There is a version of the GLP-1 story that gets told in aggregate numbers, and it is misleading. The aggregate says adoption is rising, coverage is improving, and access is broadening. KFF’s March 2026 polling found that about 18 percent of US adults have taken a GLP-1 medication at some point and 12 percent were taking one when surveyed. Those are large figures for a drug class this young.
The aggregate hides the structure underneath. Access is not broadening evenly. It is bifurcating, and the variable doing most of the sorting has nothing to do with medicine. It is the size of the company a person works for.
Two markets wearing one name
Employer-sponsored coverage for weight-management medication has expanded over the past several plan years, but the expansion has concentrated among the largest employers, the ones with the actuarial scale to absorb a high-cost pharmacy category and the benefits staff to administer it. Firms with a few hundred employees have moved far less. Many have not moved at all.
This produces two populations with the same diagnosis and radically different economics. One presents a card and pays a copay. The other discovers that weight-management medication sits outside the formulary entirely and faces the full cash price. Same condition, same clinical guidance, different financial universe.
KFF’s data suggests the split is messier still. A majority of GLP-1 users said the medication was difficult to afford, and that majority included 55 percent of users who had insurance. About 27 percent of users reported holding insurance and paying the whole cost themselves regardless. Coverage, in this category, is not a binary. It is a spectrum running from full coverage through partial coverage, prior authorisation hurdles, step requirements and outright exclusion, and a patient frequently cannot locate their own position on that spectrum until they try to fill a prescription.
Why weight management is treated differently
Plans that cover a GLP-1 for one indication often exclude it for weight management specifically. The historical reasoning traces back to a period when weight-loss medication was classified alongside cosmetic treatment rather than chronic disease management. That classification has eroded in clinical consensus faster than it has eroded in benefit design.
Benefit design moves on annual cycles and responds to cost projections. A category where a meaningful share of a workforce might become eligible, at several hundred dollars per member per month, indefinitely, is a category that makes benefits consultants cautious. The caution is rational from a plan sponsor’s position and produces an outcome that looks arbitrary from a patient’s.
What the uncovered population actually does
People who fall outside coverage do not generally stop wanting treatment. They look for another route, and the routes available to them have changed substantially.
Direct-pay telehealth grew into precisely this gap. Fact.MR valued the GLP-1 telehealth platforms market at 763.8 million dollars in 2026 and projects roughly 3.28 billion by 2036. That growth curve is not an accident of convenience. It maps onto the population that insurance left standing at the counter.
What cash-pay models offer this group is not a discount against an insured price. Most of these patients have no insured price. What they get is a number that exists, that they can see before committing, and that does not change between months. For a household trying to decide whether an expense fits, a knowable recurring figure is a different kind of object from an unknowable one.
The regulatory ground shifted underneath
Anyone comparing cash prices in 2026 is comparing products in a landscape that changed in 2025. The FDA removed tirzepatide from its shortage list in December 2024 and declared the semaglutide injection shortage resolved in February 2025. That ended the broad enforcement discretion under which large-scale compounding of both molecules had operated.
Compounded versions remain lawfully available from licensed pharmacies, but on a narrower footing. A 503A pharmacy now compounds these molecules against a prescriber-documented, patient-specific clinical rationale rather than a shortage listing. Compounded medications are prepared by a licensed pharmacy for an individual patient, are not approved by the FDA, and do not go through its premarket review for safety, effectiveness or quality. In April 2026 the FDA proposed permanently excluding semaglutide, tirzepatide and liraglutide from the 503B Bulks List, with a comment period extending into July 2026.
The practical consequence for a patient is that the question “why can this provider offer this price” now has a real answer, and providers differ in their ability to give it.
How it works in practice
A segment of the market has organised itself specifically around the uncovered patient. HealSend, a US telehealth service, publishes flat monthly pricing rather than a quote generated after intake. The company lists compounded semaglutide at 299 dollars per month and compounded tirzepatide at 399 dollars per month, each as a once-weekly injection, with FedEx shipping in discreet, temperature-controlled packaging included.
Patients complete an online health profile that a US-licensed physician reviews before any prescription is issued, and eligibility is determined by that physician. Treatment is open to adults aged 18 to 75 and is not for use during pregnancy. The company states that its plans are HSA and FSA eligible for reimbursement, which matters disproportionately for this group: a patient with no pharmacy benefit may still hold a tax-advantaged health account, and that account is often the only coverage-adjacent instrument they have. Patients searching for affordable tirzepatide can compare a published figure against their own budget before entering a consultation. HealSend states that if its provider determines a GLP-1 medication is not right for a patient, the patient receives a full refund. The service operates across most, though not all, US states.
The HSA and FSA point deserves emphasis because it is frequently overlooked. For the employee of a small firm whose plan excludes weight management, a health savings account may be the single mechanism through which pre-tax dollars reach this category at all.
Friction points and what comes next
State lines still fragment the market
Telehealth prescribing remains governed state by state. A provider’s availability map is a function of licensing, not of clinical capability, and patients in excluded states find that the cash-pay route is not uniformly available either. The uncovered population and the unserved population overlap but are not the same set.
Coverage gains may not reach the people furthest out
If employer coverage continues expanding from the largest firms downward, the last population to benefit will be employees of small businesses, contractors, gig workers and the self-employed. These are the groups with the least negotiating leverage over their own benefits and, frequently, the most variable incomes. Expansion that proceeds in order of employer size reaches them last by construction.
Price transparency is improving unevenly
Some direct-pay providers publish complete figures inclusive of shipping and clinical review. Others publish a headline that excludes both, or an introductory rate that steps up. A patient comparing across providers is frequently comparing numbers that do not describe the same thing. Until presentation standardises, the burden of translation sits with the patient, which is precisely where it is least likely to be handled well.
The structural question
The GLP-1 category is often discussed as though its central tension were clinical, a matter of weighing benefits against side effects. For a large share of the people affected by it, the central tension is administrative. Whether a person can access a medication their clinician considers appropriate turns substantially on the benefit design of an employer they may have joined for unrelated reasons.
That is not a sustainable basis for allocating a chronic-disease therapy, and the market has noticed. The rapid growth of direct-pay infrastructure is, in effect, a parallel distribution system built because the primary one sorts people by something other than need.
Whether the two systems converge or continue diverging is the open question. The answer will be written in benefit design and in regulation rather than in clinical trials, which means it will be slower and less visible than the science that produced these medications in the first place.
