The Last Fill Doesn’t Announce Itself

Why patients fall off therapy quietly—and what it takes to see it happening in time to act

Pharmaceutical vice presidents, executive directors, brand leads, marketing leads, and commercial leads:

The discontinuation that nobody decided

Somewhere in your patient population today, a patient is picking up their last fill. Not their last scheduled fill—their last fill ever. Nothing about the transaction marks it. The claim adjudicates cleanly. The pharmacist hands over the bag. Everyone involved treats it as one more routine refill, because that is exactly what it looks like.

The discontinuation happens later, and it happens as a non-event. The supply runs out, and the next fill simply never occurs. No claim rejects. No one calls. There is no moment when the patient announces a decision, because in many cases there was never really a decision—just a gap that opened and was never closed.

Depending on your data, you may see the drop-off in a quarterly persistence curve or, with specialty pharmacy feeds, within days. But seeing it and being positioned to do something about it are different things—and the people who are positioned to do something about it, the prescriber and their team, are usually the last to know. That gap between who sees and who can act is where this article is headed.

The scale of the quiet exit

The numbers on long-term therapy are familiar enough that they have stopped registering, which is its own kind of problem. The World Health Organization’s estimate has stood for years: Adherence to long-term therapy for chronic disease averages about 50% in developed countries.[1] The Centers for Disease Control and Prevention puts the front end of the problem at roughly 1 in 5 new prescriptions never filled at all, with approximately half of those that are filled taken incorrectly—wrong timing, dosage, frequency, or duration.[2] And half of patients discontinue their medication within the first year.[3]

This is not just an economic problem, though the economics are stark: Estimates of avoidable spending attributable to nonadherence run from $100 billion to $300 billion annually,[2] and the broadest analysis of non-optimized medication therapy puts the figure at $528.4 billion a year, roughly 16% of US healthcare expenditures.[4] It is a clinical problem first: Nonadherence is associated with 125,000 deaths annually and at least 10% of hospitalizations in the United States.[3] Every one of those outcomes traces back to a therapy that a clinician judged worth prescribing and a patient stopped receiving.

For a brand, there is a quieter arithmetic underneath the public health numbers. Everything you invested to get the patient started—winning the prescribing decision, clearing prior authorization, surviving the first fill—is spent whether the patient stays on therapy for 3 months or 3 years. The patient who quietly stops at month 4 takes the entire acquisition cost with them. Persistence is where that investment either compounds or evaporates.

Two different ways patients leave

It is tempting to treat non-persistence as one problem, but it is at least two—and they respond to different things.

The first is a lapse. The supply runs out, and no refill is in place. Maybe the refills were never authorized past the initial fill. Maybe the renewal request sat in a queue. Maybe the patient assumed the pharmacy would call, and the pharmacy assumed the patient would come in. Nothing about this is a decision to stop therapy. It is logistics failing quietly, in the space between the prescriber’s office and the pharmacy counter, where it is nobody’s specific job to notice.

The second is a drift. The patient decides—usually without telling anyone—that the therapy isn’t worth continuing. The evidence here is counterintuitive: Forgetting accounts for less than 30% of nonadherence. Most of it is intentional.[5] Side effects that were never discussed. A co-pay that changed. Symptoms that improved, which the patient read as “done” rather than “working.” These patients don’t call the office to announce they’re stopping. As one family-medicine review put it: We didn’t ask, and they didn’t tell.[5]

The distinction matters because the interventions differ. A lapse needs timely logistics: Catch the gap; close it. A drift needs clinical engagement: someone the patient trusts, equipped to address the actual concern—side-effect management, financial support options, a reset of expectations about what the therapy is doing. A refill reminder does not fix a side-effect concern. A patient-education brochure does not fix an unauthorized refill. Programs that treat all non-persistence as one problem end up solving neither.

The evidence that the moment is actionable

Here is what makes persistence different from most commercial problems: The research keeps showing that the drop-off responds to intervention. Three separate strands of evidence each demonstrate one link in the chain.

Timeliness closes lapses. In one of the largest randomized studies ever conducted on the question, 735,218 Medicare Part D patients who were 3 days late refilling a maintenance medication were randomized to receive a reminder call or not. The single, well-timed contact increased the initial refill rate within 14 days by 22.8% relative to control, with measurable gains in annual adherence and persistence—from one call, placed at the right moment.[6] Notice what did the work. Not the channel—it was a phone call. Not the content—it was a reminder. The active ingredient was timing: The outreach arrived inside the gap, while the prescription was still a lapse and before it hardened into a discontinuation.

Clinicians act on signals that reach them where they work. In the PROMPT-HF cluster-randomized trial, clinicians managing heart failure patients received a targeted alert in their EHR at the point of follow-up, showing the patient’s current regimen against guideline-directed therapy. More than one-fourth of patients whose clinicians received the alerts had their therapy intensified versus just under 19% in the control group—one medication change for every 14 alerts—and 80% of clinicians found the alerts helpful.[7] The alert taught clinicians nothing they didn’t already know. It put the right patient-specific signal in front of them at a moment they could act on it, and behavior changed.

Care teams equipped for the side-effect conversation keep patients on therapy. Across 9414 oral oncolytic therapy courses in multiple health systems, a pharmacist-led protocol—a proactive check-in within 14 days of starting therapy, adverse-event management and mitigation, referrals to providers for supportive care—cut overall discontinuation from 40.4% to 29% and cut discontinuation driven by drug intolerance from 6.8% to 2.8%.[8] The drift patients, in other words, are reachable. When someone who knows what to say knows when to say it, patients who would have quietly stopped stay on therapy.

Nobody’s job to notice

What’s missing in most patient populations is that no one tells the care team the moment a patient starts to fall off therapy, while there is still time to act.

A rejected claim at least produces a record—a transaction that was attempted and failed. A missed refill produces nothing at all. It is the absence of an event, and absences don’t show up in anyone’s queue.

Walk through who could plausibly notice. The patient may not register the lapse or may have quietly decided and see nothing to report. The pharmacy’s view ends at its own counter. The prescriber’s view ends at the last office visit—refill activity happens downstream, in data the practice never sees. Between visits, a patient who has stopped therapy and a patient doing fine look identical from the clinic. You may see it in your data, quickly even—but you hold no lever. You cannot call the patient. The conversation that keeps a wavering patient on therapy belongs to the prescriber and their staff, and they are the one party in the chain with no idea it is needed.

So, the persistence problem, properly diagnosed, is not primarily a patient-motivation problem; and it is not a content problem—the industry has invested heavily in both. It is a routing problem. The signal exists. It just never reaches the people who could act on it while action is still possible.

What has to change is specific. Detect the gap as it opens—not the quarter after. Distinguish the lapse from the drift. And put the signal in front of the prescriber and the staff who handle refill authorizations, patient outreach, and the conversations that keep patients on therapy—with the educational and support resources for that conversation attached.

ReCaptRx^™ watches the place where patients disappear.

This is the gap ReCaptRx was built to close, extended to the longest-value stretch of the prescription journey.

ReCaptRx monitors real prescription activity and detects persistence events as they happen: the patient whose supply is about to run out with no refills remaining, the prescription paid for but never picked up, the refill that is late, the pattern that signals a patient drifting off therapy. When one of these events is detected, the prescriber and their staff are notified through the channels they already use—EHR, email, text, or fax—with the specific reason and a path to act, while the gap is still a gap.

The notifications are actionable and compliant: refill and renewal prompts for the lapses; clinically relevant materials and patient-support resources for the drifts—side-effect guidance, affordability programs, and adherence support the care team can put to work in the moment. And they reach the staff, not just the prescriber, because the staff is who acts.

A patient who stays on therapy through month 12 is worth multiples of one who leaves at month 3, and the difference between those two patients is often a single gap that someone either saw in time or didn’t. The last fill doesn’t announce itself. The work is building the system that notices anyway.

Dave Dierk is Co-President of Aventria Health Group and a 30-year thought leader in pharmaceutical sales and marketing. If your brand is losing patients through gaps no one sees in time to close, Dave welcomes the conversation: dave.dierk@aventriahealth.com.

References

1. World Health Organization. Failure to take prescribed medicine for chronic diseases is a massive, world-wide problem. July 1, 2023. Accessed July 15, 2026. https://www.who.int/news/item/01-07-2003-failure-to-take-prescribed-medicine-for-chronic-diseases-is-a-massive-world-wide-problem

2. Centers for Disease Control and Prevention. CDC grand rounds: improving medication adherence for chronic disease management—innovations and opportunities. MMWR Morb Mortal Wkly Rep. 2017;66(45). doi:10.15585/mmwr.mm6645a2

3. Duke Health. Medication nonadherence increases health costs, hospital readmissions. November 20, 2018. Accessed July 15, 2026. https://physicians.dukehealth.org/articles/medication-nonadherence-increases-health-costs-hospital-readmissions

4. Watanabe JH, McInnis T, Hirsch JD. Cost of prescription drug–related morbidity and mortality. Ann Pharmacother. 2018;52(9):829-837. doi:10.1177/1060028018765159

5. Brown MT, Sinsky CA. Medication adherence: we didn’t ask and they didn’t tell. Fam Pract Manag. 2013;20(2):25-30. Accessed July 15, 2026. https://www.aafp.org/pubs/fpm/issues/2013/0300/p25.html

6. Taitel MS, Mu Y, Gooptu A, Lou Y. Impact of late-to-refill reminder calls on medication adherence in the Medicare Part D population: evaluation of a randomized controlled study. Patient Prefer Adherence. 2017;11:373-379. doi:10.2147/PPA.S127997

7. American College of Cardiology. Electronic alert improves guideline adherence for heart failure treatment (PROMPT-HF trial). April 3, 2022. Accessed July 15, 2026. https://www.acc.org/About-ACC/Press-Releases/2022/04/03/12/29/Electronic-Alert-Improves-Guideline-Adherence-for-Heart-Failure-Treatment

8. Hutchinson K, King J, Mourani J, Wood A. Pharmacist-led 14-day oncolytic drug check-in protocol. Abstract # CR13. J Hematol Oncol Pharm. 2023;13. Accessed July 15, 2026. https://jhoponline.com/issue-archive/2023-issues/march-2023-vol-13-special-feature/19495:pharmacist-led-14-day-oncolytic-drug-check-in-protocol

The Last Fill Doesn’t Announce Itself