The patients who stop halfway, without telling you.

Why a full schedule hides it

A patient who drops out of a program doesn’t cancel anything. They just don’t reorder, or they skip session three and never book session four. There’s no empty chair to look at, and new starts keep filling the calendar, so the practice feels busy while the back half of every program thins out.

For GLP-1 programs the scale of it is published, and I go through it in The GLP-1 Retention Cliff. A 2025 retrospective cohort of 125,474 adults found that 64.8% of patients without type 2 diabetes stopped GLP-1 therapy within a year. The exits cluster in three windows: month one (side effects), months three and four (stalled progress), and months six through nine (“I hit my goal,” or cost fatigue). Multi-session series show the same shape, as I cover in The No-Show Tax: patients who skip session three rarely finish session six without someone reaching out.

What your answers usually mean

If you picked “the front desk tries to call,” “we notice eventually,” or “they just stop coming,” the lapse itself isn’t being caught. And if patients who finish or drop the program don’t move into anything else, the relationship ends the day the program does.

Neither is a people problem. It’s that no one owns the moment a patient goes quiet.

What most practices do, and a better routine

Most practices wait for the patient to raise it. By then the decision is usually made.

A better routine:

• Anyone who misses a refill or a session hears from you within 48 hours, with a way to rebook in the message.
• Check-ins are planned around the three windows, not left to whoever has time: early on, at the plateau, and before the patient decides they’re finished.
• The conversation about what comes next starts early. In the article I suggest raising maintenance by month four, before the patient says “I think I’m done.”
• Every program has a defined next step, whether that’s a maintenance plan your providers set, a related service, or a membership.

One number to check this week

Pull everyone who started your main program or series 6 to 12 months ago. Count how many are still active. For the ones who stopped, note roughly which month they stopped and whether anyone contacted them within a week. Counts are enough; you don’t need names on a spreadsheet to do this.

If most of the exits bunch up in one window, that’s where to start.

Napkin math

Patients who stopped in the last 12 months × your monthly program price × the extra months you honestly think a check-in would have kept them

That’s an estimate, not a finding. Use conservative numbers. For scale, the article works a 50-patient GLP-1 program at $400 a month losing patients at the published rate and lands at roughly $156,000 a year. Yours will be different, and it’s worth ten minutes to see how different.