Prior authorization is the line item nobody puts on the P&L, but it shows up everywhere else: in staff turnover, delayed starts, abandoned therapies, and the quiet erosion of same-day responsiveness that concierge patients pay for. If you run a compounding or concierge practice, you are almost certainly spending more on manual PA work than you think, and the real cost is not the labor hours. It is what those hours displace.

I want to walk through the actual math, the second-order effects most owners miss, and what a reasonable path to automation looks like for a practice that cannot afford a six-month implementation.

The labor cost is the smallest part

Start with the number everyone quotes. The MGMA and CAQH indices put the fully loaded cost of a manual PA somewhere between $11 and $14, with staff time averaging around 20 to 30 minutes per authorization when you include the back-and-forth. For a mid-sized practice running 300 PAs a month, that is roughly $3,600 to $4,200 in direct labor and about 100 to 150 staff hours.

If that were the whole story, most practices could absorb it. It is not. In concierge and compounding settings, the hidden costs scale faster than the visible ones.

Delayed therapy starts and the revenue drag

A compounded hormone protocol, a GLP-1 agonist titration, a specialty injectable: each has a window between the clinical decision and the first dose. Manual PA stretches that window from hours to days or weeks. We see three things happen when that window grows:

  • Patients cool off. The urgency they felt in the exam room fades, and a meaningful percentage never start therapy at all. In our own data across partner clinics, abandonment on authorizations that take more than 7 days runs roughly 15 to 22%.
  • Refills stall. If the initial PA is slow, renewals are slower, because they get deprioritized behind new intakes. Gaps in therapy become gaps in revenue.
  • Pharmacy relationships strain. For practices that lean on a compounding partner, slow PAs mean more "status check" calls from the pharmacy, more one-off scripts, and more friction in a relationship that should be low-touch.

If your average patient contributes $180 a month in recurring revenue and PA delays cause even 20 patients a quarter to drop off before they start, that is $43,200 in annualized recurring revenue you lost without ever seeing it on a report.

The staffing cost of staffing for PA

Here is the part that quietly kills margin. When you hire a medical assistant or PA coordinator, a good chunk of their week disappears into portals, faxes, and hold music. That person was supposed to be your intake quality lead, your refill-request manager, your patient-communication backup. Instead they are logging into CoverMyMeds and waiting for a payer rep.

In concierge practices, this is especially painful because the whole model depends on responsiveness. When your clinical staff is pinned down on PA work, the message queue grows, portal replies slip past 24 hours, and the thing patients are paying a membership for starts to feel like any other clinic. Churn in concierge is a trailing indicator. By the time you see it in renewal rates, the damage was done 90 days earlier in your inbox.

Error rates and rework

Manual PAs fail for boring, preventable reasons: wrong diagnosis code pairing, missing chart notes, outdated formulary assumption, a prescriber NPI typo. Each denial triggers a rework loop that costs more than the original submission because now you are chasing documentation, re-faxing, and often calling the payer to understand what actually went wrong. First-pass approval rates in manual workflows typically sit between 60 and 75%. Every point you move that number improves staff morale as much as it improves throughput.

What this looks like on a real P&L

Let me put rough numbers on a 4-provider concierge practice with a compounding focus, running about 350 PAs per month:

  • Direct labor on PA submissions and follow-up: ~$4,500/month
  • Rework on denials (assume 30% denial rate, 15 extra minutes each): ~$1,100/month
  • Lost revenue from therapy abandonment (conservative 10% of delayed starts): $8,000 to $12,000/month
  • Opportunity cost of staff time redirected from patient communication: hard to price, but visible in membership renewal rates

Call it $14,000 to $18,000 a month in true, loaded cost. Over a year, you are looking at $170K to $215K. For most independent practices, that is a full clinical FTE, or a serious marketing budget, or the difference between a thin year and a good one.

Why compounding and concierge practices get hit harder

Three structural reasons:

Formulary complexity. Compounded medications sit in a gray zone for most payers. Coverage criteria shift, step therapy requirements vary wildly, and the "right" documentation changes by plan. Your staff is not just filing PAs. They are researching coverage policy for each one.

Volume concentrated in a few high-touch drugs. When half your PA volume is GLP-1s, testosterone pellets, or a handful of specialty compounds, a single payer policy change can double your workload overnight. Manual workflows do not flex. People just work later.

Service expectations. Concierge members expect the practice to handle this. They did not pay a membership to call their insurer. So the PA work absorbs upward: into clinical staff, sometimes into the provider's own evenings.

What automation actually needs to do

Most PA "solutions" automate the submission and leave you with the hard parts. The work that matters is upstream and downstream of the portal click. A system worth implementing should:

  • Pull the right chart elements automatically: diagnosis, prior therapies tried, relevant labs, provider notes. If a human still has to assemble the packet, you have not solved the problem.
  • Know the payer's current criteria for the specific drug and flag missing documentation before submission, not after denial.
  • Handle status checks without a human logging into a portal. This is where AI agents earn their keep.
  • Write back to the chart and notify the right person when action is needed, with context, not just a task.
  • Keep a clean audit trail. Your compliance posture should get stronger with automation, not murkier. Confirm specifics with your own counsel, but the documentation discipline a good system enforces usually helps.

The practices getting real leverage from automation are not replacing their PA coordinator. They are freeing that person to own escalations, payer relationships, and the 10% of cases that genuinely need human judgment. That is a better job and a better outcome.

A practical way to measure your own cost

Before you evaluate any vendor, run this exercise for one month:

  • Count PA submissions, denials, and resubmissions. Separate new starts from renewals.
  • Time 20 random PAs end to end, including status checks and documentation gathering.
  • Track days from provider decision to first dose for each PA'd medication.
  • Pull a cohort of patients whose PA took more than 7 days and check 60-day fill rates against patients whose PA cleared in under 48 hours.

You will have a defensible number for what manual PA is costing you, and a baseline to measure any change against. If you want to see how we approach this inside the EMR and with agentic workflows built for compounding and concierge practices, talk to our team at Qintara.

Frequently Asked Questions

How much of PA work can realistically be automated today?

In our experience, 70 to 85% of the end-to-end workflow for common compounded and specialty medications can run without human touch, including document assembly, submission, and status follow-up. The remaining work is genuine exception handling: appeals, peer-to-peer coordination, and unusual coverage situations. That is where you want your experienced staff spending time.

Will automating PA change our compliance exposure?

Done well, it should tighten your documentation trail and make audits easier because every submission is timestamped, versioned, and tied to the chart elements that supported it. Confirm the specifics of record retention, electronic signature, and payer-side requirements with your own compliance counsel, especially if you operate across multiple states.

What is a reasonable timeline to see ROI?

For a practice doing 200+ PAs a month, the labor savings alone usually cover the system within the first quarter. The larger ROI, from faster therapy starts and reduced patient drop-off, shows up in months two through six as your fill rates and renewal numbers move. Expect to spend the first 30 days tuning payer-specific logic for the drugs you actually prescribe.

Does this replace our PA coordinator?

In practice, no. It changes what that role does. The coordinators who thrive after automation become payer-relationship owners and escalation specialists. Their output per hour goes up substantially, and the work is more interesting. The practices that treat automation as a headcount cut usually leave money on the table.

How do we handle compounded medications that payers consider non-covered?

Automation helps here too, but differently. It can quickly identify which plans will not cover a given compound and route those patients to a cash-pay workflow with transparent pricing, instead of burning a week on a PA that was never going to clear. That alone recovers meaningful staff time in compounding-heavy practices.