Cost-effectiveness analysis is how health systems decide what to pay for, and its output is a single number: the cost of buying one extra year of life in good health. A recent model produced an unusually favorable one for semaglutide, and the interesting part is not the number. It is which assumption the number rests on.
What the model did
The analysis took a US third-party payer perspective and simulated a lifetime horizon for patients with non-cirrhotic metabolic dysfunction-associated steatohepatitis and moderate to advanced fibrosis — a liver condition, not weight loss. [1] It compared semaglutide 2.4 mg and two doses of resmetirom against standard of care, drawing transition probabilities from published trials, costing everything in 2025 dollars and discounting costs and outcomes at 3% a year. That is a conventional and competently specified design, quite unlike the arithmetic a buyer does at a checkout, which has no horizon at all.
In the base case, semaglutide produced 1.66 additional quality-adjusted life years at an incremental cost of $33,031, giving an incremental cost-effectiveness ratio of $19,911 per QALY gained. Resmetirom 80 mg produced 1.20 additional QALYs at an incremental cost of $415,110, and resmetirom 100 mg 1.24 QALYs at $245,991 — ratios of $346,810 and $198,607 per QALY. At a willingness-to-pay threshold of $150,000, semaglutide was cost effective in 99.5% of simulations.
Who built it
Two of the three authors are employees and shareholders of Novo Nordisk, which makes semaglutide. The third is employed by a research firm that Novo Nordisk contracted to develop the model and write the manuscript. The comparator drug, resmetirom, is made by somebody else.
The input that decides everything
An ICER is a fraction. The numerator is the incremental cost of the drug and the denominator is the health it buys. The denominator comes from trials. The numerator comes, largely, from a price — and a price is the input a manufacturer knows best and controls outright. On this roster the same input moves constantly, which is why a twelve-month plan is priced against a moving figure.
The authors evidently understood that, because they ran a scenario analysis using the direct-to-consumer cash price of $499 a month rather than a payer-negotiated cost. That scenario is named in the published abstract; its result is cut off before the figure, so this page does not report one. What the scenario establishes is that the choice of price was known to matter enough to test.
That $499 figure is worth holding next to this roster. Sellers here publish a median of $179 a month for semaglutide by injection, read September 2026 — $320 below the direct-to-consumer price the model used as its consumer-facing scenario. These are not the same product in a regulatory sense, and most of the roster is compounded rather than branded, which is a real difference and not a discount. But it does mean the cheapest input anybody could plug into that model is not the one in it. You can see the spread in the price check.
What a QALY figure is worth to you
Almost nothing directly, and it is worth being clear about why. A QALY threshold is a system-level judgment about what a population should collectively fund. It is not a statement that a given purchase is a good deal for a given person, and a buyer paying cash has no threshold, no lifetime horizon and no discount rate.
What it is good for is calibration. A drug at $19,911 per QALY and a drug at $346,810 per QALY are not close, and that ordering is unlikely to flip on a price assumption alone. Directional conclusions from a model survive better than exact ones — the same rule that applies to weighing the money against the outcome at the individual level. Treat the ranking as informative and the decimal places as somebody’s assumption.