A refill request looks like the most routine thing in a practice's inbox. The published evidence says otherwise: in a full year of measured data, more than a third carried a clinical finding.
The short answer
In a study at a federally qualified health center published in the Journal of Primary Care & Community Health, clinical pharmacists processed 1,683 refill requests across 1,255 encounters over one year. Of those encounters, 453 — 36.1% — produced at least one clinical intervention.
The findings weren't exotic:
| What the "refill request" actually was | Share of encounters |
|---|---|
| Patient due for a follow-up visit | 32.9% |
| Patient due for labs | 31.9% |
| Drug therapy problem | 12.6% |
| Medication list discrepancy — pharmacy request didn't match the chart | 11.9% |
| Should be referred to another clinical service | 7.2% |
| Coverage issue or clarification needed | 3.4% |
More than one finding could apply to a single request, which is why the shares exceed 36% in total.
Not an outlier
The obvious objection is that pharmacists reviewing refills will find things to justify the program. The study addresses it by citing two independent efforts:
- Nguyen and Zare found interventions needed in 42% of pharmacist-managed refills across more than 5,000 requests over two years
- Billups et al. found laboratory or other monitoring needed in 28% of refill authorization requests
Three studies, three settings, roughly the same answer: somewhere between a quarter and two-fifths of refill requests are carrying something.
What "drug therapy problem" means concretely
The study identified 83 drug therapy problems and categorized them. This is the part worth reading closely, because these aren't edge cases:
| Problem | Share |
|---|---|
| Dosage too low | 24.1% |
| Requires a different drug product | 21.7% |
| Dosage too high | 20.5% |
| Unnecessary drug therapy | 20.5% |
| Requires additional drug therapy | 7.2% |
| Non-adherence | 3.6% |
| Adverse drug reaction | 1.2% |
Their own examples: aspirin still being taken for cardiovascular prophylaxis by low-risk patients, therapeutic duplications, diabetes medications that current guidelines suggest replacing with preferred alternatives, and doses needing titration or adjustment for kidney function.
Roughly a fifth of the findings were medications that should have been stopped. The correct response to "please refill this" was "actually, you shouldn't still be on this."
The finding that should change how you build
Here's the number nobody quotes.
The researchers went back at three months to see what had happened to the 83 drug therapy problems the pharmacists identified and documented:
- 37.3% addressed or resolved
- 55.4% never addressed
- 7.2% indeterminate or lost to follow-up
A clinician did the work. Reviewed the chart, identified the problem, wrote the recommendation, put it in front of the primary care provider. More than half of it produced no action at all.
Note that the bar was low — an intervention counted as resolved if the provider merely acted on or responded to it, and it didn't have to be accepted as recommended. Even by that generous standard, most of it evaporated.
That's not a detection failure. It's a closed-loop failure, and it's a different problem with a different fix. Finding more problems in a system that drops half of them yields nothing.
What this means for automation
The pitch for refill automation is that most requests are routine and can be approved without a human. The evidence complicates that in a specific way.
The refill queue is functioning as an unintentional population health screen. It is one of the few moments a patient's chronic medication regimen gets examined outside a scheduled visit. Automate the routine-looking ones away and you don't just lose the catch — you lose the occasion for looking.
The 36% doesn't announce itself. Nothing in the pharmacy's fax says "this patient is overdue for an A1C." It surfaces only because someone opened the chart while handling a request that looked exactly like the other sixty-four percent.
So the goal isn't approving faster. It's surfacing what needs a human, and tracking it until someone closes it. The first half is detection. The second half is where the published evidence says the value actually leaks out.
What the study can't tell us
Two honest limitations, because they bound what anyone should conclude.
We don't know how much of the remaining 64% was genuinely mechanical. A pharmacist reviewed every request in this study. Nobody measured what a lighter-touch process would have missed, so "64% needed nothing" and "64% got nothing found because nothing was looked for" are not distinguishable from this data.
The reviewer was a clinical pharmacist, not software. These findings required someone qualified to judge that a dose was too low or a drug unnecessary. We are not claiming an agent substitutes for that judgment, and any vendor implying otherwise is describing something that doesn't exist. What software can plausibly do is assemble the context — the med list comparison, the last visit date, the overdue labs — so the judgment takes seconds instead of three screens.
The boundary the protocol drew
The study's inclusion and exclusion criteria are worth adopting as-is, because they came from a real IRB-approved protocol at a working clinic rather than a vendor's risk assessment.
Eligible for protocol handling: established patients 18 or older, seen in-office or by telemedicine within the preceding 12 months, requesting medications for chronic conditions.
Excluded entirely, forwarded to the provider untouched:
- Controlled substances
- Medications for any acute condition
- Pain medications — NSAIDs, acetaminophen, muscle relaxants
- Proton pump inhibitors
- Antibacterial and antifungal medications
That last group matters more than it looks. A refill request for an antibiotic usually signals that something didn't resolve — it's a clinical question wearing a refill's clothing.
On the numbers you'll see elsewhere
We went looking for the commonly cited refill statistics and could not verify most of them.
The AMA has published a claim that physicians can save two hours a day through annual prescription refills. Reading the source, it's attributed to an individual's assertion, supported by personal experience rather than measurement, and it describes synchronized annual prescribing — a workflow change, not software. It's a sound idea. It isn't evidence for automation.
Various figures attributed to MGMA — requests per provider per week, share fully automatable — appear on vendor blogs without a locatable report behind them. What MGMA does publish freely is a March 2026 poll of 294 practice leaders in which prescription refills ranked last among time-consuming phone tasks at 6%, behind eligibility and prior authorization at 45%.
For actual measurement, the study here is the better source: pharmacists handled those 1,683 requests in an average of 3.2 hours per week, about 10 requests per hour, with 87.8% turned around within 48 business hours.
Frequently Asked Questions
What percentage of prescription refill requests require clinical intervention?
In a year-long study at a federally qualified health center published in the Journal of Primary Care and Community Health, 36.1% of pharmacist-managed refill encounters resulted in at least one clinical intervention. The study cites corroborating research by Nguyen and Zare that found interventions needed in 42% of pharmacist-managed refills across more than 5,000 requests over two years, so the finding appears consistent across settings.
What kinds of problems are found during prescription refill review?
The most common findings are administrative-clinical rather than dramatic. In the published study, 32.9% of encounters involved a patient due for a follow-up visit and 31.9% a patient due for laboratory monitoring. Beyond those, 12.6% surfaced a drug therapy problem, 11.9% found a discrepancy between the pharmacy request and the chart medication list, and 7.2% identified a patient who should be referred to another clinical service.
What is a drug therapy problem in a refill request?
It is a medication issue that makes the request as written clinically inappropriate. In the published study the most common categories were dosage too low at 24.1%, requires a different drug product at 21.7%, dosage too high at 20.5%, and unnecessary drug therapy at 20.5%. Concrete examples given include aspirin continued for cardiovascular prophylaxis in low-risk patients, therapeutic duplications, and diabetes medications that current guidelines suggest replacing.
Do clinicians act on medication problems found during refill review?
Often not. The study followed up at three months on the 83 drug therapy problems identified and found that 37.3% had been addressed or resolved, 55.4% had not been addressed at all, and the remainder were indeterminate. The clinical work of finding the problem was completed and documented, and more than half of it produced no action, which makes tracking to closure at least as important as detection.
Should prescription refills be automatically approved?
Blanket auto-approval optimizes for the wrong outcome. The refill queue functions as one of the few moments a chronic medication regimen gets reviewed outside a scheduled visit, and the published evidence indicates a meaningful share of requests carry a finding. A system that approves the routine-looking ones faster without examining them removes the occasion for looking, so the appropriate goal is surfacing the requests that need a human rather than minimizing the number that reach one.
Which medications should be excluded from refill protocols?
The published protocol excluded controlled substances, medications for acute conditions, pain medications including NSAIDs and muscle relaxants, proton pump inhibitors, and antibacterial and antifungal medications. It also required that patients be established adults with an in-office or telemedicine encounter within the preceding twelve months. Requests outside those criteria were forwarded to the primary care provider untouched rather than handled by protocol.
How much clinician time do prescription refills actually take?
Less than commonly claimed, and the widely circulated figures are poorly sourced. The AMA has published a two-hours-per-day savings claim, but it is attributed to an individual's assertion and personal experience rather than a study, and it describes synchronized annual prescribing rather than automation. In the peer-reviewed study, pharmacists handled 1,683 refill requests in roughly 3.2 hours per week, averaging about 10 requests per hour.
What to do next
Pull your last fifty refill requests and check two things: how many involved a patient overdue for monitoring, and how many findings from the last six months were ever closed.
The second question is usually the uncomfortable one.
Source: Hurley-Kim K, Keyvani A, Ahmed R, Wong HW, McBane S. Pharmacist-Managed Refill Service Impacts on Clinician Workload and Medication Interventions in a Federally Qualified Health Center. Journal of Primary Care & Community Health, 2023. PubMed 37070677
Related: why a refill request isn't a task at all, and what happens when findings sit in an inbox.
