The Medicare GLP-1 Bridge: Three Reasons Prior Authorizations Get Denied
Your patient has been asking about Wegovy for two years. She is on Medicare, she does not have diabetes, and every previous conversation ended the same way. Then on July 1, 2026, that changed.
The Medicare GLP-1 Bridge opened a coverage pathway for GLP-1 medications prescribed for weight management. It is a real pathway with a $50 monthly copay, and it runs through December 31, 2027.
It also has criteria specific enough that a thorough, well-written note can still produce a denied request. Three things account for most preventable denials, and all three are avoidable if you know about them before you send the prescription. This article covers what the program is, who qualifies, and those three failure points, so your first submission is your only submission.
What is the Medicare GLP-1 Bridge?
The Medicare GLP-1 Bridge is a CMS demonstration program that created a coverage pathway for GLP-1 medications prescribed to reduce excess body weight and maintain weight reduction. It runs July 1, 2026 through December 31, 2027. Beneficiaries pay a $50 monthly copay. Prescribers submit prior authorization to a central processor rather than to the patient's Part D plan.
That last point is the difference worth holding onto. Every other prior authorization you submit goes to the patient's plan. This one does not. It goes to a central processor, on its own submission route, which is why the workflow your office already has does not simply extend to cover it.
The covered drugs are the GLP-1 products approved for weight reduction, but coverage is specified by formulation and not simply by brand. One of them is covered in a single delivery form only, and its other formulations are excluded. Writing the wrong formulation of a covered drug produces a denial that has nothing to do with your patient's eligibility.
Who qualifies for the Medicare GLP-1 Bridge?
Eligibility runs on three BMI thresholds. The highest qualifies on BMI alone. The two lower thresholds each require a qualifying comorbidity, and the two comorbidity lists are different from one another. The patient also has to be at least 18, enrolled in a participating plan type, and taking the medication alongside documented lifestyle modification.
Two features of that structure catch people out, and both are worth knowing before you screen anyone.
The comorbidity lists are narrower than the conditions you associate with obesity, and several entries carry specific written definitions rather than the clinical shorthand you would use in conversation. Uncontrolled hypertension is an example. It is defined by a numeric threshold and a treatment history, not by your impression that a patient's pressure runs high. A patient who is controlled on two agents sounds like she qualifies and does not meet the criterion as written.
Participating plan type is also narrower than Medicare Part D generally. Several plan types do not participate, which means a patient can meet every clinical criterion and still sit outside this program because of what she enrolled in.
The reading error that disqualifies eligible patients
The three criteria are minimums joined by “or”, and they are not exclusive bands. A patient whose BMI sits in the middle range still qualifies on a comorbidity from the lowest threshold's list, because her BMI clears that lower threshold as well. Reading the thresholds as ranges instead makes an eligible patient look ineligible.
Read as bands instead, the same patient looks ineligible. Her BMI lands in the middle range, the comorbidity she has is not on that range's list, and she gets turned away.
The banded misreading is easy to arrive at, and not because anyone is careless. The prior authorization form separately asks for a BMI range at initiation, with brackets that look exactly like tiers. That range field supports the obesity class code. It is not the eligibility test.
A patient turned away on this basis usually does not ask a second time. She hears that she does not qualify, and that is where it ends. If you have been screening Medicare patients out of this program, the criteria are worth rereading directly from CMS rather than from any summary, including this one.
Which BMI does the prior authorization actually use?
The prior authorization uses the BMI at the time the patient started GLP-1 therapy, not the BMI at today's visit. On a continuation request these are usually different numbers, because the medication worked. Reporting the current, lower BMI can disqualify a patient who genuinely qualified on the pathway she started on.
This is the failure point I would flag hardest, and it is only relevant in continuation visits, which means it has an impact later, after the workflow already feels settled.
Picture an invented patient. She is 63, she started Zepbound in March at a BMI of 31 with a prior myocardial infarction, and she qualified on the BMI 30 pathway with a qualifying comorbidity. Nine months later her BMI is 27.4. That is the outcome everyone wanted. If the renewal reports 27.4, she no longer clears the pathway she was approved on.
The BMI number the form wants is 31 (the BMI at initiation).
One consequence of that rule is easy to miss and worth acting on. The criteria apply as of the date GLP-1 therapy started, and that includes patients who started before they enrolled in Part D and patients who started before the Bridge launched on July 1, 2026. A patient who has been paying cash, or who was covered commercially before she aged into Medicare, can qualify on the BMI she had when she started. If you assumed this program was only for new starts, there may be patients already on your panel who are eligible.
The practical implication is a documentation habit rather than a clinical one. That number has to be findable in the chart a year later, when someone else is covering your panel or you are moving quickly through a renewal.
Which patients do not qualify for the Bridge?
Patients prescribed a GLP-1 for an indication that Part D already covers do not qualify for the Bridge. Those prescriptions route through the patient's own Part D plan instead. There are four such indications, and each one pairs with particular brands rather than with GLP-1s generally.
This one catches people because it runs opposite to instinct. A patient with more comorbidities feels like a stronger case. Under this program, certain diagnoses move her off the pathway entirely, because Part D already covers the drug for those indications.
Judgment applies here rather than a simple lookup. CMS directs prescribers to use clinical judgment about whether one of those diagnoses applies, and specifically to distinguish between a diagnosis that may no longer apply and a condition that is well controlled by treatment. A diagnosis sitting in an old problem list is not automatically a current one.
There is a second route off the Bridge that has nothing to do with diagnosis at all, and it turns on the patient's prescription history rather than her problem list. It is worth knowing about, because it is not something you would think to ask about unless you knew the rule existed.
The Bridge exists for the patient whose only indication is weight management. And the patients who fall outside it are not out of options. They are on a different pathway, with different drugs attached to different approved indications, and sending them down the Bridge route wastes weeks.
Free download: The Part D Pathway Quick-Reference
Four pathways, the specific brand that goes with each one, and four routing questions to run before you decide which route your patient is on. Because Ozempic is not Wegovy and Mounjaro is not Zepbound, even when the molecule is the same.
Does the order of submission matter?
It does, and this is the third failure point. The prior authorization cannot be submitted whenever you have a free minute, because it depends on something else happening first. A request that arrives before that step does not get denied on the merits. It does not process at all, and the office is left wondering where it went.
That one trips people up because nothing is clinically wrong with the request. It was complete, it was accurate, and it still went nowhere, which sends the office hunting for a documentation problem that does not exist.
The sequence is published, and it is worth reading the current version at CMS rather than working from what a colleague described in July. Operational details on this program moved during its first weeks, and one detail that was a placeholder at launch has since been filled in.
One thing to know before someone in your office spends an afternoon looking for it: there is no appeals process under this program. A denied request is corrected and resubmitted.
Why the note and the prior authorization should come from one encounter
The prior authorization asks the prescriber to attest to a specific set of facts, under penalty of perjury. Those facts are clinical, and you established them during the visit. The question is whether your note captured them in a form you can use later.
When the note and the payer submission are separate tasks, here is what actually happens. The request is denied for incomplete information. You open the chart, reread an encounter from three weeks ago, addend the note so the missing element is documented, and resubmit. Then your patient waits again. Nothing about that second pass is clinical work. It is reconstruction, and it happens after hours.
The alternative is to build the note so the attestation facts are already in it, in the same language the form uses, captured while the patient is in front of you. Then the submission is assembly rather than authorship, and someone else on your team can do it.
That is a design decision about your documentation, and it is the difference between a program that expands what you can offer your Medicare patients and one that quietly becomes a reason to avoid the conversation.
Delegating the submission
Most prescribers are not the person who files prior authorizations, and for this program delegation is close to necessary rather than optional. The administrative volume is real, and it is the reason an NP decides the whole program is not worth attempting for patients who would clearly benefit from it.
The delegation only works if the person filing knows what makes this program different from the prior authorizations they file every day, and knows exactly where to stop and hand it back to you. Those two things have to be written down and taught. Handing a colleague a new workflow when you are already behind costs you time in the first week. It stops costing you time in the second, and after that it is the only reason the program scales past a handful of patients.
The part no guideline covers
Everything above is knowable in advance. The criteria are published, the prior authorization form is public, and the submission sequence is documented by CMS. You can get every one of those things right and still watch a request sit in limbo for three weeks with no explanation attached to it.
A demonstration program in its first months has operational behavior that no clinical guideline addresses and no CMS page describes, because it is not policy. It is how the thing actually runs. Prescribers are working that out in real time right now, mostly alone, and mostly by making a phone call and learning something that would have saved the last person a week.
That is a poor way to learn anything, and it is the reason this is worth discussing somewhere other than a comment section.
Beyond the Chart
A private community for nurse practitioners, NPI verification required to join. It is where NPs compare notes on the things that do not appear in guidelines: what actually happens when a request stalls, how coverage pathways behave in practice, and how other people are handling the administrative load without absorbing it personally.
If you are prescribing under the Bridge, or you are about to be, that conversation is worth being in.
What to do before your next Bridge candidate
Screen before you prescribe, rather than after the pharmacy calls. The three failure points above, the wrong BMI, the wrong pathway, and the wrong order, are all checkable in the room in under a minute. Catching one there costs you a minute. Catching it after a denial costs you a resubmission, and it costs your patient another three weeks.
The program runs through December 2027, which is long enough to matter and short enough that the workflow you build now is the workflow you will use for its whole life.
One caution on everything above: this article reflects the program as of August 2026. CMS can revise criteria, covered drugs, and submission processes. Verify current details at CMS before relying on them.
Featured: The Medicare GLP-1 Bridge Chart Smart Kit
The documentation, the screening, and the submission workflow, already built.
Everything above tells you what the program requires. This kit is the part that turns those requirements into a visit you can actually document in the time you have.
Built by a working primary care nurse practitioner against the CMS program materials and current FDA labeling, it includes:
SOAP building blocks for starting a patient on the Bridge and for continuing her, with the attestation language already in the note
An eligibility screen to run before you prescribe
A prior authorization cover letter mapped to the CMS form, field by field
Submission instructions written for the staff who file your prior authorizations, including where they stop and hand it back to you
A plain-language patient handout covering what is common, when to call, and when to seek emergency care
Current as of August 2026. If CMS changes the criteria or the covered drug list, the kit is updated and buyers get the new version.
