Patient and Payer Incentives to Use Patented Brand-Name Drugs vs Authorized Generic Drugs in Medicare Part D

Key Points

Questions  Are authorized generic versions of insulin and direct-acting antiviral agents for hepatitis C covered by Medicare Part D plans, and do they provide value for patients, plans, and the Medicare program compared with the brand-name drug?

Findings  This cross-sectional study of Medicare formulary and pricing data for quarter 3 of 2020 found that most Part D plans offered brand-name drug–only or both brand-name and authorized generic drug coverage. Although Medicare beneficiaries who filled insulin and direct-acting antiviral prescriptions would spend less out of pocket with authorized generic drug use, plans and Medicare would likely spend more on these authorized generic drugs than brand-name drugs for beneficiaries with total drug spending below Medicare’s catastrophic threshold.

Meaning  This study suggests that authorized generic drugs available before patent expiration may lower patient spending but are unlikely to provide savings for Part D plans or Medicare, giving them little incentive to offer authorized generic drugs over brand-name drugs.

Importance  In response to scrutiny over high drug prices, manufacturers of insulin and direct-acting antiviral agents for treating hepatitis C have recently introduced authorized generic alternatives to their patented brand-name products. These authorized generic drugs have list prices at least 50% lower than the list price of the brand-name drugs, which should result in savings to patients. However, it is unclear whether these authorized generic drugs are offered on Medicare Part D formularies because they may not provide savings to plans or Medicare.

Objective  To assess Medicare Part D formulary coverage for 4 brand-name formulations of insulin and direct-acting antiviral agents and their authorized generic formulations.

Design, Setting, and Participants  This cross-sectional study used Medicare Prescription Drug Plan Formulary and Pricing Information Files from quarter 3 of 2020 and Medicare Part D plan enrollment for September 2020. Four patented brand-name drugs (sofosbuvir and velpatasvir fixed-dose combination tablets [Epclusa], ledipasvir and sofosbuvir tablets [Harvoni], insulin lispro [Humalog], and insulin aspart [Novolog]) and their authorized generic formulations for all Part D stand-alone prescription drug plans (n = 959) and Medicare Advantage prescription drug plans (n = 3148) were studied.

Main Outcomes and Measures  Beneficiary-weighted formulary coverage of brand-name and authorized generic products; beneficiary out-of-pocket costs; and prerebate plan, manufacturer, and Medicare spending on brand-name and authorized generic products.

Results  In quarter 3 of 2020, 97% of beneficiaries were in plans that covered brand-name drugs only or both brand-name and authorized generic drugs; approximately 3% were in plans that covered authorized generic drugs only. Observed authorized generic drug list prices were 67%, 62%, and 50% lower than list prices for Epclusa, Harvoni, and each brand-name insulin product, respectively. Medicare beneficiaries using authorized generic drugs could save $270 per year for 12 vials of Humalog and $2974 for a full course of Harvoni. Plans, however, have limited incentives to encourage authorized generic drug use because rebates for brands likely exceed savings available with authorized generic drugs, particularly for beneficiaries with spending that reaches the Medicare Part D coverage gap.

Conclusions and Relevance  The results of this cross-sectional study suggest that authorized generic drugs for insulin and direct-acting antiviral agents may lower out-of-pocket spending for patients but are unlikely to provide savings for Part D plans or Medicare. Instead, these drugs allow manufacturers to offer products at a lower list price without materially lowering net prices or profits.

Authorized generic drugs have been produced by brand-name drug manufacturers for decades but remain an understudied aspect of the prescription drug market.^1,2 With the exception of its labeling, an authorized generic drug is identical to the corresponding brand-name drug. Unlike traditional generic drugs, which are marketed after the brand-name drug’s patents and regulatory exclusivities have expired, authorized generic drugs can be introduced at any time under the brand-name drug’s existing New Drug Application.^2,3

Authorized generic drugs are typically introduced by brand-name drug manufacturers in the days or weeks before traditional generic drug entry is expected (ie, loss of market exclusivity) and are meant to compete with traditional generic products as they enter the market.^2,4,5 These products are designed to provide a revenue stream for manufacturers who see their brand-name drug sales decrease with generic drug substitution. More recently, however, several manufacturers have introduced authorized generic drugs during a period in which the brand-name drug faced no immediate threat of generic drug competition (eg, traditional generic drugs for Harvoni [ledipasvir and sofosbuvir tablets] and Epclusa [sofosbuvir and velpatasvir fixed-dose combination tablets] are not expected to be approved until approximately 2030⁶). This strategy has been used for epinephrine injectors (EpiPen) in 2016, hepatitis C treatments (Harvoni and Epclusa) in 2018, and insulin (Humalog and Novolog) in 2019 and 2020. In each case, the product’s manufacturers faced congressional investigation and public criticism regarding their high launch prices or substantial list price increases over time,^7-9 which might have motivated the timing of authorized generic drug entry.⁵ Indeed, manufacturers indicated in press releases^10-12 that these authorized generic products were offered, in part, to provide insured patients access to a lower list price option that reflected discounts that their health care plans received on the brand-name drug.

Despite list price reductions of at least 50% from the brand-name drug’s list price,^10-13 authorized generic drug uptake by patients and payers may be limited. Specifically, plans may decide not to offer authorized generic drugs if the net price for brand-name drugs (ie, the price after rebates paid by manufacturers to plans to encourage or reward use of brand-name drugs) is lower than the net price of the authorized generic drug. Given the number of brand-name competitors for insulin and hepatitis C drugs and the expectation that rebates are largest for drugs used to treat conditions where therapeutic substitutes are available, rebates that exceed 50% are likely for these medications.^14,15 Incentives to adopt authorized generic drugs may also be particularly low among private plans that administer the Medicare Part D program because the share of spending that plan sponsors must pay decreases as total drug spending increases (Figure 1).

In this study, we assessed Medicare Part D formulary coverage for 4 brand-name drugs (sofosbuvir and velpatasvir fixed-dose combination tablets [Epclusa], ledipasvir and sofosbuvir tablets [Harvoni], insulin lispro [Humalog], and insulin aspart [Novolog]) and their authorized generic drugs. We selected these drugs because we believe they are the only ones that had authorized generic formulations launched more than 1 year before expected patent expiration and faced no traditional generic competition as of the third quarter of 2020. The manufacturers of these products also faced intense scrutiny over their brand-name drug’s prices or price increases, which may have motivated their early introduction of authorized generic formulations. We estimated potential savings for Medicare Part D beneficiaries using the authorized generic vs the associated brand-name product in 2021, focusing on beneficiaries without low-income subsidies to limit their out-of-pocket spending. To highlight changes in financial incentives faced by patients, plans, manufacturers, and Medicare across a range of drug spending, we also estimated prerebate spending for brand-name drugs and authorized generic drugs, assuming that authorized generic drugs received a 50% lower list price compared with the brand-name drug.

We used the Prescription Drug Plan Formulary and Pricing Information Files¹⁶ from quarter 3 of 2020 to summarize Medicare Part D formulary coverage and point-of-sale prices for authorized generic and brand-name versions of sofosbuvir and velpatasvir fixed-dose combination tablets (Epclusa), ledipasvir and sofosbuvir tablets (Harvoni), insulin lispro (Humalog), and insulin aspart (Novolog) for all Part D stand-alone prescription drug plans (n = 959) and Medicare Advantage prescription drug plans (n = 3148). Because coverage did not differ by drug dose, we included a representative dose for each product using the default dose recorded in the Medicare Part D plan finder. We also included both pen and vial formulations for each insulin product. Because authorized generic drugs cannot be uniquely identified in the Medicare formulary files when traditional generic drugs are also available, we focused on 4 products for which authorized generic formulations but not traditional generic formulations were available as of quarter 3 of 2020. Because these files are publicly available for download and not human participants research, this study did not undergo institutional review board approval. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

For each product, we assessed the relative frequency of 4 formulary coverage patterns across all Part D plans: only the brand-name drug, only the authorized generic drug, both, or neither. We weighted our analyses by Part D plan enrollment in September 2020¹⁷ to capture the number of beneficiaries who faced each coverage option. No data were missing for the key variables used in our study.

Next, for covered products we summarized point-of-sale prices and patient out-of-pocket spending for a full (Epclusa and Harvoni) or annual (Humalog and Novolog) course of treatment. The point-of-sale price represents the amount paid by the patient and payer before application of rebates and approximates the product list price. We used the quarter 3 of 2020 formulary file to obtain the mean per-fill price at in-area retail pharmacies for a product offered on the same plan during the prior quarter. This value is aggregated by the Centers for Medicare & Medicaid Services and is expressed as a per-unit cost. For insulin, we multiplied the per-unit cost to reflect a standard fill (eg, 10-mL vial of 100 U/mL and 5 pens of 3 mL of 100 U/mL); for direct-acting antiviral agents for hepatitis C, we assumed 84 units (pills) in total. We calculated out-of-pocket spending for each treatment, applying the 2021 Part D standard benefit. This benefit included a $445 deductible, 25% coinsurance during the initial coverage phase (until total drug spending reached $4130) and coverage gap (for total drug spending of $4130-$10 045), and 5% coinsurance during the catastrophic coverage phase (for total drug spending >$10 045). Because most plans used copayments instead of coinsurance for insulin filled during the initial coverage phase, we replaced the standard 25% coinsurance during that phase with the median plan copayment for insulin (ranging from $0 to $45, depending on the product) to avoid overestimating out-of-pocket spending.

To highlight plan incentives to select brand-name drugs over authorized generic drugs, we estimated prerebate spending by patients, plans, manufacturers, and Medicare for a brand-name drug with a $500 list price and an authorized generic drug with a 50% lower list price filled under the initial coverage phase (where patients pay 25% and plans pay 75%), the coverage gap (where patients pay 25%, plans pay 5%, and manufacturers pay 70%), and the catastrophic coverage phase (where patients pay 5%, plans pay 15%, and Medicare pays 80%) of the Medicare Part D benefit (Figure 1).

To provide greater context to potential motivations for plans to cover brand-name drugs over authorized generic drugs, in sensitivity analyses we estimate the minimum rebate needed for Part D plans to achieve the same net price for the brand-name drug and the authorized generic drug (thus making plans indifferent to which one beneficiaries use) across a range of drug list prices ($500-$100 000) and accounting for prior drug spending.

In quarter 3 of 2020, coverage of hepatitis C treatments and insulins varied across plans. Most beneficiaries were in plans that covered brand-name drugs only or both brand-name and authorized generic drugs; few were in plans that covered authorized generic drugs only (Figure 2¹⁸; eTable 1 in the Supplement). Exclusion of both the brand-name drug and authorized generic drug from formularies ranged from 0.2% for the hepatitis C drug Epclusa to 54% for the Humalog pen and Humalog vial. Among plans that covered the brand-name drug, authorized generic drug, or both, coverage of brand-name drugs only ranged from 47% for Epclusa to 87% for Novolog vials or pens. By contrast, coverage of authorized generic drugs only was rare, with only 3% to 4% of beneficiaries in plans that covered authorized generic drugs alone for any product studied. Between 10% and 60% of beneficiaries were in plans that covered both the brand-name and authorized generic product when at least 1 of the products was covered.

When considering prices per 84-day course of Epclusa or Harvoni (Table 1), we found that point-of-sale prices for the authorized generic drug were 67% lower than for the brand-name drug for Epclusa and 62% for Harvoni. Compared with the corresponding brand-name drug, the point-of-sale price for the authorized generic drug was 48% lower for 1 year of insulin vials to 51% lower for pens. Even with the lower point-of-sale prices for authorized generic drugs, these products remained expensive, with expected total costs for an 84-day course of treatment of roughly $25 000 with Epclusa and $37 000 with Harvoni and between $3300 and $6800 for 12-monthly fills of insulin. As expected, patients experienced lower out-of-pocket spending with authorized generic drugs than with brand-name products (Table 1). Out-of-pocket savings ranged from $270 per year for 12 vials of Humalog insulin to $2974 per course of Harvoni.

When comparing estimated prerebate spending by plans on a brand-name drug with a list price of $500 vs an authorized generic drug with a 50% lower list price, we found that plan incentives to select a brand-name drug change as drug spending increases. Specifically, in the initial coverage phase, plans would have spent $187 more on the brand-name drug than the authorized generic drug, before accounting for rebates (Table 2). In this case, if plans could have negotiated a brand-name drug rebate of at least 37.5%, they would have spent less on the brand-name drug than the authorized generic drug. For the same drug prescription filled in the coverage gap, where plans are responsible for only 5% of drug spending, plans would have spent only $13 more on the brand-name drug than the authorized generic drug, before accounting for rebates. In the coverage gap, plans would only need to have negotiated a 2.5% rebate on the brand-name drug to ensure that they spend less on the brand-name drug than the authorized generic drug. Finally, in the catastrophic coverage phase, plans would have paid $38 more and Medicare $200 more for the brand-name drug, before accounting for rebates. In this phase, plans and Medicare together would have needed a rebate of 47.5% to make the brand-name drug less costly than the authorized generic drug.

In sensitivity analyses that vary input drug prices, the level of previously accumulated drug spending at the time the brand-name or authorized generic drug prescription was filled (eTable 2 in the Supplement), and expected discounts for authorized generic and brand-name drugs (eTable 3 in the Supplement), we found similar patterns to those described above. Specifically, we found that among scenarios tested, rebates needed to make the brand-name drug less costly than the authorized generic drug were lowest for beneficiaries with spending under the catastrophic threshold and in the coverage gap and highest when spending occurred primarily during the catastrophic coverage phase. A similar pattern emerged when considering list price reductions of 30% or 70% for the authorized generic compared with the brand-name drug.

In this cross-sectional study, compared with brand-name drugs, authorized generic versions of direct-acting antiviral agents for hepatitis C and insulin would have resulted in lower out-of-pocket spending for Medicare beneficiaries. However, authorized generic drug coverage was limited for some Part D beneficiaries, with many beneficiaries in plans with coverage for only brand-name drugs. For Part D plan sponsors, this decision likely reflects rational economic behavior because net prices (after rebates) for brand-name drugs in these classes may be similar to or lower than net prices for authorized generic drugs. This behavior is particularly true in cases where drug spending would reach the coverage gap because plans are responsible for only a small percentage of drug spending in that phase. For context, expected per-beneficiary spending on insulin alone was nearly $4000 in Medicare Part D in 2016,¹⁹ making it likely that many insulin users would fill at least some of their insulin prescriptions in the coverage gap. On the other hand, were a product filled entirely under the catastrophic phase (once patient out-of-pocket spending exceeds $6550 or approximately $10 000 in brand-name drug spending in 2021), plans would have needed brand-name drug rebates of 47.5% to have similar net prices for the brand and authorized generic drugs. Given their high list prices, users of Harvoni or Epclusa would have been responsible for substantial spending in the catastrophic phase of the benefit, suggesting that plan incentives to offer coverage for only the brand-name drug may have been slightly lower for these products. Even assuming that net prices were similar between brand-name and authorized generic drugs, plans may have preferred the brand-name drug if it resulted in lower use of a high-priced product (ie, patients may use less if facing a higher out-of-pocket cost), lower health plan premiums, or the ability to negotiate bundled rebates across multiple products (eg, rebate walls).^20-22

Although plans have limited incentives to encourage authorized generic over brand-name drugs in the context studied, patients have financial incentives (ie, lower out-of-pocket spending) to select authorized generic drugs when offered by their plans. Despite this, the proportion of claims filled in Part D that were for the authorized generic drug were only 8.3% for Harvoni and 12.7% for Epclusa in 2019, suggesting that even beneficiaries in plans that cover the authorized generic drug may face other barriers to filling prescriptions for these products. For example, an investigation by US senators Elizabeth Warren and Richard Blumenthal suggested that there were substantial barriers to patients accessing insulin because many pharmacies were unaware of or did not stock the authorized generic version of the drug.²³ For patients to take advantage of these lower list prices, they must be aware that the authorized generic drug is available, be on a plan that provides coverage for the product, and find a pharmacy that has this product in stock. These requirements represent substantial barriers to getting authorized generic drugs to patients.

For the products considered, the manufacturers have noted that their motivation for releasing an authorized generic product was to provide patients with treatment options that more closely reflect the price of the drug paid by the patients’ plan.^10-12 They do this by offering an alternative to the brand-name drug at a lower list price, a way to provide rebates at the point of sale, thus lowering patient out-of-pocket expenses. Despite this, manufacturers face no real financial loss for offering an authorized generic product, at least for insulin and hepatitis C drugs, because these drugs are almost certainly at or near the brand-name drug’s net price. In fact, manufacturers could benefit from increased use of authorized generic drugs over brand-name drugs if their use resulted in lower drug spending in the coverage gap (where manufacturers pay 70% of the brand-drug’s price) or through increased sales if patients are able to fill more prescriptions because of lower costs. Manufacturers also gain reputational benefits by offering a drug at a steeply discounted list price. Plans and Medicare, on the other hand, face greater potential financial loss for patients using authorized generic drugs in this context and would eventually pass along any increased spending to beneficiaries in the form of higher premiums.

The potential for patient savings associated with authorized generic drug use highlights broader concerns regarding how Medicare Part D plans cover drugs with high list prices and large rebates. This potential has led to proposals for plans to provide rebates at the point of sale, when beneficiaries would pay a percentage of the estimated net price (after rebate) rather than a percentage of a drug’s prerebate price. Although this is conceptually appealing, it is important to recognize that not all drugs receive large rebates, including many high-priced drugs offered under Medicare Part D. As a result, the benefits of this proposal would be limited to users of prescription drugs that have sufficient competition to generate large rebates. Perhaps more problematic, estimates from the Congressional Budget Office suggest that providing rebates at the point of sale under Part D would increase federal spending by $170 billion throughout 10 years. This increased spending is largely driven by the assumption that manufacturers would not reduce drug list prices by the full discounts received by plans and pharmacy benefit managers today,^14,24 consistent with testimony from industry representatives shortly after the proposed rule was issued. Indeed, if spending increased as projected, it is likely that some Medicare Part D beneficiaries would see their premiums increase.^14,24

Current Medicare Part D reform plans advanced by the US Congress and recommended by the Medicare Payment Advisory Commission would alter Part D’s benefit design to help avoid incentives for plans to select drugs with higher list prices and larger rebates. However, even under the redesigned benefit options that have been advanced, the standard benefit specifies the use of coinsurance (ie, a percentage of list price) instead of flat-fee copayments for drugs filled. For plans that choose to use coinsurance over copayments, beneficiaries would continue to pay more for the brand-name drug than the authorized generic for the products considered in this study. When plans elect to cover a drug with a higher list price rather than one with a lower list price (as documented for the products studied), they should consider offering the higher list-price product under flat-fee copayments or limit cost sharing to the amount that would have been paid for the lower-priced product. This approach would allow plans to select the drug with the lowest net price without patients overpaying for their plan’s choice.

This study has some limitations. First, the study focuses on a small number of authorized generic products launched well before traditional generic drug entry. We believe that the products constitute the subset of brand-name drugs that had authorized generic formulations launched more than 1 year before traditional generic drug entry and for which no traditional generic drug competitors had entered the market by quarter 3 of 2020. Second, our results generalize only to beneficiaries covered by Medicare Part D who do not receive low-income subsidies. Medicare’s unique design, including the manufacturer coverage gap discounts and shifting financial responsibility from patients and plans to the Medicare program as spending increases, may present unique challenges for authorized generic drug adoption. Third, we were unable to include actual rebates for brand-name and authorized generic drugs. Because patient out-of-pocket spending under Part D is calculated based on point-of-sale prices, this limitation did not affect our key measure of out-of-pocket spending. We also assumed that there were no additional rebates for authorized generic drugs and thus that lower prices were achieved through list price reductions only.

Ultimately, the availability of authorized generic options for products not facing traditional generic competition is insufficient to improve affordability for Medicare beneficiaries. The results of this study suggest that efforts to modify the Medicare Part D benefit should also ensure that incentives for plans and beneficiaries are well aligned and that beneficiaries do not overpay for drugs that are a better value for their health plan and the Medicare program.

Accepted for Publication: August 27, 2021.

Published Online: October 18, 2021. doi:10.1001/jamainternmed.2021.5997

Corresponding Author: Stacie B. Dusetzina, PhD, Department of Health Policy, Vanderbilt University School of Medicine, 2525 W End Ave, Ste 1203, Nashville, TN 37203 (s.dusetzina@vanderbilt.edu).

Author Contributions: Dr Dusetzina had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: Dusetzina, Sarpatwari, Keating, Huskamp.

Drafting of the manuscript: Dusetzina.

Critical revision of the manuscript for important intellectual content: Sarpatwari, Carrier, Hansen, Keating, Huskamp.

Statistical analysis: Dusetzina, Sarpatwari, Keating, Huskamp.

Obtained funding: Dusetzina, Huskamp.

Administrative, technical, or material support: Sarpatwari, Carrier.

Conflict of Interest Disclosures: Dr Dusetzina reported receiving grants from Arnold Ventures and the Commonwealth Fund during the conduct of the study; grants from Robert Wood Johnson Foundation and the Leukemia and Lymphoma Society; and personal fees from Institute for Clinical and Economic Review, West Health, and the National Academy for State Health Policy outside the submitted work. Dr Dusetzina also serves as a commissioner for the Medicare Payment Advisory Commission (the views expressed are those of the authors and not of the commission). Dr Sarpatwari reported receiving grants from Arnold Ventures during the conduct of the study and personal fees from West Health outside the submitted work. Dr Hansen reported receiving personal fees from Takeda Pharmaceuticals for expert testimony outside the submitted work. Dr Keating reported receiving grants from Arnold Ventures and the Commonwealth Fund during the conduct of the study and grants from the National Cancer Institute and contract salary from the Centers for Medicare & Medicaid Services outside the submitted work. Dr Huskamp reported receiving grants from the Commonwealth Fund and Arnold Ventures during the conduct of the study. No other disclosures were reported.

Funding/Support: This work was funded by the Commonwealth Fund and Arnold Ventures (Drs Dusetzina, Huskamp, and Keating).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.