Every compounding practice I've worked with tracks new patient acquisition obsessively. Very few track refill turnaround time with the same discipline, which is strange, because refills are where the revenue actually lives. A patient who reorders on time for eighteen months is worth ten times more than one who churns at month three because their second refill took nine days to authorize.

If you run a concierge or compounding-adjacent practice, refill speed is not a customer service metric. It's a revenue metric. Here's what that actually looks like in the numbers.

The math nobody wants to run

Take a practice with 800 active patients on compounded therapies averaging $340 per month in program revenue. Assume a healthy 6% monthly churn baseline. Now assume that every refill delayed beyond 72 hours increases churn probability for that patient by roughly 2 to 3 percentage points over the following 90 days. In our internal review across a handful of clinics, that number was closer to 4 points on GLP-1 and HRT programs where patients were price-shopping.

Do the arithmetic. If 20% of your refills lag beyond the 72-hour mark, and that lag adds even 2 points of churn to those patients, you're losing 32 to 40 patients a quarter you would otherwise keep. At $340 a month with an average remaining tenure of eight months, that's somewhere between $85,000 and $110,000 in annualized revenue walking out the door, per quarter, from a single operational failure.

That's the visible loss. The invisible one is worse: the patient who didn't churn but told two friends the refill process was a mess, so those friends never signed up.

Where refills actually stall

When we've audited refill workflows across compounding practices, the delays almost never come from the pharmacy side. They come from the clinic. Specifically, from four choke points:

  • Clinical review queues. A refill request lands, and it sits in a provider's inbox for two or three days because the provider is seeing patients and hasn't been given protected async time.
  • Missing labs or check-ins. The refill can't be authorized because a required quarterly check-in wasn't scheduled, and nobody flagged it 30 days before the refill was due.
  • Payment friction. Card on file expired, autopay failed, and the refill is technically approved but sitting behind an unresolved billing task.
  • Handoff gaps. The MA processes intake, the NP reviews, the front desk collects payment, and the pharmacy fulfills. Nobody owns the end-to-end clock.

In practice, the single most common pattern I see is the second one: a refill is due, the patient hasn't had their check-in, and now the clinic is scrambling to schedule a visit reactively instead of having queued it 30 days out. The patient waits. The patient gets frustrated. Sometimes the patient calls a competitor.

What "good" looks like

The practices that hold onto patients aren't necessarily faster clinically. They're better sequenced. When I benchmark refill operations, I look for a few concrete numbers:

  • Median time from refill request to clinical authorization: under 24 business hours.
  • 90th percentile: under 48 business hours. If your P90 is five days, your patients are noticing.
  • Percentage of refills where required labs or check-ins were completed before the refill window opened: above 90%.
  • Percentage of refills held for payment issues: under 3%.

These aren't aspirational. I've seen well-run practices operate at these levels with three clinicians and a shared MA pool. The difference isn't headcount. It's whether the clinic runs on scheduled prompts and structured queues, or on memory and email.

Refill speed as a leading indicator

The reason I push practice owners to put refill turnaround on their weekly dashboard, alongside new patient starts and revenue, is that it moves before revenue does. Churn is a lagging indicator. By the time you notice quarterly revenue softening, the operational failure that caused it happened 60 to 90 days ago.

Refill lag, on the other hand, tells you today what your revenue is going to look like next quarter. If your P90 authorization time creeps from 36 hours to 72 hours in March, you're going to feel it in June retention numbers. Watching the leading indicator gives you a chance to fix the workflow before the churn shows up in the P&L.

Where AI agents and structured workflows actually help

I'm skeptical of AI as a magic-fix pitch, but there are two very specific places where automation moves refill speed materially, and both are boring:

First, the pre-refill scheduling nudge. Thirty days before a refill window opens, the system checks whether the patient has a valid check-in scheduled or completed. If not, it reaches out to the patient with scheduling options, all before any human at the clinic touches the file. This alone eliminates the majority of "we can't refill because we haven't seen you" delays.

Second, the clinical review queue itself. When a refill request comes in with all supporting data (recent vitals, labs, patient-reported symptoms, adherence notes) pre-assembled in a single review view, a clinician can authorize in 45 seconds instead of 8 minutes. Multiply that across 200 refills a week and you've given your provider back most of a clinical day, which they can use for new patient consults.

Neither of these requires the AI to make clinical decisions. The provider still authorizes. The system just removes the reasons the authorization gets stuck. If you want to see what that infrastructure looks like in practice, you can talk to our team about how compounding clinics are running these workflows today.

What to change on Monday

If you're reading this and don't currently measure refill turnaround, start there. Pull the last 60 days of refills, calculate median and P90 time from request to authorization, and break it down by clinician and by delay reason. You'll almost certainly find one or two patterns that account for most of the lag.

Then pick the one that's cheapest to fix. Usually it's the pre-refill check-in scheduling problem, because it's a pure workflow issue that doesn't require any new clinical capacity. Solve that, watch your P90 come down, and watch your quarterly retention numbers respond about 60 days later.

Frequently Asked Questions

Isn't refill speed mostly a pharmacy problem?

In my experience, no. The pharmacy side is usually the fastest link in the chain once it has a valid authorization. The delays live inside the clinic: clinical review, missing check-ins, and payment friction. If your pharmacy partner is genuinely slow, that's a separate conversation, but audit your internal cycle time first.

How do I benchmark my current refill turnaround if we don't track it?

Pick 50 recent refills at random. For each, note the timestamp of the patient's request and the timestamp of clinical authorization. Calculate the median and the 90th percentile. It's not perfect, but it's directionally accurate and takes about two hours to compile.

What's a realistic P90 target for a small practice?

Under 48 business hours is achievable for practices with three or more clinicians and structured queues. Solo-provider practices should target under 72 hours with proper async workflows. If you're above five business days at P90, patients are quietly leaving.

Does automating parts of the refill workflow create compliance exposure?

Automating the scheduling, data assembly, and payment steps doesn't change your clinical decision authority: the licensed provider still authorizes each refill. The compliance questions to work through with your own counsel are around documentation of clinical review, state-specific rules on async visits for controlled or compounded therapies, and record retention. The automation itself isn't the risk; unclear ownership of the clinical decision is.