🧬 Longevity
Medicare's GLP-1 Bridge Costs $245 a Month. The Break-Even Price Is $150. Nobody Changed the Number.
CMS launched its 18-month GLP-1 demonstration on July 1 at a negotiated MFN price 39% above the fiscal break-even identified in a JAMA pricing model. At base-case enrollment, the gap costs $3.4 billion per year in excess federal spending, and the program sunsets five years before generic semaglutide arrives.
Ninety-five dollars. That is the monthly distance between what Medicare agreed to pay for GLP-1 obesity drugs and the price at which the program would actually save money. On July 1, 2026, the Centers for Medicare and Medicaid Services launched its GLP-1 Bridge Program, an 18-month demonstration covering Wegovy, Zepbound, and the newly approved Foundayo pill for qualifying beneficiaries at a negotiated Most Favored Nation price of $245 per month. Beneficiaries pay $50 out of pocket. CMS absorbs the rest.
Four months earlier, a research team at Brigham and Women's Hospital published the math CMS apparently did not use. In a March 2026 JAMA Network Open study, Thomas Hwang and colleagues modeled the fiscal impact of expanded Medicare GLP-1 coverage across every plausible combination of uptake, adherence, and pricing, running the numbers through a 10-year horizon with three-way sensitivity analysis on the variables that actually determine whether the program pays for itself or hemorrhages money until the demonstration expires. At $245 per month, no scenario produced net federal savings over ten years, not at 10% uptake, not at 30%, not at 90%, because every single cell in their sensitivity table showed net cost to the federal government. At $150 per month with 40% real-world adherence, the picture reversed completely: net savings at every uptake level, from $4.5 billion at 10% enrollment to $30.5 billion at 90%, a clean fiscal win that required nothing more exotic than paying less for the same molecule. Break-even sits right around $150. CMS negotiated $245.
The Gap, in Dollars
CMS director Chris Klomp told reporters the agency expects "single-digit millions" of enrollees. Call it 3 million, roughly 10% of the 30 million Medicare beneficiaries estimated to be clinically eligible. At that scale, here is what the pricing gap costs.
Net federal outlay per enrollee per year: ($245 − $50) × 12 = $2,340, which for 3 million enrollees means $7.02 billion annually in drug acquisition costs alone, and over the full 18-month demonstration window, approximately $10.5 billion in federal spending before a single medical offset is counted. Apply the Hwang model at a negotiated price of $150 per month and 40% real-world adherence, and that same 3 million-person cohort generates $4.5 billion in net savings over a decade through reduced hospitalizations, fewer cardiovascular events, and lower rates of type 2 diabetes conversion, while at the actual negotiated price of $245, the identical cohort produces $5.0 billion in net cost even after accounting for every medical offset the model can find.
| MFN Price | 10% Uptake (3M) | 30% Uptake (9M) | 90% Uptake (27M) |
|---|---|---|---|
| $245/mo (negotiated) | +$5.0B | +$17.6B | +$47.7B |
| $150/mo (break-even) | −$4.5B | −$11.0B | −$30.5B |
| Annual gap (3M) | $3.4 billion/year | ||
Per patient, the difference is $1,140 per year, which across the base-case population translates to $3.4 billion annually in spending above the break-even threshold, or roughly $5.1 billion over the 18-month demonstration window paid to Novo Nordisk and Eli Lilly for the privilege of launching a program designed to demonstrate cost savings at a price point mathematically incapable of producing them.
Why Higher Adherence Makes It Worse
Most drug cost models assume that better adherence means better outcomes and therefore lower total spending. GLP-1s break that logic.
In the earlier Hwang et al. JAMA Health Forum analysis (April 2025), increasing adherence from 40% to 70% at $245 per month and 10% uptake tripled the net fiscal cost from $5.0 billion to $15.8 billion over a decade, because weight loss plateaus after 12 to 18 months of continuous treatment while drug costs march forward at the same rate indefinitely, meaning that patients who stick with the medication keep paying full freight for diminishing marginal weight reduction while CMS absorbs the difference between an escalating cumulative drug bill and a flattening curve of medical savings.
At $150, the math flips: medical offsets outpace drug spend at every adherence level the model tested. At $245, the offsets never catch up, not at 40% adherence, not at 70%, not at any plausible real-world persistence rate. This is not a problem of pharmacology but a problem of arithmetic, and the arithmetic was published four months before CMS set the price.
The Sunset Cliff
Set aside pricing for a moment and consider timing. Medicare's Bridge Program expires December 31, 2027. Generic semaglutide does not arrive until 2032, when Novo Nordisk's exclusivity lapses. That leaves a five-year gap between the end of the demonstration and the availability of affordable alternatives.
What happens when 3 million people stop GLP-1s simultaneously?
We know, because Novo Nordisk ran the experiment themselves. In the STEP 1 extension trial (Wilding et al., Diabetes Obesity Metabolism, 2022), 327 participants who had lost an average of 17.3% of body weight on semaglutide were followed for one year after discontinuation, and they regained 11.6 percentage points, erasing two-thirds of their weight loss in 52 weeks while cardiometabolic improvements in blood pressure, HbA1c, and inflammatory markers all reverted toward baseline as though the treatment had never occurred.
Now scale that to a national program where 3 million beneficiaries lose weight for 18 months, begin accumulating cardiovascular protection (the SELECT trial showed significant CV benefit emerging around month 12), and then face a hard stop because the demonstration expired and generics are half a decade away. Congress would need to extend the program, negotiate a new price, or watch the health gains evaporate in real time across a population the size of Chicago, Houston, and Phoenix combined. Extension without renegotiation locks in $245 for another window. Renegotiation requires leverage CMS has already surrendered by setting the precedent. Expiration creates a population-scale natural experiment in drug withdrawal that no institutional review board would approve if it were proposed as clinical research.
What Went Wrong
CMS did not invent the $245 figure from nothing. Most Favored Nation pricing pegs Medicare reimbursement to the lowest price among six reference countries (Australia, Canada, France, Germany, Japan, and the United Kingdom), but in those markets semaglutide for obesity is either not covered at all or priced under restrictions so narrow that the effective per-patient cost is substantially lower than the list price, which means MFN captures the sticker price rather than the net-of-restrictions cost and delivers a "most favored" number that lands 39% above the domestic fiscal break-even.
An earlier, more ambitious approach existed and collapsed. In April 2026, CMS shelved the Balance program, which would have created a permanent Part D obesity drug benefit administered by commercial insurers, after CVS and UnitedHealthcare both declined to participate, leaving CMS without the insurer infrastructure the program required and forcing a pivot to the Bridge as a stopgap demonstration project requiring only executive authority, no congressional approval, and no insurer cooperation, all of which could be stood up in months rather than years.
Speed came at a cost: a demonstration program cannot negotiate drug prices the way a permanent benefit can, because manufacturers know the clock is ticking and have no incentive to offer concessions on a revenue stream guaranteed to expire. Novo Nordisk and Eli Lilly offered $245 knowing the program ends in 18 months, that CMS had no fallback after Balance collapsed, and that Morgan Stanley estimates 10% uptake alone generates roughly $5.5 billion per year in pharma revenue. For manufacturers, the Bridge is a risk-free revenue window with a built-in expiration date and no renegotiation clause.
Limitations
This analysis relies on Hwang et al.'s modeled scenarios, not on observed Bridge enrollment data, which does not yet exist (the program is three days old). Actual uptake could be far below 3 million if prior authorization requirements and lifestyle program certification requirements suppress enrollment, as similar barriers have done in Medicaid GLP-1 programs in Arkansas and West Virginia. Real-world adherence in a Medicare population with high comorbidity burdens may differ substantially from the 40% and 70% rates modeled in the JAMA studies, which drew primarily from commercial insurance claims data. CMS may also negotiate supplemental rebates not reflected in the published $245 MFN price, though no such rebates have been disclosed. Finally, the CBO's $35 billion estimate over nine years and the Morgan Stanley/BMO revenue projections use different enrollment assumptions and time horizons, and none of these models account for potential congressional action to extend or restructure the program before its December 2027 expiration.
Strongest Counterargument
Fiscal cost is not the only metric that matters, and arguably not even the most important one. Medicare covers chemotherapy regimens costing $150,000 per year that extend median survival by four months. It covers joint replacements at $50,000 that improve quality of life but generate no net savings. Demanding that obesity treatment achieve fiscal neutrality applies a standard that almost no other Medicare benefit is required to meet. If 3 million beneficiaries lose 15% of their body weight, reduce their cardiovascular risk, and avoid type 2 diabetes, the clinical benefit is real regardless of whether the price sits at $245 or $150. A program that costs $10.5 billion and prevents tens of thousands of heart attacks may be worth funding even if a cheaper price would have been better. Letting the perfect price kill the program helps nobody.
What You Can Do
If you are a Medicare beneficiary with a BMI above 30 and a qualifying cardiovascular condition, enrollment opened July 1 and you should contact your Part D plan immediately about Bridge Program eligibility, because prior authorization is required, your physician must certify participation in a lifestyle modification program, and administrative delays consumed months in early Medicaid GLP-1 programs across states like Arkansas and West Virginia.
If you work in health policy or advocacy, the Hwang JAMA model is public, the sensitivity tables are in the supplement, and the $150 break-even figure is fully reproducible, which means that before Congress debates Bridge extension in 2027, the price question needs to be the central issue on the table rather than the coverage question, which the Bridge has already answered.
If you are an investor tracking the pharma upside, watch enrollment velocity against the Morgan Stanley 10% base case closely, because if uptake exceeds projections, the political pressure to extend the Bridge at the current price intensifies dramatically: cutting off millions of enrolled beneficiaries is orders of magnitude harder than letting a small pilot expire quietly, and both Novo Nordisk and Eli Lilly are pricing in a revenue floor, not a ceiling.
The Bottom Line
Medicare launched the largest federal obesity drug program in American history on July 1, and it did so at a price that the published peer-reviewed literature identifies as incapable of generating savings under any modeled scenario. The $95 monthly gap between the negotiated price and the break-even price is not a rounding error but $3.4 billion per year at base-case enrollment, compounding across an 18-month window that ends five years before generics arrive and leaves no bridge to the bridge. When the program expires in December 2027, the weight comes back, the cardiovascular protection dissipates, and the federal government will face a choice between extending a program priced above break-even and watching the health gains of 3 million people reverse in a single calendar year while generic semaglutide sits half a decade away on the other side of a patent wall. That choice will be considerably easier if someone changes the number before December 2027, but three days into the program, nobody has.