Every week, care teams celebrate a patient starting GLP-1 therapy. The prescription is written, the prior authorization clears, the first fill is picked up. On paper, the program is working.
Then, quietly, a large share of those patients stop. Not in a single dramatic moment, but through a missed refill, a delayed appointment, a month where the copay felt impossible. Nobody calls. Nobody notices until the next visit, or the next claim, or the next admission.
The numbers behind the drop off
A peer reviewed analysis of more than 4,000 commercially insured adults without diabetes who started a GLP-1 for obesity found that only 32 percent were still on therapy a year later. Fewer than half made it to six months. That is not an edge case. It is the typical course.
And stopping has consequences. In the extension of the STEP 1 trial, participants who discontinued semaglutide regained roughly two thirds of the weight they had lost within a year. For patients managing diabetes, the rebound shows up in glycemic control, and eventually in the emergency department.
The result is the worst of both outcomes: drug spend that never produced its intended benefit, followed by the downstream cost of the condition the drug was meant to manage.
Why today's workflow misses it
Pharmacy data lives in one place. Clinical data lives in another. The behavioral signals that predict discontinuation, such as cost pressure, side effect friction, and missed appointments, are rarely captured together at all.
Claims data confirms a missed refill weeks after it happened. By then the patient has already been off therapy long enough that re engaging them is harder, and the window where a simple call or a copay assistance referral would have worked has closed.
What intelligence should change
The goal is not another dashboard of adherence rates. It is a short, ranked list each morning of the patients most likely to stop, with a plain reason attached to each one.
High dropout risk, primary driver out of pocket cost, recommended action copay assistance referral. A care coordinator can act on that in a minute. A percentage on a population report gives her nothing to do.
From insight to action
Knowing is not doing. The intervention has to be specific to the driver: financial navigation for cost pressure, a pharmacist conversation for side effects, scheduling help for appointment friction, a pharmacy change when the dispenser is unreliable. Generic check in calls rarely move anything.
Measure what happens next
Close the loop on every flagged patient. Was the outreach made? Was the refill picked up? Is the patient still on therapy at ninety days, at six months, at a year? Those are the numbers that show whether a GLP-1 program is producing outcomes or just prescriptions.
Preventra is building treatment continuity intelligence designed to help care teams and health plans see discontinuation risk early, prioritize outreach, and measure what happens next.
Sources
- Gleason PP, et al. Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes. Journal of Managed Care & Specialty Pharmacy, 2024. View study
- Wilding JPH, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes, Obesity and Metabolism, 2022. View study