The prescription that never makes it home

The gap between the prescription and the patient

A prescription is written. It is sent to a pharmacy. It is filled, bagged, and set on a shelf behind the counter with the patient's name stapled to it.

And then, often enough, nothing happens.

The medication sits there for a week. Then two. Eventually it gets returned to stock, and the patient goes back to their doctor months later with a condition that has quietly gotten worse. Nobody made a decision to stop treatment. There was no conversation about side effects, no second opinion, no change of mind. The prescription simply never made it home.

This is one of the least visible failures in American healthcare, and one of the largest.

How often prescriptions never get filled

The clearest look at the problem comes from electronic prescribing data, where researchers can match what a clinician sent to a pharmacy against what a patient actually picked up.

In a study of 75,589 patients published in the Journal of General Internal Medicine, Fischer and colleagues found that 28% of new electronic prescriptions were never filled at all. The numbers were worse for exactly the conditions where consistency matters most: 28.4% for hypertension, 28.2% for high cholesterol, and 31.4% for diabetes.

That is not patients quitting a medication partway through. That is treatment that never began.

Abandonment at the counter — a prescription that gets filled, then left unclaimed — is a smaller slice of the problem, but it moves sharply with cost. IQVIA Institute data on insulin patients found abandonment sitting near 3% when out-of-pocket costs held steady, and climbing to 11–12% when a patient's costs rose by more than $50.

What it costs

The consequences are measured in both dollars and outcomes, and the dollar figures are easier to pin down than the mortality ones.

In 2021, 8.2% of US adults aged 18–64 who take prescription medication — roughly 9.2 million people — reported not taking it as prescribed in order to save money, according to the CDC's National Center for Health Statistics. They skipped doses, took less than directed, or delayed filling a prescription. Among uninsured adults, that figure was 22.9%. Among adults in fair or poor health — the people with the least margin for a missed dose — it was 18.0%.

The broader economic estimate most often cited comes from Watanabe, McInnis and Hirsch in Annals of Pharmacotherapy, which put the cost of prescription drug–related morbidity and mortality at $528 billion a year. It is worth being precise about what that number covers: it is the cost of non-optimized medication therapy overall — wrong drugs, wrong doses, interactions, and non-adherence together — not non-adherence alone. Non-adherence is one contributor to a very large bill, not the whole of it.

Mortality is harder to quantify honestly. Widely circulated death-toll figures for non-adherence exist, but the most-repeated one rests on a derivation researchers have questioned, and we would rather not build an argument on a number we cannot stand behind. What is not in dispute is the direction: untreated hypertension, untreated diabetes, and untreated high cholesterol are among the best-understood drivers of preventable cardiovascular events in medicine.

Why prescriptions get abandoned

The reasons are rarely medical. They are logistical, financial, and geographic — which is both the frustrating part and the hopeful part.

Cost

Cost is the single best-documented driver, and the IQVIA and CDC figures above make the pattern clear: as out-of-pocket spending rises, pickup falls. This is a benefit-design and pricing problem. No delivery service fixes it.

Distance

Getting to a pharmacy is not a given. Research by Wittenauer and colleagues in Health Affairs Scholar found that 15.8 million people — 4.7% of the US population — live in a pharmacy desert, defined by both low income and low access to a nearby pharmacy.

That footprint has been shrinking further. A 2024 Health Affairs study by Guadamuz and colleagues found that nearly one in three US retail pharmacies closed between 2010 and 2021, and the closures were not evenly distributed: 37.5% of pharmacies in Black neighborhoods and 35.6% in Latino neighborhoods closed, against 27.7% in White neighborhoods.

Transportation

Even where a pharmacy exists, getting to it can be the obstacle. Wolfe, McDonald and Holmes, writing in the American Journal of Public Health, estimated that about 5.8 million Americans delay or forgo medical care each year because of transportation barriers — a share that held stubbornly flat at roughly 1.8% from 1997 through 2018, through two decades of otherwise significant change in how care is delivered.

Time, language, and everything else

Then there is the ordinary friction that does not show up in any dataset. A pharmacy counter open only during the hours you are at work. A three-day flu that makes the trip impossible on precisely the days the antibiotic matters. A parent managing prescriptions for a child and an aging parent in the same week. A patient who reads a pharmacy label, an insurance letter, and an automated phone tree in a language that is not the one they think in.

Which of these barriers delivery actually solves

It is worth being direct about the limits here, because overstating the case is how good ideas lose credibility.

Delivery does not lower a copay. It does not change a formulary, resolve a prior authorization, or reopen a pharmacy that has closed. If a patient leaves a prescription at the counter because the price at the register was three times what they expected, the last mile was never the problem.

What delivery does address is a narrower but very real category: the prescription that is filled, paid for, and sitting on a shelf that the patient cannot reach today. The distance problem. The transportation problem. The hours problem. The I-am-too-sick-to-drive problem. On the evidence above, that category is large — millions of people, concentrated in exactly the communities where pharmacies have been closing fastest.

It is also the category where the fix is unglamorous and entirely logistical. Somebody has to go get it.

Where Pills2Me fits

Pills2Me is not a pharmacy, and does not fill, dispense, or request prescriptions. The model starts one step later than people usually assume.

The prescription is already filled and waiting at the patient's own pharmacy — their CVS, their Walgreens, their local independent. The patient tells the Pills2Me AI Assistant, by text or by voice, in the language they speak at home. A trained Pills2Me caregiver collects it and brings it to the door, the same day. (The step-by-step version is here.)

Three details matter more than the speed:

• No pharmacy switch. Patients keep the pharmacist who knows their history. Continuity of care survives the convenience.
• Trained handlers, not general couriers. Pills2Me caregivers complete HIPAA and medication-handling training. A prescription is not a burrito.
• Ordering that meets people where they are. A family caregiver can arrange a delivery for a parent while making dinner, in their own language, without a portal or a phone tree.

None of this is a treatment. It is the removal of one specific obstacle standing between a patient and a medication a clinician already decided they needed.

The part worth fixing

Medication non-adherence is not one problem. It is a stack of them — pricing, benefit design, pharmacy economics, health literacy, transportation, and plain daily logistics — and they need different tools.

But it is worth noticing how many of the barriers in the research above have nothing to do with medicine. A pharmacy that closed. A car that will not start. A shift that ends after the counter does. Those are solvable problems, and they are the ones we work on — with patients, pharmacies, health plans and health systems.

The prescription is already filled. Getting it the last two miles should not be the hard part.

---

Pills2Me is not a pharmacy. Prescriptions are dispensed by licensed partner pharmacies. Pills2Me is HIPAA-compliant. This article is for general information and is not medical advice — talk to your pharmacist or clinician about your own medications.

Sources: Fischer MA et al., Journal of General Internal Medicine (2010) · CDC/NCHS Data Brief No. 470 (2023) · Watanabe JH, McInnis T, Hirsch JD, Annals of Pharmacotherapy (2018) · Wittenauer R et al., Health Affairs Scholar (2024) · Guadamuz J et al., Health Affairs (2024) · Wolfe MK, McDonald NC, Holmes GM, American Journal of Public Health (2020) · IQVIA Institute, Emergence and Impact of Pharmacy Deductibles