Explore this article's sources with AI
Follow PeptideMethods on Google
When Medicaid GLP-1 coverage ended
A 2026 study found a steep fall in Medicaid-reimbursed GLP-1 dispensing after five states withdrew obesity coverage.
Why we wrote this. Coverage headlines need a clear boundary between reimbursed dispensing, treatment interruption and patient outcomes.
In this article (5 sections)
When five state Medicaid programmes withdrew obesity coverage for GLP-1 medicines, reimbursed dispensing fell sharply within one quarter. A 2026 quasi-experimental study estimated an 84.7% decline in obesity-indicated dispensing relative to continuing states[1]. The result measures reimbursed prescriptions, not health outcomes, private purchases or what happened to every individual enrollee.
The policy question includes medicines such as semaglutide and tirzepatide. Our semaglutide overview and tirzepatide overview explain the clinical evidence separately. Coverage determines whether a programme pays. It does not by itself decide whether a medicine is appropriate for a particular person.
How the study was designed
Researchers used quarterly Medicaid State Drug Utilization Data for all 50 states and the District of Columbia from 2021 through the first quarter of 2026. They compared states that withdrew coverage with 12 states that continued it. Each state's dispensing of diabetes-indicated GLP-1 medicines served as an additional within-state comparison[1].
This difference-in-differences design looks for a change in the withdrawing states beyond the change seen in comparison states. It can strengthen an observational analysis when pre-policy trends are similar. It is not random assignment. Other state-level changes, coding choices or differences in implementation may still affect the estimate. The authors report that pre-announcement differences were small and not statistically significant, which supports the design without eliminating every source of confounding. The estimate is therefore more informative than a simple before-and-after count, but it remains an observational policy result.
The main outcome was reimbursed prescriptions for obesity-indicated GLP-1 receptor agonists per 1,000 enrollees per quarter. The underlying utilisation dataset is published by the Centers for Medicare and Medicaid Services[2]. In practical terms, these records capture paid dispensing at programme level, not whether a person used the medicine, obtained it elsewhere or experienced a clinical outcome. That boundary matters when a large prescription decline is discussed as though it were already a measured change in health.
What changed after withdrawal
The five withdrawing states represented 2,428,709 of 3,725,732 Medicaid-reimbursed obesity-indicated GLP-1 prescriptions in 2025. The study estimated 29.1 fewer prescriptions per 1,000 enrollees per quarter after withdrawal, alongside the 84.7% relative decline[1]. The confidence interval for the relative decline was 64.5% to 93.4%.
In the first quarter without coverage, individual withdrawing states saw obesity-indicated dispensing fall by 61% to 96%. Four states that withdrew coverage on 1 January 2026 reimbursed 661,353 such prescriptions in the preceding quarter and 134,647 in the next, a raw decline of 79.6%[1]. Those counts describe reimbursement, not a direct count of interrupted patients.
The comparison with diabetes indications helps narrow the interpretation. The researchers detected no significant change in diabetes-indicated GLP-1 dispensing, with an estimate of minus 3.2% and a confidence interval spanning substantial decreases and increases[1]. This pattern is consistent with the obesity coverage change being the main driver, though it cannot prove every causal pathway.
Other medicines did not fill the gap
Dispensing of older anti-obesity medicines increased by no more than 2.9 prescriptions per 1,000 enrollees, far below the estimated GLP-1 decline. Two states that narrowed criteria instead of fully withdrawing coverage saw declines of 23% to 37%[1]. The data therefore do not show a one-for-one shift to older medicines.
That does not tell us why substitution was limited. Prescriber decisions, contraindications, patient preferences, prior authorisation and supply can all matter. Nor does the study establish that a different medicine should have been used. A policy analysis can show a dispensing gap without prescribing the clinical response to it.
What we don't yet know
The authors say protecting existing users, such as by honouring current authorisations, might reduce interruption, but they also state that this approach has not been evaluated[1]. The study does not report subsequent weight change, cardiovascular events, adverse effects, out-of-pocket spending or treatment obtained outside Medicaid.
That is the limit. The analysis ends with the first quarter of 2026, so it cannot show whether dispensing later recovered or whether states changed policy again. The small number of withdrawing and continuing states also warrants care, particularly because state averages can conceal different effects by age, disability status, geography, enrolment category and the reason a medicine was prescribed.
What the result means for readers
The study provides strong evidence that coverage rules can change reimbursed access quickly. It does not support stopping treatment without a plan or changing indications to obtain payment. People affected by a coverage change should speak with their prescriber and Medicaid programme about current authorisation, appeal and continuity rules. Those rules are state-specific and can change.
Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.
Frequently asked
How much did dispensing fall after coverage withdrawal?
The study estimated an 84.7% relative decline in obesity-indicated GLP-1 dispensing compared with continuing states.
Did the study measure patient health outcomes?
No. It measured Medicaid-reimbursed prescriptions, not medicine use, weight change, cardiovascular events or other clinical outcomes.
Did older obesity medicines replace the lost dispensing?
Only slightly. Their increase was far smaller than the estimated decline in obesity-indicated GLP-1 dispensing.
Can the findings prove every treatment was interrupted?
No. Claims data cannot show whether a person obtained treatment through another payer, paid privately or stopped using the medicine.
Sources
- [1]Nyambod et al. Medicaid coverage withdrawal and GLP-1 dispensingTier 1 · primary↩
- [2]CMS State Drug Utilization DataTier 1 · primary↩
No revisions yet. First published .