The copay that leaves no trace: Medicare's new $50 GLP-1 deal charges the sick and the poor the most

On July 1, for the first time, MedicareMedicareThe US federal health insurance program mainly for people 65 and older. began covering weight-loss drugs. The Medicare GLP-1GLP-1A class of drugs (like Wegovy and Zepbound) that mimic a gut hormone to reduce appetite and are used for weight loss and diabetes. Bridge, a federal demonstration running through December 2027, lets eligible beneficiaries fill Novo Nordisk's Wegovy, the KwikPen version of Eli Lilly's Zepbound, or Lilly's new Foundayo pill, the oral GLP-1 [approved April 1](https://www.prnewswire.com/news-releases/fda-approves-lillys-foundayo-orforglipron-the-only-glp-1-pill-for-weight-loss-that-can-be-taken-any-time-of-day-without-food-or-water-restrictions-302731485.html), for a flat $50 a month. Manufacturers supply the drugs to the government at a [net price of $245 a month](https://www.kff.org/medicare/what-to-know-about-the-balance-model-for-glp-1s-in-medicare-and-medicaid/), a fraction of list prices that have run near $1,000. The first prescriptions are being dispensed this week.

The launch coverage has treated the $50 as what it looks like: a discount. This article asks a narrower question. Who pays more under the Bridge than they would have paid if the same drugs, at the same $245 price, had gone into the regular Medicare drug benefit, and by how much per person per year?

To see why those sentences matter, you need one piece of plumbing. Medicare Part DMedicare Part DThe part of Medicare that pays for prescription drugs, run through private insurance plans., the drug benefit covering [more than 50 million people](https://www.kff.org/medicare/key-facts-about-medicare-part-d-enrollment-premiums-and-cost-sharing-in-2025/), keeps a running meter of what each enrollee spends out of pocket, the deductible first (up to [$615 in 2026](https://www.medicare.gov/health-drug-plans/part-d/basics/costs)), then roughly 25 percent coinsurancecoinsuranceA percentage of a drug's price you pay instead of a fixed amount, such as 25 percent of each fill. per fill. Since the Inflation Reduction Act's redesign, the meter has a hard stop: once out-of-pocket spending on covered drugs reaches $2,100, every additional covered drug is free for the rest of the year. Think of it as a punch card: every dollar punches the card, and a full card means you stop paying.

The Bridge is the first flat-copaycopayA fixed dollar amount you pay for each prescription fill, like $50 per month. drug deal Medicare has offered that is booked entirely off that card. It runs under Section 402 demonstrationSection 402 demonstrationA 1967 legal authority that lets the government test new ways of paying for care without changing the underlying benefit. authority, a 1967 statute letting the HHS Secretary test ['changes in methods of payment or reimbursement'](https://www.cmscmsThe Centers for Medicare and Medicaid Services, the federal agency that runs Medicare..gov/medicare/coverage/prescription-drug-coverage/medicare-glp-1-bridge) without changing the benefit, and CMS states plainly that it ['will operate outside of the Medicare Part D benefit's coverage and payment flow.'](https://www.cms.gov/medicare/coverage/prescription-drug-coverage/medicare-glp-1-bridge/information-part-d-plans) Your Part D plan never sees the claim; a central processor pays it. So the $50 never punches the card. It does not count toward the deductible or the $2,100 cap, cannot be smoothed through Medicare's new Prescription Payment PlanPrescription Payment PlanA Medicare option that lets enrollees spread their yearly drug costs into equal monthly installments., and is not reduced by Extra Help, the low-income subsidy that normally caps a poor enrollee's brand-name copay at [$12.65 per prescription in 2026](https://www.phlp.org/en/news/what-you-need-to-know-about-medicare-extra-help-and-part-d-costs-in-2026).

Not only do the costs not count toward your out-of-pocket cap, your deductible or anything like that, if there is a problem, you need to go to Medicare.

We can price the alternative precisely, because the alternative was written down and then shelved. The Bridge exists only because the BALANCE modelBALANCE modelThe postponed CMS plan that would have covered these same weight-loss drugs inside Part D starting in 2027., the CMS plan that would have put these same drugs inside Part D in 2027, [was postponed on April 21](https://www.aha.org/news/headline/2026-04-22-cms-delays-part-d-portion-balance-model-expansion-glp-1-access) after Part D plan sponsors ['were reluctant or unwilling to participate'](https://www.kff.org/medicare/what-to-know-about-the-balance-model-for-glp-1s-in-medicare-and-medicaid/) and CMS's 80 percent participation threshold went unmet. BALANCE's published cost-sharing schedule is the counterfactual: $245 per fill in the deductible phase, then $50 a month in enhanced plans or $125 in basic plans, then $0 once the cap is reached, with every dollar punching the card and Extra Help stacking on top.

Read left to right. For a healthy beneficiary with no other prescriptions in a basic plan, the Bridge is a genuine bargain: $600 against roughly $1,740. For a typical enrollee whose deductible is met, the sticker is identical, BALANCE's enhanced-plan copay was the very same $50, except that inside Part D the $600 would have pulled every other medicine she takes closer to free.

Then the gradient flips. A beneficiary sick enough to reach the $2,100 cap on her other medicines, an estimated [3.2 million such people outside the low-income subsidy in 2025](https://www.aarp.org/pri/topics/health/prescription-drugs/medicare-part-d-out-of-pocket-spending-cap-prescription-drug-costs/), per AARP's Public Policy Institute, within [about 11 million total cap-reachers](https://aspe.hhs.gov/reports/impact-ira-2000-cap) counted by HHS, pays $600 a year for a drug Part D would have handed her for $0 at the margin. An Extra Help enrollee, whose subsidy would have capped the copay at $12.65 a fill (about $152 a year), pays the full $600: roughly four times as much, drawn from the smallest incomes in the program.

For lower-income beneficiaries, a recurring $50 charge outside the Part D cap and outside LIS protections is significant.

The crossover is computable from published parameters. A year on the Bridge costs $600. Inside Part D, the drug's cost-sharing would count toward the $2,100 cap, so its marginal costmarginal costWhat one additional unit costs you; after the cap is hit, the marginal cost of another covered drug is $0. can never exceed $2,100 minus what you already spend on other medicines. The Bridge therefore flips from bargain to surcharge once other out-of-pocket drug spending passes $2,100, $600 = $1,500 a year in 2026. In 2027 the cap rises to $2,400 and the line moves to $1,800, but the copay stays $50, so the wedge persists by design.

There is a second-order penalty the chart cannot show. The Bridge's $600 buys no ledger credit for itself and fails to push the enrollee toward the cap on everything else. Someone $200 short of the cap would, inside Part D, fill that gap with GLP-1 copays within months and then take all her drugs free; on the Bridge she pays $600 and her other medicines keep costing full cost-sharing until they exhaust the gap on their own. For anyone near the cap, the design is strictly dominated by the benefit Medicare already runs.

The fairest reading of the off-ledger design is administrative necessity. BALANCE died because plans would not carry the risk; Section 402 routes claims around the plans entirely, which is why ['Part D sponsors will not carry risk... and do not have to opt in'](https://www.cms.gov/medicare/coverage/prescription-drug-coverage/medicare-glp-1-bridge/information-part-d-plans), there is no participation threshold left to miss. The ledger is plan-administered plumbing: a copay that never passes through a Part D plan has no natural way to post to that plan's accounting. And keeping the program's [$1.3 billion to $3.3 billion projected cost](https://www.kff.org/medicare/nearly-four-million-medicare-beneficiaries-could-be-eligible-for-the-temporary-medicare-glp-1-bridge-program-covering-these-drugs-for-weight-loss/) outside Part D shields all 50-plus million enrollees from premium pass-through.

The competing reading is that the accounting is the product. The off-ledger structure concentrates the program's savings on low-utilizers and its costs on the two populations the IRA cap was written to protect, and it does so silently: no public material walks an enrollee through the arithmetic above. The exclusion appears as a technical note, not as the pivotal fact that determines, person by person, whether $50 a month is a good deal. CMS chose the only channel that did not need the plans, and the channel came with an accounting shadow. It did not choose to tell the 3.8 million people standing in it.

Both readings are live. The test between them is documentary: whether CMS's Bridge guidance or the BALANCE postponement record treats the TrOOP exclusion as legally forced by Section 402's limits or as a discretionary design decision. We could not find a primary document that settles it either way, which is itself notable for a program of this size.

Flat, round-number copays on politically salient drugs are a familiar product. What is new is the accounting. The Part D Senior Savings Model, launched in 2021, offered insulin at ['a flat $35 copayment during all phases of the benefit'](https://www.kff.org/medicare/insulin-costs-and-coverage-in-medicare-part-d/), through more than 1,750 participating enhanced Part D plans, inside the benefit. The IRA's statutory $35 insulin cap, effective January 2023, went further in the opposite direction from the Bridge: under [CMS guidance](https://www.cms.gov/files/document/frequently-asked-questions-medicare-part-d-insulin-benefit.pdf), not only does the $35 count toward the ledger, the amount the enrollee would otherwise have owed also counts toward the deductible and true out-of-pocket costs, actively speeding insulin users toward catastrophic coverage. The insulin cap subsidizes the sick twice; the Bridge copay, once booked, simply vanishes.

The closest structural ancestor is the 2004-2005 Medicare drug discount card, the last time CMS ran a drug program outside an actual benefit. It, too, was explicitly a bridge, ['a transitional program... while preparations are made for implementation of the Medicare drug benefit under Part D in 2006'](https://www.cms.gov/newsroom/fact-sheets/medicare-prescription-drug-discount-card-and-transitional-assistance-program), and it carried a [$600 annual credit for low-income enrollees](https://www.cms.gov/newsroom/press-releases/600-credit-and-discounts-combine-give-low-income-medicare-beneficiaries-significant-discounts), the very population today's Bridge charges full freight. The analogy breaks at the destination: the discount card had a benefit waiting at the far end. The GLP-1 Bridge's intended successor is indefinitely postponed; the span now ends December 31, 2027, over open water.

• Cap-hitting beneficiaries: roughly 3.2 million non-LIS enrollees reached the out-of-pocket cap in 2025 per AARP, within about 11 million total counted by HHS. Any of them who joins the Bridge pays $600 a year for a drug that would be free at the margin inside Part D.
• Extra Help enrollees: the low-income subsidy does not apply, so the poorest participants pay about four times what BALANCE would have charged them ($600 vs roughly $152).
• The BMI-30 eligibility tier: the Bridge reaches down to BMI 30 only with stage 3 kidney disease, heart failure, or uncontrolled hypertension on multiple drugs, conditions that come with long medication lists, meaning the program may select for the very people its accounting treats worst.
• Healthy, higher-income eligibles: the clear winners, saving up to about $1,140 a year versus a basic-plan counterfactual.
• Manufacturers and taxpayers: Novo Nordisk and Eli Lilly gain a new federal channel at $245 a month; Medicare bears $1.3-3.3 billion at baseline take-up, paid outside Part D premiums.

Note what the eligibility screen does to the exposed population. KFF's [3.8 million eligible count](https://www.kff.org/medicare/nearly-four-million-medicare-beneficiaries-could-be-eligible-for-the-temporary-medicare-glp-1-bridge-program-covering-these-drugs-for-weight-loss/), out of 13 million meeting BMI thresholds, excludes people with type 2 diabetes, sleep apnea, and MASHMASHMetabolic dysfunction-associated steatohepatitis, a serious fatty liver disease; having it routes patients out of the Bridge and into regular coverage rules., and current Part D GLP-1 users, stripping out many of the heaviest drug utilizers. That purifies the comparison but also means the share of participants above the $1,500 line is genuinely unknown.

• Base case: take-up lands in KFF's 10-25 percent band and skews toward healthier, lower-utilizing enrollees for whom the Bridge is a real discount; the surcharge is real but affects a minority. Evidence that would confirm it: early enrollment data showing participants' other drug spending clustered well below $1,500.
• Upside: CMS or Congress patches the accounting, counting Bridge copays toward TrOOP, restoring LIS cost-sharing help, or reviving BALANCE for 2028, before the demonstration expires. Evidence to watch: any successor rulemaking or revised Bridge guidance addressing TrOOP treatment before December 31, 2027.
• Downside: enrollment concentrates in the BMI-30 comorbidity tier with long medication lists, so a substantial share of participants pay $450-600 a year above the shelved alternative, and the whole channel expires at the end of 2027 with no successor, leaving them a coverage cliff. Evidence to watch: first-fill demographics and the 2027 plan year showing no BALANCE revival.

• The counterfactual is a shelved design: BALANCE's cost-sharing schedule was published but never implemented, and actual plan behavior under it can never be observed.
• The $1,740 archetype assumes the statutory-maximum $615 deductible and nine months at the $125 basic-plan copay; enrollees in enhanced plans or plans with lower deductibles would have paid less.
• No take-up or enrollment-composition data exists yet; the demonstration's stated purpose is to generate it.
• How many of the 3.8 million eligible exceed $1,500 in other out-of-pocket drug spending is not published; the harmed population is bounded, not measured.
• The 3.2 million and roughly 11 million cap-reacher figures are 2025 projections built for the $2,000 cap, applied here to 2026's $2,100 threshold.

What is already certain, four days into the program, is the shape of the deal. Medicare's first weight-loss benefit charges every participant the same $600 a year and returns ledger credit to none of them, worth the most to those who need the least other care, break-even for the typical enrollee, and $450 to $600 a year more than the shelved alternative for the cap-hitting sick and the Extra Help poor. The punch card was Medicare's promise that the sickest stop paying. The Bridge is the first deal Medicare has ever sold that never touches it.