Confidential agreements between the Trump administration and drugmakers could wipe out most of the Medicare savings expected from new rules that tie U.S. drug payments to prices abroad, according to a modeling study in The Lancet published September 13. Excluding medicines made by the first 17 companies with deals would cut projected savings by 71%, the researchers estimated.
The lead author says the loss could now be larger. After a new round of deals announced on August 31 raised the reported total to 26 companies, Dr. Thomas Hwang of Brigham and Women's Hospital updated the estimate to nearly 80%, according to the journal's press release. That figure is his calculation after publication, not a result reported in the paper.
The findings are projections, not observed outcomes. But they matter to the roughly 68 million people on Medicare and to taxpayers, because the promised savings depend on which medicines the rules actually cover, and the deals that determine that remain private.
Rebates Tied to 19 Countries Drive the Estimate
The policy works through two proposed Medicare models. The GLOBE model covers drugs given in doctors' offices and hospital outpatient settings under Part B, and the GUARD model covers pharmacy drugs under Part D. Both would require manufacturers to pay extra rebates when Medicare's price for a medicine is higher than a benchmark drawn from 19 wealthy countries, including Canada, Germany, Japan, and the United Kingdom.
The approach goes further than earlier reforms. Medicare was barred from negotiating drug prices until the 2022 Inflation Reduction Act allowed negotiations on a small number of costly medicines. These models would instead align Medicare payments for many brand-name drugs with prices in comparable countries, adjusted for purchasing power.
The researchers analyzed 195 patented medicines that account for $87.9 billion in yearly Medicare spending. Because actual U.S. discounts are confidential, they estimated Medicare's net prices and compared them with prices in the reference countries. Without exemptions, they projected savings of $5.2 billion, or 16%, under GLOBE and $6.4 billion, or 18%, under GUARD during the initial phase, which would apply to a randomly selected 25% of beneficiaries over five years.
If expanded to all Medicare enrollees, those savings could reach $21 billion and $25.5 billion. The lowest price used as an international benchmark was 71% below what Medicare pays, and South Korea, Norway, and Australia most often set the reference price, the researchers found.
The evidence has clear limits. This is a peer-reviewed simulation built on estimated net prices and public foreign price data that may miss other countries' confidential discounts. It cannot predict how companies will respond, and its medicine list may not include newly approved drugs. The study was funded by Arnold Ventures, and the authors disclosed consulting and expert-witness work.
Exemptions for 26 Companies Shrink the Savings
The 17 companies with reported deals as of June account for 131 of the 195 medicines studied, or 67%. Removing those drugs would leave about $3.3 billion in initial-phase savings: $0.9 billion under GLOBE and $2.4 billion under GUARD. The White House has said its approach would save Medicare $26 billion over several years, STAT reported.
Hwang said that if manufacturers can avoid the models by striking side deals, "most of those savings might not be realised." Because the agreements are not public, he said, it is hard for policymakers and the public to judge whether the policy delivers what was promised.
Co-author Dr. Aaron Kesselheim, also of Brigham and Women's Hospital, said the models' reach is limited, noting that "their scope is limited by various exemptions and will likely face legal challenges," according to the press release. The terms of the individual deals, including exactly which drugs they shield, have not been disclosed.
Effects Could Reach Patients Abroad and Medicare Enrollees at Home
The model also points to effects outside the United States. For 101 of 138 medicines with available sales data, the estimated Medicare cut was larger than the drug's entire annual sales in the country used as its benchmark. That gives companies a financial reason to raise prices abroad, turn discounts into confidential rebates, develop different formulations, or delay launches in reference countries, changes that could affect when patients in those countries get new treatments. Co-author Dr. Kerstin Vokinger of ETH Zurich and the University of Zurich urged policymakers to make sure U.S. rules do not delay important medicines elsewhere, and Hwang noted that reference countries have limited budget room to absorb higher prices.
For U.S. households, the effects would come through program costs and patients' share of costs. The proposed GLOBE rule would lower coinsurance for affected Part B drugs for beneficiaries in the model group, so smaller savings could mean less relief for some patients who receive infused or injected medicines in clinics. Medicare enrollees taking expensive brand-name drugs, including people with cancer or autoimmune conditions who rely on Part B infusions, have the most at stake.
Nothing changes at the pharmacy counter today. The proposed rules call for GLOBE to begin October 1, 2026, and GUARD to begin January 1, 2027, but more specific details were still pending when STAT published its report. Enrollees should not stop or switch medications because of this study. They can compare plan drug coverage during Medicare open enrollment, which runs October 15 to December 7, with free counseling available from State Health Insurance Assistance Programs.
Key unknowns include the final rule text, which drugs the 26 agreements exempt, and whether legal challenges will delay the models. The researchers were scheduled to present the findings on September 14 at the Nordic Pharmaceutical Forum Summit, and MedicalDaily will report on the final rules when CMS releases them.
Key Questions Answered
What did the Lancet study find? It estimated that Medicare's proposed most-favored-nation models could save $5.2 billion and $6.4 billion in their initial phase, but that exempting medicines from 17 companies with confidential deals would cut those savings by 71%.
Where does the nearly 80% figure come from? Lead author Dr. Thomas Hwang calculated it after the reported number of companies with deals rose to 26. It is an updated estimate, not a result published in the paper.
Is this based on real Medicare spending changes? No. It is a peer-reviewed simulation using estimated prices, and it cannot predict how drug companies or insurers will actually respond.
When would the Medicare pricing models start? The proposed rules list October 1, 2026, for GLOBE and January 1, 2027, for GUARD. Final details had not been released as of September 13.
Will my Medicare drug costs change right away? No immediate change has been announced. The GLOBE proposal would lower coinsurance for some Part B drugs for beneficiaries in its model group, but final details are pending.
Could this affect patients in other countries? The researchers say companies could respond by raising prices or delaying launches in benchmark countries, which could slow access to new medicines there.
