Medicare patients who take certain physician-administered drugs could see their share of the bill fall sharply under a mandatory pricing model the Centers for Medicare & Medicaid Services finalized this week. The agency’s own illustration shows 94 percent of the drugs it modeled carrying coinsurance of between 2 and 12 percent, compared with the standard 20 percent in Original Medicare Part B. The model reaches only a slice of enrollees, roughly a quarter of those in Original Medicare Part B, chosen by geography rather than by choice.
The 2% to 12% coinsurance illustration against the standard 20%
The Global Benchmark for Efficient Drug Pricing Model, known as the GLOBE Model, was published as a final rule in the Federal Register on October 2, 2026 and takes effect November 30, 2026. In it, CMS writes that “94 percent of illustrative GLOBE Model drugs would have a beneficiary coinsurance percentage between 2 and 12 percent, well below the standard 20 percent coinsurance.”
The wording matters. The figure comes from an illustration that CMS built on 2024 data for a list of example drugs, not from a guarantee about any particular medicine. Actual coinsurance will depend on how the model’s benchmark prices compare with what Medicare pays for each drug once the model runs, so the 2% to 12% range describes what the agency modeled, and a given patient’s bill could land differently.
The same rule estimates that the model would produce $177 million in total out-of-pocket savings for beneficiaries, alongside $440 million in Medicare Part B net spending savings over the seven-year payment period. The Part B spending figure is the subject of separate coverage; the beneficiary-side number is the one that tracks the coinsurance illustration.
A quarter of Part B enrollees, chosen by random ZIP Code Tabulation Area
The coinsurance relief would not apply to everyone in Original Medicare. The rule describes a cohort of “approximately 25 percent of beneficiaries who are enrolled in OM Part B.” Which quarter is decided by place: according to the rule, “the GLOBE Model geographic areas will be determined through a random selection of ZIP Code Tabulation Areas (ZCTAs).”
Two patients with the same diagnosis, the same drug and the same Part B coverage could therefore face different coinsurance because one lives in a selected area and the other does not. The rule describes no application, enrollment step or opt-out for patients. Selected beneficiaries are included by where they live.
CMS’s own announcement of the rule frames the timing cautiously. In its press release, the agency says selected beneficiaries “may begin to see reduced out of pocket costs beginning April 1, 2027.” The word “may” carries weight, because the savings depend on the drugs a patient is actually prescribed in a covered area.
Because the cohort is defined by ZIP Code Tabulation Area, a patient cannot sign up for the lower coinsurance or decline it. The practical question for each patient is whether the home address falls in a selected area and whether the prescribed drug is on the list of GLOBE Model drugs. The rule’s own illustration covers a sample of drugs from its Table 4, so a medicine missing from that sample may behave differently, and neither the 2% to 12% band nor the 94 percent share should be read as a promise about every Part B drug.
Who pays under the model: manufacturers, not doctors
The model is built around drugmakers. According to the rule, it “will require mandatory participation for all manufacturers” of Part B rebatable drugs that are GLOBE Model drugs. Those manufacturers owe rebates when the Medicare price runs above a benchmark derived from prices in 19 reference countries: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, France, Germany, Ireland, Israel, Italy, Japan, the Netherlands, Norway, South Korea, Spain, Sweden, Switzerland and the United Kingdom.
Doctors, hospitals and other providers are not asked to change how they bill. The rule states that providers and suppliers who furnish GLOBE Model drugs “will not be model participants” and will continue to buy and bill for the drugs as usual. For a patient, that means the office visit and the billing process look the same; the difference, where it occurs, shows up in the coinsurance calculation.
The spending threshold that determines which products fall under the model is covered in a separate Financial Wire report.
The 2027 to 2034 timeline: effective in November, payment period from April
Three dates apply. The final rule is effective November 30, 2026. The CMS press release lists a model start of January 1, 2027. The payment period, per the rule, “will begin on April 1, 2027 and will end on March 31, 2034,” a seven-year span, and the April date is the one CMS ties to any lower out-of-pocket costs for patients.
The rule is cited as 91 FR 62936 under docket CMS-5545-F and runs from page 62936 to page 63116, according to the Federal Register’s own record of the document. That length reflects how much of the model, including the drug list, the benchmark method and the rebate mechanics, sits in the rule text and not in the summary figures.
The coinsurance range is a projection built on 2024 prices, and the rule’s 2% to 12% band remains CMS’s illustration of the benefit, to be tested against actual claims once the payment period opens.
Tracking Part B drug costs when coinsurance depends on ZIP code
The Medicare Cost & Coverage Protection Kit is written for Medicare beneficiaries who need to keep prescription and treatment costs organized and understand what coverage rules apply to them. The GLOBE Model’s illustrated coinsurance range of 2% to 12% is tied to random ZIP Code Tabulation Areas, which leaves a patient’s own cost for a given drug uncertain until bills arrive.
The Medicare Cost & Coverage Protection Kit includes a medication and cost tracker and a guide to the new Part D out-of-pocket cap, two items that help a patient keep a running record of what each drug actually costs.
Click here to get The Medicare Cost & Coverage Protection Kit →
This article was written with AI assistance and verified line by line against the primary records linked in it.



