Prior authorization has arrived in Original Medicare - decided with AI and paid on the denial. The hospitals that come through the next two years in one piece are the ones fixing their documentation and their data pipes right now. For forty years, Original Medicare paid first and asked questions later. That changed this year, in six states, with almost no fanfare. Most coverage says prior authorization has finally reached traditional Medicare, which is true enough. The part worth a rural CFO's time is how the new reviewers get paid. They earn a percentage of the spending their denials prevent. Pay someone to find waste, and they will find it, whether or not it is there.
The program is called WISeR, short for Wasteful and Inappropriate Service Reduction. It runs from 2026 through 2031 in New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington. Private technology companies use AI, backed by human clinicians, to check a short list of non-emergency services against Medicare's existing coverage rules before the bill is paid. The services are familiar ones that most Medicare Advantage plans already gate: skin substitutes, nerve stimulators, epidural steroid injections, and knee arthroscopy for osteoarthritis. CMS is right to point out that it has not changed a single coverage rule. On the surface, this is housekeeping.
Look closer at the payment, and a well-documented failure mode comes into view.
Guest Author: Yemi Mateola, MBA, CPHIMS - Founder & CEO, Stratevora Technologies | Healthcare Vertical AI for Hospital Revenue Cycle, Interoperability & Adoption.
A number stands in for the goal
In 2019, Michael Harris and Bill Tayler, writing in Harvard Business Review, named the trap "surrogation." Give people a fuzzy goal and a sharp number meant to measure it, and they start serving the number. They borrow three conditions from Daniel Kahneman and Shane Frederick, and WISeR meets all three. The goal is abstract: appropriate, medically necessary care. The number is concrete and visible: dollars saved for the contractor, a clean approval rate for the hospital. Once the number is what pays, the swap happens on its own, without anyone choosing it.
Wells Fargo is the case that Harris and Tayler reach for. The strategy was to build lasting customer relationships. The measure was accounts per customer. Staff opened as many as 3.5 million accounts nobody asked for, hit their targets, and torched the relationships the targets were meant to build. Steven Kerr described the same problem in 1975 under a blunter title: the folly of rewarding A while hoping for B. WISeR pays for averted spending and hopes for good care. A clinician signing each denial does not change which way the money points.
CMS clearly saw the risk and built a guardrail. The contractor's cut rises and falls with quality scores and provider-experience ratings, so sloppy denials cost the reviewer money. Credit where it is due. Even so, a quality modifier sits on top of the main incentive, and the main incentive is to deny. Economists have a colder name for the setup. Jensen and Meckling called it the agency problem back in 1976: hand your decisions to an agent whose interests run a different way, and you own the gap between you. WISeR runs that gap three deep. The patient sits at the top. CMS acts on the patient's behalf. The contractor acts on CMS's behalf and gets paid to say no. Better oversight can tighten the bolts. The shape is baked in.
We have run this experiment
Medicare Advantage has already run this experiment at scale. In 2024, MA plans handled almost 53 million prior authorization requests and turned down 4.1 million of them, about 7.7%. The number that should stop a CFO cold comes next: among the denials that were appealed, 80.7% were reversed. When someone pushes back, the denial usually does not survive.
Now the other half. Only 11.5% of denials were appealed at all. Put the two figures together, and the design reveals itself. Most denials are never appealed, and the ones that are usually collapse. The friction does the work. Fighting every denial costs hours and staff that a stretched hospital does not have, so far too many denials stick. In 2022 alone, hospitals spent an estimated $19.7 billion chasing denied claims, by Premier's estimate, and across the industry, roughly two-thirds of denials are never resubmitted.
The spread between insurers gives the game away. In 2024, the overturn rate on appeal ranged from about 51% at one large plan to better than 95% at another, both working under the same rules. A denial rate looks less like a fact than a setting each payer chooses, and each plan turns the dial to fit its own math. Rural and community hospitals feel it hardest, because they have the fewest people free to chase an appeal and the least cash to wait on.
In its defense, CMS built WISeR to touch only low-value services with published criteria, with a human clinician on every denial. The early results from Washington are already in, and they are not encouraging. A report this April from Senator Maria Cantwell, built on data from 16 Washington hospitals, found that procedures once cleared in about two weeks were taking four to eight weeks under WISeR. At one health system, roughly 100 patients were left waiting for epidural steroid injections. CMS set response targets of one to three days. The hospitals are seeing fifteen to twenty. The criticism now spans both parties in Congress, which rarely happens on health policy.
The Collision No One is Pricing In
Set WISeR next to a rule that rarely comes up in the same conversation. CMS-0057-F, the Interoperability and Prior Authorization rule, requires the major payers, including Medicare Advantage, Medicaid, and exchange plans, to run prior authorization through standardized FHIR data connections, with the prior-authorization and provider-access components generally due by January 1, 2027. The insurers' own 2025 pledge targets the same horizon: by 2027, at least 80% of electronic prior authorizations are expected to be returned in real time. One force is making prior authorization more burdensome within Original Medicare. Another is making it electronic across everything else. Both land in the same eighteen months.
Digital and electronic are different things, and the difference is about to matter. A PDF dropped into a payer portal is digital, and it still waits in a human queue. Electronic means your system builds the request directly from the chart, checks it against the rules, and sends it machine-to-machine. The work even has named building blocks: the HL7 Da Vinci implementation guides for coverage requirements and documentation that tell your system what a payer needs before you submit a claim. A hospital with that capability handles WISeR's added load and the 2027 shift in stride. A hospital still running prior authorization on fax machines, phone calls, and portals absorbs every delay both create. The capability takes quarters to build, and the WISeR clock is already running. A CIO who starts the data work in late 2026 is behind before the first request goes out.
The Gate is the Whole Game
Here is the part almost everyone has missed, and the part a rural operator should care about most. WISeR comes with a way out. CMS has described a gold-card path: a provider with a roughly 90% affirmation rate can be exempted from prior authorization and prepayment review for those services for a period, with the final details due in 2026. Sit with that for a second. The model is grading paperwork. Often, a denial here means the record did not prove the need on the first pass, even when the patient needed the procedure all along.
The denials that wreck a first-pass rate are mundane. A conservative-treatment note is missing before a knee scope. A wound measurement that never reached the chart before a skin graft. A nerve study, sitting in a scanned image, that the reviewer's software could not open. Each is a small documentation gap, the kind a machine flags in seconds and a busy clinic never sees coming. The medicine was right. The paperwork lagged.
That turns the problem from a threat into a project, the kind with owners and steps. Clean intake, complete documentation captured the first time, and a data layer that passes it on without anyone retyping it: those are things a CFO and a CIO can own. The work is unglamorous: measure the first-pass approval rate by service line, trace every denial back to the gap that caused it, and close that gap at registration and at the point of care, where it starts. The hospitals that treat that 90% as a performance target and hit it can earn their way out. Those who treat it as Washington's problem remain in the model for six years.
The timing makes the gap cruel. Payers are already using AI to decide. Barely one in five providers use AI anywhere in the denials process, per Bain and KLAS, while 61% of physicians tell the AMA they believe insurers' AI is already pushing up denials. A hospital answering an automated denial engine by hand loses on speed and volume before anyone weighs the medicine. I have called this the AI arms race for a while now, and the rural hospital starts on the wrong side of it. The medicine was usually fine. What is missing is the tooling and the discipline to prove it.
What to do Before the Gate Closes
WISeR may not survive its political fight. Congress has already moved to cut its funding, and the courts could still reshape it. The question that outlasts all of that is simpler: can your hospital prove medical necessity, on the first submission, at the speed a machine reads? That capability pays off no matter what happens to WISeR, because Medicare Advantage already works this way, commercial plans already work this way, and 2027 makes the electronic version the cost of doing business. WISeR just started the clock on a weakness that was always going to surface.
A human reason sits under the spreadsheet, too. Behind those 4.1 million denials are patients who waited for care their doctor had already ordered, and many who gave up. A hospital that can respond quickly protects its margins and its patients in one move. The same clean record that clears the gate is the one that spares someone the wait. That is the work we do at Stratevora: enterprise capability on a rural budget. The cheapest denial is the one that never happens, and the cheapest time to lay the pipes is before the gate closes.
If you run or sit on the board of a community hospital in one of the six states, the next move is concrete. Pull your selected service documentation against the published Medicare criteria. Measure your true first-pass approval rate against the 90% bar. Find out, honestly, whether your prior authorization is electronic or just digital. If you sit on the audit or technology committee, ask management one question this quarter: What is our first-pass approval rate for these services, and who owns moving them forward? The model is betting your answer is no. Prove it wrong.
#HealthcareLeadership #PriorAuthorization #RuralHealth #HealthIT #AIinHealthcare #HealthcareInnovation #HealthcareAI #HealthIT #RevenueCycleManagement #PriorAuthorization #Medicare #Interoperability #FHIR #DigitalHealth #HospitalLeadership #RuralHealth #HealthcareTechnology #RevenueIntegrity #ClinicalDocumentation #HealthcareTransformation #AIinHealthcare #ValueBasedCare #HealthcareOperations #HealthTech #Stratevora
References
Centers for Medicare and Medicaid Services. WISeR Model overview, Fact Sheet, and Frequently Asked Questions (services, six states, 90 percent affirmation gold-card pathway, with details being finalized for 2026). Federal Register, "Implementation of Prior Authorization for Select Services for the WISeR Model," July 1, 2025.
U.S. Government Accountability Office. Decision B-337994, May 2026 (WISeR Notice is a rule under the Congressional Review Act; participants paid a quality-adjusted percentage of averted spending).
U.S. Senate Committee on Finance. Ranking member release on the WISeR Congressional Review Act resolution of disapproval, May 20, 2026.
U.S. House Committee on Appropriations. FY2027 Labor-HHS appropriations markup, June 9, 2026.
KFF. "Medicare Advantage Insurers Made Nearly 53 Million Prior Authorization Determinations in 2024" (4.1 million denied, 7.7 percent; 11.5 percent appealed; 80.7 percent overturned; insurer overturn range about 51 percent to 95 percent).
Premier, Inc. Trend Alert on claims denials, 2022 (hospitals spent an estimated $19.7 billion overturning denials). Resubmission estimate (roughly two-thirds of denials never resubmitted): HFMA, Change Healthcare, and AMA industry reporting.
Office of U.S. Senator Maria Cantwell and the Washington State Hospital Association. Report on WISeR prior authorization delays, April 2026 (16 hospitals; two-week approvals now four to eight weeks; about 100 patients awaiting epidural steroid injections).
Centers for Medicare and Medicaid Services. Interoperability and Prior Authorization Final Rule (CMS-0057-F), FHIR API requirements for impacted payers, compliance beginning January 1, 2027.
AHIP. "Health Plans Take Action to Simplify Prior Authorization," 2025 (industry pledge; at least 80% of electronic prior authorizations answered in real time by 2027).
American Medical Association. 2025 Prior Authorization Physician Survey (61 percent of physicians concerned that health plans' use of AI is increasing denials).
Bain and Company and KLAS Research. 2025 healthcare AI survey (about one in five providers use AI in denials management).
Harris, Michael, and Bill Tayler. "Don't Let Metrics Undermine Your Business." Harvard Business Review, 2019. Kerr, Steven. "On the Folly of Rewarding A, While Hoping for B." Academy of Management Journal, 1975. Jensen, Michael, and William Meckling. "Theory of the Firm." Journal of Financial Economics, 1976.