The Fight Over Who Pays for an Obesity Drug

On 1 July 2026, American pharmacies began processing a kind of claim Medicare had never paid. Under a demonstration called the Medicare GLP-1 Bridge, eligible beneficiaries with prescription-drug coverage could obtain certain GLP-1 medicines for obesity for a copayment of $50 a month, until the end of 2027 1. The manufacturers had agreed to supply the drugs at a net price of $245 for a month's supply, and the payments ran outside the ordinary Part D drug benefit altogether, counting towards neither deductibles nor out-of-pocket limits 2.

The route was a workaround, and why it had to be one is the subject of this piece. Federal law has barred Medicare, since its drug benefit began, from covering medicines when they are used specifically for weight loss 2. The argument over whether to change that — who should pay for a medicine that a large share of adults are eligible for, how to value it, and who gets it first when a system cannot pay for everyone — has run for years on both sides of the Atlantic. This is an account of that argument. It contains no clinical advice.

A law older than the drugs

The exclusion was written when obesity pharmacology was a small and discredited field, and drugs for weight loss were grouped with those for hair growth and cosmetic purposes as things a public insurer should not buy. The trials that changed the field's reputation are told elsewhere on this site. What they did not change was the statute.

In November 2024 the outgoing administration proposed to reinterpret the law, treating obesity as a chronic disease whose treatment fell outside the weight-loss exclusion, so that Medicare and Medicaid could cover the drugs from 2026. In April 2025 its successor declined to finalise that part of the rule 3. The reason given was cost, and the cost had already been estimated with unusual precision.

In October 2024 the Congressional Budget Office had modelled an illustrative policy authorising Medicare coverage from 2026. It found that federal spending would rise by about $35 billion, net, between 2026 and 2034, with the federal cost per user around $5,600 in 2026, falling to $4,300 by 2034 as prices were expected to decline. The offsetting savings from better health were small inside that window: under $50 million in 2026, rising to about $1 billion by 2034 4.

Good value, and unaffordable

A year later the Institute for Clinical and Economic Review, an independent American body that assesses whether drugs are worth their price, reached what looked like the opposite conclusion. Its December 2025 report found that injectable semaglutide, oral semaglutide and tirzepatide all represented high long-term value for money at current net prices, which it put at $6,829 a year for injectable semaglutide and $7,973 for tirzepatide — below the price ranges it judged the health benefits could justify 5.

The two findings are not in conflict; they answer different questions. A cost-effectiveness analysis asks whether each person treated gains enough health, over a lifetime, to justify what their treatment costs. A budget score asks what a programme costs a particular payer over a particular window — and most of the gains from preventing heart attacks and diabetes arrive after the window closes, often to a different payer. A medicine can be excellent value per person and still unaffordable in aggregate, simply because so many people qualify. The Institute said as much: even at those prices, it wrote, the population in need is so great that the American health system will be strained to provide the drugs affordably to most people who need them 5.

That is the core of the fight, stated without rhetoric. Nobody serious argues that the drugs do nothing. The argument is about who pays for benefits that arrive slowly, to a very large number of people, out of a budget that has to balance this year.

The deal and the workaround

What broke the stalemate in the United States was negotiation rather than legislation. In November 2025 the administration announced agreements with the two manufacturers under which Medicare and Medicaid would pay $245 a month for their GLP-1 medicines. The first vehicle, a model called BALANCE, was designed to run through the private insurers that administer Medicare drug plans, and it needed most of them to take part. They would not. In April 2026, with major plan sponsors reluctant or unwilling to join, the agency said it would not implement the model in Medicare Part D for 2027 2.

The Bridge was the replacement: the government would pay directly, outside Part D, with a prior-authorisation process confirming that each patient met body-mass-index and clinical criteria 2. It is time-limited, and it rests on demonstration authority rather than on any change to the statute. What happens to the people enrolled in it after December 2027 has not been decided.

The British version: eligibility without access

The National Health Service answers the same question through a different institution. The National Institute for Health and Care Excellence recommended semaglutide for obesity in March 2023, but only within specialist weight-management services and only for a maximum of two years — restrictions that reflected both the cost and the thin capacity of those services 6.

Its December 2024 guidance on tirzepatide went further, allowing use in primary care. It also came with a funding variation that NHS England requested and NICE accepted. Rather than funding everyone eligible at once, the service would phase access over the first three years to around 220,000 people, selected by health need and clinical benefit, out of an estimated 3.4 million eligible under the guidance — with up to a further nine years allowed for full implementation. Prescribing in primary care began, in phases, in June 2025 7.

That is the British shape of the problem. An expert body says a medicine is good value; a health service accepts the verdict; and then it rations the medicine by time rather than by price, because it cannot afford to give it at once to everyone the verdict covers. Eligibility and access become two different things.

Outside the NHS a private market absorbed the difference, and discovered its own exposure. In August 2025 Lilly announced that from September its UK list price for tirzepatide would rise by up to 170 per cent, with the price providers paid for the highest-strength pens going from £122 to £330 a month, to bring UK prices in line with other European countries. NHS prices, set under separate agreements, were unaffected 8. Overnight, the cost of the private route to the same medicine was reset by a decision taken largely with other countries in mind.

The equity problem

Every system that rations a scarce or expensive medicine ends up rationing it unequally. In the United States the Institute singled out high out-of-pocket costs and what it called onerous clinical eligibility criteria as barriers likely to worsen disparities in who is treated 5. In England a phased rollout by clinical priority is fairer on paper than first come, first served, but it leaves most of the eligible population waiting while those who can pay privately do not have to.

In most high-income countries obesity and its complications are more common further down the income distribution, so the people least able to pay are, on average, among those with the most to gain. A payer that restricts access by price reaches the wrong end of that distribution last. One that restricts it by queue reaches everyone eventually, but slowly. And wherever the wait is long or the price high, some people go looking for supply outside the regulated system — which is how the payment question and the grey-market question, told in this cluster's companion pieces, turn out to be the same question asked from different ends.

Why the price fell in 2026

In 2026 the price of these medicines fell, for Americans paying cash, faster than at any point since their launch — and not because of any single decision. Four forces arrived at once.

• Negotiated public prices: the November 2025 agreements fixed what Medicare and Medicaid pay at $245 a month.
• Cash-price cuts: the semaglutide manufacturer cut its monthly cash price for existing self-paying patients from $499 to $349 in November 2025.
• A cheaper formulation: an oral version for obesity launched in the United States in January 2026, with its starting strength priced at $149 a month for cash-paying patients.
• Generics abroad: patent expiries in Canada, India, China and Brazil pushed prices down in those markets and put pressure on the originator's pricing everywhere else.

The two consumer moves were announced publicly at the time: the cash-price reduction in November 2025 9 and the oral launch in January 2026 10. Together with the negotiated public prices, they narrowed the gap between what the approved medicines cost and what the compounded and grey-market alternatives had been offering — which may, in the end, have done more to shrink those alternatives than any enforcement action.

None of it settles the argument. Lower prices make the budget arithmetic kinder and the cost-effectiveness case stronger, but they do not change the size of the eligible population, and they do not tell a payer how long a patient will stay on treatment. For a medicine whose effect lasts only while it is taken, that is the variable that decides what coverage actually costs.

Where the story stops

That variable is where the economics becomes technical. Every estimate in this piece — the budget score, the value benchmarks, the phased rollout — rests on assumptions about how long people stay on treatment, what happens to their weight and their cardiovascular risk when they stop, and how far into the future the savings are counted. Change those assumptions and the answer to who should pay changes with them. How those models are built, and which of their assumptions the evidence can actually support, is where a technical article would begin.