Everyone keeps treating GLP-1 dropout like a mystery. It isn't. We asked people why they stopped, and they told us.
The reasons were never the hard part.
What patients actually say
A 2026 cross sectional survey published in the Journal of Managed Care and Specialty Pharmacy screened just over seven thousand adults and surveyed the 440 who had started GLP-1 therapy between 2022 and 2025 and later stopped. What they reported: cost, 36.1 percent. Adverse effects, 33.3 percent. Lack of insurance coverage, 28.2 percent. Reaching a weight loss goal, 26.1 percent.
Look at what's sitting at the top of that list.
Two of the top three have nothing to do with the medication. Cost and coverage aren't clinical findings. They're questions about whether a person could get the drug and keep getting it. In those cases the therapy didn't fail. The pricing and paperwork around it did.
A different method, the same shape
Surveys have a known weakness. They're recall, gathered after the fact, from people explaining a decision they've already made.
So it's worth setting a different method beside it. Truveta's real world analysis of GLP-1 discontinuation, presented at ISPOR in 2025, skipped the survey. It pulled the reasons documented in clinical notes at the time, for nearly 79,000 patients, most of them taking the drug for diabetes. Side effects came first at 28.2 percent. Cost came in at 12.8 percent.
Different method, different ordering, same conclusion underneath. A meaningful share of this was never a clinical problem.
And the stakes haven't gone away. Persistence has improved since the shortage years: Prime Therapeutics found that 60.9 percent of commercially insured adults without diabetes who started Wegovy or Zepbound in the first half of 2024 were still on therapy a year later, against 33.2 percent of 2021 starters. That's real progress, and it still means more than a third stopped within a year. We looked at the older, much worse numbers in October.
Every one of these has a date on it
This is the part I'd defend in any room.
Each of the top three reasons comes with a date you can see in advance.
A copay change exists in the benefit design months before anyone feels it at the pharmacy counter. An expiring prior authorization has an end date sitting in a field somewhere. A step up to the next dose needs a new prescription, often brings a fresh round of side effects, and sits on a schedule somebody wrote down weeks ahead.
None of that is hidden. It just never gets pulled together around one named patient while there's still time to act on it.
The authorization piece deserves saying plainly. In the AMA's 2025 survey of 1,000 practicing physicians, 95 percent said prior authorization delays necessary care, and 79 percent said patients abandon treatment because of it. That survey wasn't GLP-1 specific, so I wouldn't present it as a GLP-1 finding. The lesson carries over anyway: treatment can fail without anyone ever making a clinical decision to stop it.
The most predictable discontinuation event of the year
If you want somewhere to test all of this, it's six weeks away.
January 1. Plan years reset. Formularies change. Deductibles start over. Some prior authorizations have to be redone. For most of a panel, it all happens on the same day.
This January makes the point on its own. Fourteen percent of employers in a Business Group on Health survey have dropped or plan to drop GLP-1 weight loss coverage by 2027, and Novo Nordisk's list price cut for Wegovy and Ozempic takes effect January 1. Some patients will pay less in January. Some will lose coverage. Both were decided months ago.
It's the single most predictable discontinuation event on the calendar, and almost nobody works it in advance. The patients who will fall out in January can be identified in November. The work is knowing which ones, and reaching them before a failed fill at the counter is how they find out something changed.
Nobody learns about January on January 1. They learn about it in February, from a report.
Aggregate reporting answers the wrong question
Reporting tells you what happened. That's its job. It answers "why did patients discontinue last quarter" with real precision, and that answer matters for formulary design, contracting and strategy.
What it can't answer is "which of my patients is about to." Those are two different questions that need two different kinds of machinery, and most organizations have only built the first.
A discontinuation report is a postmortem. Often a very good one. Still a postmortem.
What a serious program would watch
Three measures. The first is the hardest to build, and the one that matters most.
Patients with a coverage or authorization change effective in the next sixty days. That's a list with names on it, not a rate.
Days from a detected gap to human contact. Not a mailing, not an app notification. A conversation, which is the same standard we argued for on the readmissions side.
Persistence through the plan year boundary, measured separately from annual persistence, because January doesn't behave like any other month, and blending it in hides exactly the thing you need to see.
That's the layer we're building at Preventra. It pulls refill activity, authorization end dates, coverage changes and appointment history together around one patient, and puts the patients with a change coming in front of the people who make the call, with as much lead time as the data allows. A January problem should show up on somebody's list in November.
The reasons patients quit were never the missing piece. The missing piece is knowing which patient, and knowing early enough to do something about it.
Sources
Journal of Managed Care and Specialty Pharmacy. Cross sectional survey of reasons for GLP-1 receptor agonist discontinuation, 2026. DOI 10.18553/jmcp.2026.32.8.903.
Truveta Research. Real-world temporal and indication-specific variation in drivers of GLP-1 RA discontinuation. Poster presented at ISPOR, 2025.
Marshall LZ, Gleason PP, Friedlander N, Farley J, Urick BY. Trends in 1-year persistence and adherence among initiators of high-potency, weight loss-indicated glucagon-like peptide 1 receptor agonists. Journal of Managed Care and Specialty Pharmacy, 2026;32(3):281-291.
American Medical Association. 2025 AMA Prior Authorization Physician Survey.
Human Resources Director. More employers are dropping GLP-1 weight loss coverage before open enrollment, 2026.
Novo Nordisk. Wegovy and Ozempic list price reduction, effective January 2027.
