not already available on formulary before being selected for the MDPN. Second, the MFP lowers the list price, which is the basis of co-insurance. The co-pay, however, is not related to the list price, so people in a plan with a co-pay for their MFP drug may not see cost savings from the MDPN. However, Medicare beneficiaries who receive a low-income subsidy (LIS) pay a fixed amount per prescription, which does not change with the MFP. Therefore, they will not experience cost savings from the MDPN.
When a drug is on a plan’s formulary as a result of the IRA, the amount a beneficiary pays OOP counts toward the OOP maximum of $2,100 in 2026. Patients taking specialty drugs with co-insurance are more likely to reach the OOP cap.[6] Beneficiaries who would have exceeded the OOP maximum prior to the IRA may spend less in total. Alternatively, they may spend more if prescription abandonment declines. Nevertheless, as the MFP reduces the list price of the drug, beneficiaries taking a drug with an MFP will reach the OOP maximum more slowly than a person taking a higher-list-price drug.[7]
Plans use other formulary tools to manage drug utilization, including prior authorization, which requires the provider to submit justification for the prescription to the plan for approval, and step edits, which require a beneficiary to use a typically less expensive drug before potentially receiving a costlier prescription from their healthcare provider. The coverage requirement for the MDPN-selected drugs does not preclude a drug plan from using other utilization management tools.
In 2025, the federal government reduced the subsidy provided to plans for beneficiaries who use enough medicines for their spending to exceed the OOP limit, or have “catastrophic costs,” which include many who use specialty medicines.[8], [9] As of 2025, plans were responsible for 60 percent of the beneficiaries' “catastrophic costs,” up from 15 percent when the benefit was established, where a government subsidy covered 80 percent of those higher costs in the Medicare drug benefit and the beneficiary paid 5 percent. The plans can manage this additional expense by (1) increasing premiums and/or reducing access to medicines through narrower formularies and more utilization management and/or (2) increasing patient cost sharing to an extent within the limits of the actuarial value defined for the Medicare Part D benefit.[10]
Methodology: Formulary analysis
We evaluate Medicare formularies for the medications selected for the MDPN and other therapies that treat the same conditions as those drugs selected using the United States Pharmacopeia Medicare model guidelines. We assess changes in formulary coverage, utilization management, and cost sharing as of January 2026, before comparing them to data from January 2025, 2024, and 2021. Data were sourced from the publicly available CMS formulary files.[11], [12], [13], [14], [15]
Findings
Formulary coverage and cost sharing for MFP selected drugs
- Full coverage for selected drugs: Plans are required to cover all selected drugs with a price set by the MDPN; we find all ten selected drugs on 100 percent of formularies in 2026 when the MFPs were established. Prior to the IRA in 2021, eight of the ten drugs were covered on 98 percent or more formularies in 2021.[16]
- Cost sharing – shifting to co-insurance:
- Cost sharing for tier 3 or lower changed over the study period.
- Prior to the IRA in 2021, coverage on tier 3 or below for the MDPN non-specialty drugs was most commonly associated with a co-pay.
- Over time, all plans evaluated increasingly use co-insurance, particularly PDPs for non-specialty drugs. In 2026, 35–36 percent of formularies used a co-pay structure for the selected non-specialty drugs (a change from 78–91 percent in 2021). Lower-tier co-insurance rates have increased over the study period.
- In 2026, for PDPs, only 10 percent of formularies used a co-pay structure for the non-specialty drugs. This is compared to 38–41 percent for MA-PD plans.
- Insulin OOP costs are limited to $35 per month beginning in 2023. Beneficiaries who take one selected drug have cost sharing reduction due to that provision, although formulary data lists a co-pay of $41–$43 in 2024–2026.
- Specialty drugs are labeled tier 4 or higher, with co-insurance in 100 percent of formularies over the study period.[17]
- Drugs with co-insurance will have lower OOP cost per prescription filled from the MDPN, as the list price is reduced. However, total expenses will also depend on the number of prescriptions filled and on reaching the OOP cap.[18]
- For non-specialty drugs, the MDPN reduced the average cost sharing with co-insurance to less than the average cost sharing with co-pay by $5 or more for three drugs. For two non-specialty selected drugs, co-insurance is still greater than the average co-pay.
- Co-pays for selected non-specialty drugs declined by $1–$3 over the study period.
- Utilization management: We find that utilization management was higher in all years of the study period for specialty drugs (75–100-percent prior authorization), was low for non-specialty drugs, and did not change for the majority of MFP drugs. The use of prior authorization declined for three drugs (by 25 percent, 11 percent, and 1 percent). The use of step edits increased for one drug by 17 percent and declined for four drugs by 3 percent and 7 percent.[19]
Table 1: Formulary access selected drugs (analysis unit is an individual formulary)[20]
We compare the co-pay or co-insurance over time for a one-month prescription fill. To calculate the amount a beneficiary would pay with co-insurance, we use the 2023 prices that CMS announced when selecting the drugs for the 2026 MDPN, in years 2021–2025, and the 2026 MFP price, multiplying these prices by the average co-insurance percentage.[1]
Table 2: Drugs selected for 2026 MDPN with estimated cost in co-pay and co-insurance for one month
While all drug plans have begun shifting to co-insurance, the trend is more pronounced in PDPs, where only 10 percent of formularies use a co-pay for non-specialty selected drugs (table 3), compared to 38–41 percent of MA-PD plans using co-pay for these drugs. Co-pays in PDPs are, on average, higher than in MA-PDs by $4–$5, but co-insurance rates are 1–4 percent lower.
Table 3: Drugs selected for 2026 MDPN (PDP vs. MA-PD) with estimated cost in co-pay and co-insurance for one month (the pre-2026 price set at the 2023 price; the 2026 price is the CMS-announced MFP)[21]
Formulary coverage and cost sharing for non-selected drugs: Therapeutic alternatives
We identify all therapeutic alternatives to the selected drugs without approved generics and assess formulary coverage. We find that coverage for non-selected drugs' therapeutic alternatives is declining, particularly for cardiovascular and metabolic diseases such as diabetes and stroke, as well as for gastrointestinal diseases, which are covered less frequently or not at all. If a drug is not covered on the formulary and a patient pays OOP, that does not count toward the OOP maximum.
We find that formulary coverage has declined over the study period in five therapeutic areas (blood glucose regulators by 1 percent, blood products and modifiers by 17 percent, cardiovascular agents by 2 percent, central nervous system agents by 10 percent, and gastrointestinal by 16 percent). Coverage has increased for four therapeutic areas (antineoplastic by 1 percent, dermatological agents by 5 percent, immunological agents by 4 percent, and respiratory tract by 8 percent).
Table 4: Formulary coverage for therapeutic alternatives to drugs selected for MDPN
Of the therapeutic alternatives we evaluated, 23 (or 16 percent) are no longer included on 5 percent or more Medicare drug plan formularies. Of those 23 drugs, 7 (or 30 percent) had orphan indications for immunological or gastrointestinal diseases.
Table 5: Drugs not selected but in the same category as selected medicines, losing 5 percent or more formulary access, by orphan drug designation
Methodology: Proportion of beneficiaries with cost savings from the MFP
The MDPN will have an effect on affordable access to covered medicines for beneficiaries who are (1) taking a selected drug, (2) enrolled in a plan that uses co-insurance rather than co-pay for the selected drug, and (3) not receiving LIS.
We estimate the proportion of beneficiaries affected by the MFP using the 2023 Medicare Part D spending-by-drug data and formulary analysis. We identify the number of people taking a selected drug in 2023 (10.5 million), divided by the Medicare Part D population in 2023 (50 million), then estimate the proportion of those beneficiaries who do not receive LIS (76 percent).[22] We determine the proportion of that group in a PDP (42 percent) or an MA-PD (58 percent) and, for each, determine the proportion in a plan using co-insurance (90 percent for PDP; 61 percent for MA-PD) or co-pay for their non-specialty MFP selected drugs.
Projected effect on beneficiaries
We estimate that 11 percent of all Medicare beneficiaries using the Medicare Part D benefit may experience cost savings on their drugs selected for the MDPN as they are taking a selected drug, are not receiving LIS, and are part of a plan that uses co-insurance for the selected drugs.[23]
Of that group, just over half are in a PDP. PDPs have experienced premium growth, which may offset some savings from the MFP.[24] Sixty-one percent of all beneficiaries are not on LIS and are not taking a selected drug, so they will not receive cost savings from the MFP. However, they may experience higher premiums (particularly in PDPs where 26 percent of beneficiaries not taking MFP drugs are estimated to have coverage), more narrow formularies, and greater cost sharing.
In 2027, three of the ten selected drugs will no longer have a set MFP due to the entrance of a biosimilar or generic. This is estimated at 44 percent of all the beneficiaries taking a selected drug, reducing the group with savings from the 2026 MFP to 6 percent of the Medicare population in 2027.[25]
Figure 2: Distribution of beneficiaries’ expectation of savings from MFP
Conclusion
As Medicare Part D plans shifted to the use of co-insurance for both specialty and non-specialty drugs, a proportion of Medicare beneficiaries experienced lower cost sharing for the drugs selected for the MDPN with the January 2026 implementation of the MFP. However, savings will be limited as three of the selected drugs will no longer have an MFP in 2027, meaning those beneficiaries will benefit from cost savings from a generic or biosimilar, but not from the MFP. Meanwhile, premiums are growing for beneficiaries who do not take a drug with an MFP, resulting in growing costs without offsetting savings. We do not estimate the proportion of beneficiaries who will lose coverage for a selected drug or experience higher cost sharing on a drug not selected for the MFP.
It was estimated that the MFP will result in a $6.4 billion reduction in net Medicare spending for the selected drugs for 2026 in one year. Of that, 32 percent of the savings derives from the three drugs (Entresto, Stelara, and Xarelto) that will no longer have an MFP in 2027.[26] Meanwhile, the Congressional Budget Office has estimated that the federal subsidy to reduce Medicare Part D premiums costs $5 billion in one year.[27]
As the IRA is implemented, other changes – including premium growth, more restrictive formularies, and increasing use of co-insurance – are likely to offset cost savings from the MDPN to beneficiaries and affect a larger portion of the Medicare population than the group that may save money from the MFP. Other provisions, such as the OOP cap, are expected to lower OOP costs for 9.5 million people in 2025.[28] Policymakers and administrators at CMS are encouraged to monitor the evolution of formularies as IRA implementation continues. They may also consider taking action to preserve beneficiary access for both selected and non-selected medicines.
Limitations
The IRA includes other elements that affect beneficiary costs including an OOP cost maximum, changes to the Medicare benefit design and federal subsidies, penalties for price inflation, and price caps on insulins, which are not the focus of this analysis. It is not possible to determine from our analysis whether the changes we observe on formulary are directly related to the MFP, other elements of the IRA, or other changes such as the health of the Medicare beneficiary population, the cost of drugs, or new entrants to the market.
Further, drug plans and manufacturers may have had existing rebate agreements for certain drugs selected for the MDPN. These rebates reduce the net cost of the drug, thereby lowering total plan expenditures, beneficiary costs through lower premiums or cost sharing, and federal costs. If these privately negotiated rebates are reduced or eliminated with the imposition of the MFP, the net cost of the drug may remain unchanged or decrease to a lesser amount relative to the difference between the list price and the MFP. We do not have insight into how rebates may or may not have changed.
Our estimate of the proportion of beneficiaries likely to experience cost savings from the MFP is both an over- and under-projection. We count each beneficiary individually who is taking a selected drug, where in reality there may be overlap (e.g., in the case of a beneficiary taking multiple MFP drugs). Our estimate does not account for changes in consumption if the population of treated beneficiaries is expanded due to the MFP. Beneficiaries in MA-PD with a co-pay structure have received co-pay reductions of up to $3, and we believe the MFP is not reducing these populations’ costs. We do not estimate the total OOP cost for beneficiaries taking other drugs that are not subject to the MFP, which may include drugs whose cost sharing has increased or who have lost formulary coverage. Other effects, such as the OOP limit on insulin, are not considered in forming this estimate. Further, the estimate does not include the effect of the OOP cap, where the cost sharing for their drugs – whether MFP selected or otherwise – is no longer imposed.
When this analysis was conducted, enrollment data by plan for 2026 was not available. As such, we estimate beneficiary distribution based on historical enrollment in MA-PD or PDP. Moreover, we estimate the proportion of the population in specific cost sharing plan types (co-pay or co-insurance) based on the count of formularies, not beneficiaries, which may also be over- or underestimated.
For more information, please contact the authors.
[1] CMS announced the removal of Entresto, Stelara, and Xarelto from the selected drugs list effective January 1, 2027 after determining that at least one approved generic or biosimilar version of each drug is being marketed.
[2] Costing $950 or more per month in 2024.
[3] Avalere, “Medicare Prescription Payment Plan May Help Enrollees Facing More Coinsurance in 2025,” 2025.
[4] We determine that a drug is a “specialty” drug if the cost per claim is over $1,000 in 2022 and if it is on a formulary tier higher than tier 3 more than 90 percent of the time. Using Medicare Part D Dashboard, 2023 is the most recent data available.
[5] PDP is a drug plan that covers only Medicare Part D drug costs, typically the medicines picked up at a pharmacy. An MA-PD is a plan that integrates pharmacy drug coverage and other medical care such as physician and hospital visits into one insurance plan.
[6] In 2024, patients had no catastrophic cost sharing. The OOP began in 2025 at $2,000 and is adjusted annually after 2025 at the projected rate of growth in Medicare Part D costs per enrollee.
[7] Milliman Report, “Expected Impact of Inflation Reduction Act Medicare Drug Price Negotiation Program on Medicare Part D Beneficiary OOP Costs,” 2024.
[8] Explanation of Medicare Part D redesign.
[9] The MA-PD will be more able to manage the additional risk across their medical and prescription drug spending as a result of other changes in the IRA where the PDP drug plan has only the drug benefit to manage. Avalere’s August 24, 2023 analysis provides an explanation of the risk shifting.
[10] Actuarial value explanation in CMS Factsheet.
[11] We evaluate 331 formularies in 2021, 300 in 2024, 260 in 2025, and 201 in 2026. This includes formularies from 4,539 plans in 2021; 4,671 plans in 2024; 4,189 in 2025; and 3,729 in 2026. CMS Formulary Files.
[12] We collected all unique Concept Unique Identifier (RXCUI) codes, which is a system for uniquely identifying drugs, to then identify all unique National Drug Codes (NDC) codes. The NDC codes were then used to identify whether a drug is covered on a given formulary (using the plan information data and basic drug formulary data). Useful explanation provided by NIH and Duke University.
[13] When a drug is “on formulary,” that means it is covered within the plan’s benefit and expenses for that drug count toward deductibles and other out-of-pocket limits. “Prior authorization” means an additional approval by the plan provider is required to receive the drug that is prescribed by the health provider. For an explanation see here.
[14] We matched these drugs to the formulary file with NDC codes identified through Prescription Drug Data Collection Code (RxDC).
[15] We used the USP Medicare Model Guidelines 9.0 to identify the therapeutic class for each drug.
[16] The remaining two, Farxiga and Fiasp, had been covered on 76 percent and 64 percent respectively in 2021 prior to the IRA, and had 97 percent coverage and 66 percent coverage respectively in 2025 prior to the application of the MFP.
[17] There are multiple FDA approved biosimilars to Stelara.
[18] Adherence may be affected by the OOP costs; other studies have indicated that cost is a barrier to adherence, meaning that if patients fill more prescriptions their total costs may go up.
[19] Jardiance, where no formulary had step edits in 2021 or 2025 and 19 percent of formularies used step edits for the drug in 2026. 20 percent of MA-PD used step edits for Jardiance compared to 11 percent for PDPs.
[20] The cost sharing structure is either co-pay or co-insurance. We report only the percent of formularies using co-pay; the balance is co-insurance.
[21] Medicare Drug Price Negotiation Program: Negotiated Prices for Initial Price Applicability Year 2026.
[22] KFF. Key Facts About Medicare Part D Enrollment, Premiums, and Cost Sharing in 2024.
[23] We find that 10.5 million people take a selected drug based on the CMS announcement of the 2026 MFPs. We use KFF data to project LIS and non-LIS enrollment by plan type.
[24] Beneficiaries in PDPs are experiencing higher premiums on average. We find that PDPs are $22 higher on average in 2026 relative to 2021 (data on file).
[25] Entresto, Stelara, and Eliquis will not have an MFP in 2027.
[26] Brookings, “Impact of federal negotiation of prescription drug prices.”
[27] CBO letter to House Committee on the Budget, Energy and Commerce and Ways and Means, dated October 2, 2024.
[28] ASPE, “Medicare Part D Enrollee Out-Of-Pocket Spending: Recent Trends and Projected Impacts of the Inflation Reduction Act Inflation Reduction Act,” 2023.
