Refill requests are the workflow that quietly eats concierge practices alive. Every clinic I've worked with underestimates the volume until they actually count it. A panel of 400 patients on chronic therapies can generate 60 to 120 refill touches a month, and if your NP or MD is personally reading each message, opening the chart, checking labs, and typing a response, you've just spent 15 to 20 clinical hours on work that should take three.

The practices that scale past 500 patients without adding a second clinician full-time have all figured out the same thing: refills need a protocol, and the protocol needs to run mostly without the clinician in the loop. Below is what actually works, based on what we see in the compounding and concierge clinics running on Qintara.

Why refills consume disproportionate clinical time

A refill request looks like a two-minute task. In practice it's ten steps: read the message, open the chart, check the last visit date, verify labs are current, confirm no drug interactions with anything added since the last fill, check the pharmacy on file, decide on quantity and refills authorized, send the script, message the patient back, document the encounter. If any of those steps surface a gap (labs overdue, form needs updating, pharmacy changed), the two-minute task becomes twenty.

The bottleneck isn't the clinical decision. For 80% of stable, chronic patients on established protocols, the decision is trivial. The bottleneck is the retrieval, verification, and documentation wrapped around the decision. That's the part async protocols are built to compress.

The three-tier refill protocol

Every high-volume practice I've seen handle this well uses some version of a tiered system. The tiers aren't about skill, they're about how much clinician judgment the request actually requires.

Tier 1: Auto-eligible refills

These are patients on stable, established therapies where the criteria for a refill are pre-defined and documented in the treatment plan. Typical examples in a concierge or compounding practice: continuation of a peptide protocol at an established dose, ongoing HRT at a stable regimen, GLP-1 continuation for a patient who's been on therapy 90+ days without dose changes.

The criteria for Tier 1 eligibility are written down and specific:

  • Last clinical visit within the required interval (often 90 or 180 days depending on the protocol)
  • Required labs on file and within range
  • No new medications, conditions, or symptoms reported on the last check-in
  • Patient current on membership and consent forms
  • Dose and formulation unchanged from previous fill

If a request meets every criterion, staff or an AI agent can prepare the script for clinician signature in a batch. The clinician reviews a queue once or twice a day, signs, and moves on. Time per refill drops from 10 minutes to under 60 seconds because the clinician sees a summary card, not a chart to dig through.

Tier 2: Requires clinical review but not a visit

These are refills where something has changed or is close to changing: labs are approaching expiration, the patient mentioned a new symptom on their last check-in, they're due for a dose reassessment. The clinician needs to actually think, but they don't need a full visit. A well-run Tier 2 queue includes a pre-built summary: what changed, what's flagged, what the recommended action is. The clinician either approves, adjusts, or bumps to Tier 3.

Tier 3: Requires a touchpoint

Labs are overdue, the patient reports something material, or they're at a natural reassessment point. These get routed to scheduling for a telehealth or in-person visit before any refill is authorized. This is the tier where clinics leak revenue and patient trust if they don't have a clean handoff, because the patient just wants their medication and now they're being asked to book a visit. The messaging matters. So does the speed of getting them on the calendar.

What actually needs to happen before you can run this

The protocol above assumes a few things that most practices don't have in place. Getting these right is the real work.

Treatment plans that specify refill criteria at the point of prescribing

When a clinician starts a patient on a therapy, the note should capture the refill criteria as structured data, not free text buried in the assessment. What labs, at what interval, what dose range is acceptable without re-evaluation, when does the patient need to come back in. If this lives only in the clinician's head, you cannot delegate anything.

A single source of truth for patient status

Staff and any automation need to see, at a glance: last visit, labs on file with dates, current medications, consent status, membership status, pharmacy on file. If any of this requires clicking through five screens or checking a separate system, the protocol collapses under its own weight. Most legacy EMRs weren't built for this, which is a large part of why practices end up building spreadsheets and Airtables around them.

Clear escalation rules

Staff need explicit permission to move requests between tiers and explicit rules for when to escalate. Ambiguity here is where burnout comes back in, because staff start pinging the clinician on Slack for every edge case and you've lost the whole benefit.

Where AI agents fit

The parts of this workflow that lend themselves to AI are the retrieval and drafting steps: reading the incoming request, pulling the relevant chart data, checking it against the Tier 1 criteria, and preparing either a signature-ready packet for the clinician or a message to the patient explaining what's needed next. The clinician still owns the clinical decision and the signature. What changes is that they're not spending eight minutes gathering context for every one.

In the clinics we work with, this typically cuts refill-related clinician time by 60 to 75% once the protocol is dialed in. That's not a marketing number, that's what we measure when we watch the queue before and after. If you want to see how the agent and EMR layer handle this end to end, talk to our team.

Common failure modes

A few patterns show up over and over when practices try to implement this and struggle:

  • Protocols that live in a Google Doc no one reads. If the criteria aren't enforced by the system, they aren't enforced. Staff default to "just ask the doctor."
  • Clinicians who won't let go of Tier 1. Some clinicians want to personally review every refill. That's fine at 150 patients. At 500 it's the reason they're working Sundays.
  • No SLA on the clinician queue. If the batched refill queue only gets reviewed when the clinician remembers, patients wait, staff field angry messages, and the whole thing feels worse than the old way.
  • Skipping the documentation step. Every refill is a clinical encounter and needs a note. If your protocol doesn't auto-generate the documentation, you're building a compliance problem.

On that last point, refill protocols intersect with state prescribing rules, telehealth requirements, and controlled substance regulations that vary by state and by drug class. Nothing above is legal advice. Run your specific protocol past your own counsel and your medical director before you turn it on.

Frequently Asked Questions

How many refills can one clinician realistically handle per week with a tiered protocol?

In practice, a single clinician with well-defined Tier 1 criteria and a batched review queue can sign off on 150 to 250 refills per week in roughly two to three hours of dedicated time, assuming the retrieval and drafting steps are handled by staff or an agent. Without a protocol, that same volume can easily consume 15+ hours.

Do patients notice the difference between a Tier 1 auto-prepared refill and one the clinician handled personally?

They notice speed. A refill that comes back in 90 minutes on a Tuesday afternoon feels better than one that took 36 hours because the clinician was in visits. The clinical content of the response is the same; the turnaround is what shapes the experience.

What's the right SLA for refill response times?

Most concierge practices we work with commit to 24 business hours for routine refills and communicate that explicitly at onboarding. The practices running tight async protocols usually beat that by a wide margin, but setting expectations at 24 hours gives you breathing room on the days someone's out.

Should front-desk staff or clinical staff run the refill queue?

Clinical staff (MA, LPN, or RN depending on your state's scope rules) should own Tier 1 and Tier 2 triage. Front-desk staff can handle the patient communication and scheduling piece for Tier 3, but the clinical eligibility check needs someone with the training to recognize when something in the chart should stop the refill from moving forward.