Prior Auth Burden Hours and Cost for Physician Practices

Prior authorization now costs a typical physician 13 to 16 hours a week and somewhere between $2,161 and $3,430 a year per physician in direct prior authorization costs. This piece maps that burden line by line, using the numbers that professional associations, government analysts, and health policy researchers have actually published, so a practice can see what it's spending before anyone starts pitching a fix.

Prior authorization, as CMS defines it, is a payer's requirement that a physician get approval before delivering a medication, test, procedure, or piece of equipment, and before the practice ever submits a claim for payment. The idea started narrow: insurers wanted a checkpoint against low-value or unnecessary care, a way to flag the rare outlier before it became an expensive one. That's not what it is anymore. The requirement has spread across nearly every specialty and now touches routine prescriptions and standard procedures that used to sail through without a second look. Major commercial payers and Medicare Advantage plans broadly expanded their prior authorization lists somewhere between 2024 and 2026, and practices that didn't update their internal checklists in 2025 are now getting denials on services they've billed the same way for years. The service didn't change. The rule around it did.

Part of the friction is structural. Each insurer runs its own approval criteria, and those criteria shift without much warning. Reporting in Medical Economics notes that many authorizations still move by phone and fax, not through any electronic system built for the job. The obvious consequence follows: because provider EHRs often don't talk to insurer systems, staff end up re-entering the same clinical data by hand, or faxing it, because there's no pipe connecting the two records. That's the environment the hours and dollars below actually happen inside of.

The weekly time bill: how many hours prior authorization consumes per physician

Three separate organizations, using three different methodologies, landed in roughly the same place. The AMA's 2025 Prior Authorization Physician Survey, based on 1,000 practicing physicians, found an average of 40 PA requests per week consuming about 13 hours of physician and staff time. RAND's July 2025 commentary put it a bit higher: 43 requests weekly, over 16 hours spent filling out forms, sitting on hold, and drafting appeals. The American College of Physicians landed at 45 requests per physician per week and 14 hours of combined staff and physician time. Different surveys, different physician samples, same order of magnitude: 40 to 45 requests, 13 to 16 hours. Medical Economics summed it up: nearly two full business days a week, gone.

That's per physician. Run the arithmetic for a three-doctor practice and the number looks like a second full-time job spread across the office, which is exactly what's happened at a meaningful share of practices. The 2025 AMA survey found that 40% of physicians now employ staff whose entire job is prior authorization, not staff who handle it alongside billing or scheduling, but people hired specifically to chase approvals.

The per-transaction time is where the range gets wide, and the width itself is a data point. A 2026 AJMC systematic review pooling eleven separate studies found that a single prior authorization takes anywhere from 25 to 64 minutes to complete, a range wide enough to be a data point in itself. That's not sloppy measurement. It reflects the fact that a simple prescription refill authorization and a complex imaging or surgical authorization are not the same task, and lumping them into one average would hide more than it reveals.

What those hours cost in dollars: per-transaction, per-physician, and per-year figures

Money follows time, and the same AJMC review put a dollar figure on each transaction: $21 to $49 per prior authorization, a range that reflects the meaningful variation in complexity across different types of requests. Neither number is "the" cost. Depending on what's being authorized, the cost may be $21 or $49 per prior authorization.

Zoom out to the annual level and the American College of Physicians estimates primary care practices spend $2,161 to $3,430 per physician per year on prior authorization specifically. It's primary care. Specialty practices, especially ones ordering advanced imaging, biologics, or surgical procedures with heavier documentation requirements, sit well above that range.

Prior authorization is also just one line item inside a much bigger bill. A ReferralMD analysis cited in Medical Economics put total health-plan interaction costs, meaning phone time, peer-to-peer reviews, appeals, and rework combined, at roughly $68,000 per physician per year. That figure is not a prior authorization number on its own; it's the whole administrative relationship with payers, and conflating it with the ACP's narrower estimate would overstate what PA alone costs. Reporting in Medical Economics describes practice spending on prior authorization staffing as having grown substantially in recent years, a trend consistent with the volume and dedicated-staff figures above.

Physicians clearly feel the financial injury, even where reimbursement hasn't caught up to it. A Physicians Practice survey cited in Medical Economics found that 97% of physicians believe they should be paid for prior authorization work. Currently, payment for that work is not standard practice.

The dedicated-staff question: when PA volume forces practices to hire for it

The AMA survey found that forty percent of physicians now employ staff dedicated exclusively to prior authorization. Not a biller who squeezes in PA calls between other tasks. A role that exists solely to obtain approvals and generates no clinical revenue on its own.

For a large health system, that overhead disappears into a bigger budget. For an independent practice or a small group, it's a real line item: salary, benefits, training, and the opportunity cost of a hire who could otherwise be doing something that bills. Medical Economics describes the trend bluntly, noting that some groups now run full-time authorization specialists whose only job is chasing approvals, overhead that small offices can ill afford.

There's a compounding cost caused by staff turnover and inconsistent payer rules that doesn't appear in any salary line. A biller who's spent two years learning that one insurer wants step therapy documentation submitted in a specific format, while another wants the peer-to-peer scheduled before the denial letter even arrives, is carrying institutional knowledge that took real time to build. When that person quits, the practice doesn't just lose a headcount. It loses the accumulated, payer-specific know-how that made the process faster, and has to rebuild it from zero, at cost.

That turnover risk connects directly to burnout. The AMA survey found that 94% of physicians say prior authorization contributes to burnout. When burnout pushes a physician or a staff member out the door, the PA workload doesn't vanish. It redistributes onto whoever's left, which raises their load and their own risk of leaving next.

How payer behavior drives the volume: denial rates, insurer variation, and shrinking appeal windows

Denial rates aren't uniform, and the gaps between insurers are large enough to change how a practice should think about payer mix. KFF's 2025 analysis, covering the 14 largest insurers in each market (25 million Medicare Advantage enrollees, 36 million Medicaid MCO enrollees, and 10.8 million ACA Marketplace enrollees), found average standard prior authorization denial rates of 12% in Medicare Advantage, 14% in Medicaid managed care, and 18% in ACA Marketplace plans.

Averages flatten what's actually a wide spread. In Medicare Advantage, denial rates ranged from 5% at Elevance to 17% at UnitedHealth Group. In Medicaid managed care, a large regional plan denied 2% of requests while Independence Health Group denied 23%. Care Health Plan denied 2% of requests while Independence Health Group denied 23%. In the ACA Marketplace, GuideWell sat at 3% while Centene reached 25%. A practice serving a heavy UnitedHealthcare or Centene population is dealing with a structurally different denial environment than one serving mostly Elevance or a large regional plan. Staffing decisions should reflect that reality.

Physicians rate the burden by payer in similarly stark terms. The AMA survey found 75% of physicians describe UnitedHealthcare's prior authorization burden as high or extremely high, followed by Humana at 65%, Anthem/Elevance and Aetna both at 61%, Cigna at 59%, and Blue Cross Blue Shield at 56%.

Meanwhile the window to fight back is closing. UnitedHealthcare cut its peer-to-peer review request window from 30 days to 14 days from the denial date, effective in the first quarter of 2025. Humana Medicare Advantage plans shortened expedited appeal windows from 72 hours to 48 hours for urgent requests. Several Blue Cross Blue Shield state plans cut standard appeal windows from 60 days to 30 days on non-urgent denials. None of that changes how often a service gets denied. It just shrinks the room a practice has to respond once it does.

muni.health's 2026 guide identifies the most common stated reason for denial as "not medically necessary," the dominant reason cited across major commercial insurers, with incomplete step therapy documentation and "experimental or investigational" classifications trailing behind. And this isn't a static problem: 74% of physicians in the 2025 AMA survey report that denials have increased over the past five years. Sixty percent are worried that AI-assisted review will push that number higher still, which is exactly where the burden heads next.

Algorithmic denials and what AI-assisted review adds to the practice's workload

Speed and fairness are not the same thing, and the shift toward algorithmic review has made that gap obvious. Decisions that used to take several business days for a human reviewer now come back in hours, but the AMA's 2025 survey data, as cited in industry reporting, shows denial rates drawing concern from physicians who worry AI-assisted review favors speed over accuracy. Sixty percent of physicians say they're concerned augmented intelligence will push denial rates even further up.

Reporters and lawmakers made the lack of transparency public rather than theoretical. Reporters uncovered an automated tool, referred to as PXDX, that rejected hundreds of thousands of claims at an average review time of 1.2 seconds each, and Connecticut lawmakers responded with a 2025 bill that would bar insurers from making coverage decisions using AI alone. Lawsuits against UnitedHealth and Humana over algorithmic denial tools are working their way through the courts now, and the outcomes will decide whether a denial produced by an automated system creates a different kind of legal liability than a human one does.

Regulators are moving too, if slowly. CMS has moved to require greater transparency from payers around prior authorization decisions, an attempt to keep an algorithm from overruling a physician's judgment with no explanation attached.

None of this actually shortens the practice's workload, though. A faster denial still triggers the same appeal, the same peer-to-peer scheduling, the same documentation gathering on the practice's end. Speed helps the insurer's side of the ledger; it doesn't touch the practice's. If anything, it compresses the window the practice has to respond, since faster denials paired with shorter appeal deadlines leave less runway on both ends.

The "human review" backstop that insurers point to doesn't hold up well under scrutiny, either. Only 24% of physicians in the 2025 AMA survey say medical necessity denials are consistently reviewed by an appropriately qualified clinician, and only 16% of physicians who've gone through a peer-to-peer review say the health plan's representative often or always had the right qualifications for the case. The safety net insurers describe and the one physicians actually encounter appear to be two different things.

The appeal calculus: high overturn rates that practices rarely act on

KFF's 2025 data shows 67% of appealed Medicare Advantage prior authorization denials get overturned. Medicaid managed care overturns 47% of appeals, and ACA federally facilitated Marketplace plans overturn 43%, correction rates far from marginal. Those aren't marginal correction rates. They suggest the initial denial is wrong close to half the time, or more, once someone actually pushes back.

The insurer-level spread on overturns tells the same story from a different angle. In Medicare Advantage, Centene overturns 93% of appealed denials, while Kaiser Permanente overturns 40%. In Medicaid, UnitedHealth Group overturns 81% against CVS's 22%. Whatever's driving initial denial accuracy, it clearly differs a great deal by payer, not just by clinical circumstance.

The KFF report shows that prior authorization denials are rarely appealed at all. That's the gap that matters most in this entire discussion. If most appeals succeed, and most denials never get appealed, then a meaningful share of revenue a practice is legitimately owed simply gets written off because nobody on staff had the hours, the documentation trail, or the institutional memory to fight it, even though the payer was wrong. HFMA analysis drawing on Kodiak Solutions data found that up to 65% of denied claims never get reworked at all, a figure spanning all denial types but especially damning here given how often PA appeals actually succeed when someone bothers to file them.

Layer the shrinking appeal windows from earlier onto this math and the picture gets worse, not better. Less time to file means fewer appeals get filed in the first place, which means the practice recovers less money even in cases where the original denial was flatly incorrect. That's not a collections problem that better follow-up habits would fix. It's a structural gap between what a practice is owed and what it actually collects, and it compounds every billing cycle it goes unaddressed.

What the 2025 insurer reform pledge did and did not change

On June 23, 2025, more than 60 health insurers, including CVS Health's Aetna, UnitedHealthcare, Cigna, Humana, Elevance Health, and Blue Cross Blue Shield, publicly pledged to voluntarily reform their prior authorization programs, with implementation staggered across 2025 through 2027, Duane Morris reported. One of the specific commitments: every medical necessity denial would be reviewed by a licensed, qualified clinician.

Measure that commitment against what physicians actually report and the gap is immediate. The 2025 AMA survey found only 24% of physicians say that kind of review is happening consistently. The pledge was made in June. The experience on the ground, at least as physicians describe it, hadn't caught up by the time the survey closed.

Skepticism runs deep on both sides of the transaction. Only 33% of physicians believe this pledge will produce a meaningful difference, a number shaped by prior rounds of insurer promises that didn't stick. A July 2025 KFF poll found only about four in ten adults believe insurers will actually follow through, while roughly six in ten call it "not too likely" or "not at all likely.""

The pledge is voluntary, which is exactly why the skepticism is warranted, and it differs from something with actual teeth: the CMS Interoperability and Prior Authorization Final Rule, known as CMS-0057-F. That rule requires Medicare Advantage, Medicaid, and CHIP plans, though not ACA marketplace issuers on federally facilitated exchanges, to respond to routine prior authorization requests within seven calendar days and urgent requests within 72 hours, starting in 2026, with FHIR-based electronic submission pipelines mandated by 2027. One of these is a promise. The other is a compliance deadline with a regulator behind it. Practices tracking their own exposure should be watching the rule, which carries a compliance deadline with a regulator behind it.

What the burden costs a practice when

Line these figures up together and a clear pattern emerges: the cost of prior authorization isn't one number, it's a stack of numbers that hit a practice at different points in the calendar and different points in the revenue cycle. Thirteen to sixteen hours a week land on staff time before a single claim ever gets denied. Twenty-one to forty-nine dollars per transaction lands the moment a request goes out, regardless of whether it's approved. Two thousand to thirty-four hundred dollars a year lands on the practice's books whether or not anyone ever tallies it as a distinct line item.

Then the denial arrives, and a second cost clock starts. If the practice appeals, the odds of reversal run as high as two-thirds in Medicare Advantage, and the money was very likely owed all along. If the practice doesn't appeal, and most don't, that revenue simply disappears because nobody had the bandwidth to make the case a second time. Shrinking appeal windows and the speed of automated denials are squeezing that second clock from both directions at once. The 2025 insurer pledge addresses some of the process on paper; the 2026 CMS rule enforces a piece of the timeline by law. Neither one, on its own, touches the hours or the dollars documented above. Those keep accruing on the practice's ledger regardless of what insurers promise next.