Every practice administrator eventually stares at the same line item and hesitates over it. You have a stack of prior authorizations that keeps growing, a front desk that keeps getting pulled off the phones to chase them, and a nagging sense that maybe it is time to just hire someone whose whole job is authorizations. Then you look up prior authorization full-time staff salary numbers, see mid-forties to mid-fifties, quietly add benefits in your head, and hesitate again. Is a dedicated seat actually the right call, or are you about to spend seventy-plus thousand dollars a year on a problem that automation and a smarter workflow could absorb?
For a six-provider cardiology group, this is not an abstract staffing debate. Cardiology sits at the wrong end of the prior-authorization curve. Stress tests, nuclear imaging, CT angiography, echocardiograms, cardiac MRI, implantable devices, and most interventional procedures land on payer authorization lists. Where a family medicine group might need an auth on a small fraction of its orders, a cardiology group needs one on a large share of the high-dollar ones. So before you post the job, it is worth running the actual numbers rather than the gut feeling.
Count Auth-Required Orders, Not Patients
The single most common mistake in this decision is measuring the wrong thing. Administrators reason from total patient volume: six providers, roughly a hundred-plus encounters a day, surely that is enough to keep a full-time authorizer busy. But patient count tells you almost nothing. What matters is auth-required orders per week, because that is the only work an authorization specialist actually touches.
Run the count honestly for one week. Pull every order that hit a payer authorization requirement: the imaging, the device implants, the elective interventional cases, the higher-cost cardiac testing. A busy six-provider cardiology group commonly lands somewhere between 40 and 60 auth-required orders in a week, and in a heavy imaging practice it runs higher. Now attach time. A clean authorization that sails through takes 20 to 30 minutes of gather-submit-confirm work. One that bounces into additional-clinical-information territory, or requires a peer-to-peer, can consume an hour or more spread across days of phone tag and portal checking. At 50 orders a week with even a third of them turning messy, you are well past a full-time workload before anyone counts denials and appeals.
That volume test is exactly why the answer diverges by specialty. The same headcount of six providers in behavioral health or general primary care would generate a fraction of these auth-required orders, and the math would say do not dedicate a seat. Cardiology clears the bar that most practices do not.
flowchart TD
A[Order placed by cardiologist] --> B{Requires prior auth}
B -->|No| C[Schedule and perform]
B -->|Yes| D[Gather clinical documentation]
D --> E[Submit to payer portal or fax]
E --> F{Payer decision}
F -->|Approved| G[Schedule the procedure]
F -->|More info needed| H[Pull records and resubmit]
F -->|Denied| I[Peer to peer review]
H --> F
I -->|Overturned| G
I -->|Upheld| J[Procedure delayed or abandoned]
J --> K[Lost revenue and delayed care]The Fully Loaded Cost of One PA Seat
Once volume clears the bar, the next honest step is to cost the hire correctly, because the salary you budget is never the cost you carry. Base pay for a prior authorization specialist generally sits between 42,000 and 55,000 dollars, and someone with real cardiology and imaging authorization experience commands the top of that band. But the base is roughly two-thirds of the true number.
Layer in the rest. Payroll taxes run about 8 percent. Health benefits, retirement contribution, and other benefits typically add 20 to 30 percent of base. Paid time off means you are paying for a seat that is empty two to three weeks a year. Then the operational tail: a workstation, a phone line, portal and software seats, and the manager time it takes to supervise the role. Add it up and a 48,000-dollar base becomes a 62,000 to 78,000 dollar all-in commitment. In a high cost-of-living metro, budget higher.
Set that against what the seat protects. If a single delayed authorization bumps a stress echo or a CT angiogram, you have an empty high-value slot and a patient whose care slipped. A handful of those a month is real money, and it is genuinely part of the case for the hire. The problem is not that the role lacks value. The problem is what happens to that value when the one person holding it is unavailable.
The Single-Point-of-Failure Trap
Here is the failure mode nobody puts in the job description. You hire one dedicated authorizer, the pipeline runs smoothly, and everyone exhales. Then that person takes a week of vacation, or catches the flu during a busy imaging stretch, and the entire authorization pipeline simply stops. Nobody else knows the payer portals, nobody has been tracking which cases are mid-appeal, and cardiology cases start getting bumped for lack of authorization at exactly the moment your schedule is full.
A dedicated seat concentrates risk. All the tribal knowledge of which payer wants which clinical fields, which nurse reviewer to ask for on a peer-to-peer, which cases are three days from a deadline, lives in one head. When that head is out, you are not down 20 percent of capacity, you are down close to 100 percent of a critical function, and the backlog it creates takes weeks to unwind. For a cardiology group where the auth pipeline gates high-dollar procedures, a stalled week is not an inconvenience, it is a measurable revenue dent and a patient-safety exposure when time-sensitive testing waits.
This is the flaw in framing the decision as hire-versus-do-nothing. The real choice is about how the work gets structured so that it never depends on one person having a free hour or a healthy week. Some of the work genuinely needs a clinician-adjacent human: the peer-to-peer review, the judgment call on which additional records will actually satisfy a reviewer, the appeal narrative. But a large share of it, the gathering, the submitting, the status-checking, the resubmitting, is repetitive queue-work that punishes you precisely because it depends on human availability.
Where a PA Outsourcing Company Fits, and Where It Doesn't
The obvious alternative to hiring is handing the whole function to a prior authorization outsourcing company. For some groups this works, and it neatly solves the single-point-of-failure problem because the vendor carries the coverage. But it comes with tradeoffs a cardiology group should weigh with open eyes.
Outsourcers typically price per authorization or as a percentage of the work volume, which means your cost scales with your auth load rather than staying flat, and cardiology's auth load is high. You also hand off control of turnaround: when a stress echo authorization is sitting in a vendor's queue behind other clients' work, you cannot walk over to a desk and ask about it. And PHI leaves your building, so the vendor relationship lives or dies on a signed business associate agreement and real security posture, not a handshake. For a straightforward, lower-volume auth book, outsourcing can be cleaner than hiring. For a high-volume, time-sensitive cardiology pipeline where a bumped procedure is expensive, the loss of direct control and the per-auth pricing curve often make full outsourcing a harder fit than it first appears.
That leaves a third path that has quietly become the strongest option for volume-heavy specialties: keep the function in-house, but stop staffing it as if it were all human work.
Automating the Repetitive Half So the Pipeline Never Stops
The reason the hire-versus-outsource debate feels stuck is that both options treat prior authorization as one indivisible job. It is not. It splits cleanly into repetitive queue-work and genuine clinical judgment, and only the second half needs a person. AI prior authorization software in 2026 has gotten good enough at the first half, the documentation gathering, the payer submission, the relentless status-checking and resubmission, that the economics of the whole decision change.
This is where CallSphere Health reframes the problem for a cardiology group. Instead of buying a full-time seat to sit in payer portals all day, you route the repetitive submission and follow-up work through automation that does not take vacation, does not get pulled onto the phones, and does not concentrate all the payer knowledge in one fragile head. The hands-off billing and claims capability keeps authorizations and their downstream claims moving on their own cadence, with automated follow-up on the ones that stall, so a busy week or a sick day never stops the pipeline. Just as importantly, the AI front desk answers 100 percent of calls around the clock and books appointments on its own, which pulls the phone load off the exact staff who otherwise abandon authorization work to pick up line two. The interruptions that starve the auth queue simply stop landing on your people. You can see how those pieces connect on the /features page, and the /pricing page lays out the flat monthly cost against the 62,000-to-78,000-dollar all-in reality of a dedicated hire.
The point is not that software replaces clinical judgment. Peer-to-peer reviews and appeal strategy still need a human who understands cardiology. The point is that you no longer pay a full salary for a person to spend most of their day on portal clicks and hold music. You reserve the human for the judgment calls and let automation carry the volume that made the seat feel necessary in the first place.
Making the Call for Your Group
So walk it back to a decision you can actually make this quarter. First, count your auth-required orders for one real week, not your patient volume. If a cardiology group is running 40-plus a week with a meaningful denial tail, you have enough work to justify dedicating resources to it, full stop. Second, price the hire honestly at its fully loaded 62,000-to-78,000-dollar number, not its base salary, and set that against the revenue your bumped procedures cost you. Third, and this is the step most administrators skip, ask which half of the work you are actually buying. If most of the seat is repetitive submission and status-chasing, a full-time hire is an expensive way to solve a problem that automation absorbs more reliably and without the single-point-of-failure risk.
The honest answer for most six-provider cardiology groups is that yes, you have the volume to justify serious investment in prior authorization, but no, a lone dedicated human is not the most durable way to spend it. Automate the queue-work so the pipeline never stops, keep a person for the clinical judgment, and you get the coverage without buying the fragility. Run your one-week order count before you post the job, and the numbers will tell you which half of that sentence applies to you.