When the Algorithm Says No: What Medicare's WISeR Model Means for Your Practice's Cash Flow

For six months, a small group of practices in six states has been running an experiment none of them volunteered for. On January 1, 2026, the Centers for Medicare & Medicaid Services launched the Wasteful and Inappropriate Service Reduction model — WISeR — bringing algorithm-assisted prior authorization into Original Medicare for the first time. New Jersey, Ohio, Oklahoma, Texas, Arizona and Washington drew the short straw.

Now the Government Accountability Office has weighed in with a determination that opens the model to congressional review, and with it the real possibility of repeal. Suddenly a program built to run through 2031 has an expiration date nobody can predict.

Here is the part most coverage is missing. Whether WISeR survives the year or not, the operational lesson it has already taught is permanent — and every independent practice in the country should be taking notes.

What actually changed on January 1

Original Medicare has historically been the one payer that did not make you ask permission first. You documented, you billed, you got paid, and any fight over medical necessity happened on the back end. That asymmetry is precisely why so many practices treated Medicare denials as a rounding error and built their denial-management muscle around commercial and Medicare Advantage plans instead.

WISeR inverted that. Under the model, selected services in the six participating states require a prior authorization request submitted through a model participant's electronic portal. The participants use machine-assisted review to triage those requests, with licensed clinicians making every final non-payment recommendation. CMS has said it will audit participants for inappropriate denials and hold them to an accuracy metric on their determinations.

The early numbers from Texas tell the story better than any policy summary. Roughly 62 percent of prior authorization requests were approved on first pass. After a physician reviewed the initial non-authorizations, approval climbed to about 84 percent.

Read that gap again. Somewhere around one in five requests that the system initially declined was, on human review, a perfectly legitimate claim. Those requests were not wrong. They were incomplete, or they were submitted in a form the review process could not act on.

The gap between 62 and 84 is a documentation problem

This is the sentence we would put on the wall of every practice we work with: the difference between a first-pass approval and a first-pass denial is almost never clinical. It is almost always documentation.

The American Hospital Association made this exact point in its comments to CMS on the model. The submission platforms in play do not use FHIR transactions, which means they do not pull from the EHR. Staff have to manually collect and extract clinical information and re-key it into a portal — and then do it a completely different way for commercial patients than for Original Medicare patients. Two workflows, two sets of habits, two chances to leave out the one note that would have gotten a first-pass yes.

Multiply that by a practice running thin on front-office staff, and the 62 percent figure stops being surprising and starts being predictable.

The financial consequence is not the eventual denial. Most of those cases got approved. The consequence is the delay — the days added to accounts receivable while a request bounces, gets supplemented, and goes back through review. For a practice with a tight cash position, thirty extra days in A/R on a meaningful slice of Medicare volume is not an administrative inconvenience. It is a payroll problem.

Why the repeal question does not let you off the hook

It would be easy to read the GAO news and conclude that this is a temporary irritation for four states, two states, or none. That would be a mistake, for three reasons.

• First, the direction of travel is one-way. CMS's broader 2026 posture is not subtle. The agency's own materials on the CY 2026 Physician Fee Schedule describe expanded use of data review to flag outliers and thin documentation. The skin substitute reclassification in the same rule — moving most products to incident-to supply status at a flat rate of $127.14 per square centimeter — came after Part B spending on those products went from $252 million in 2019 to more than $10 billion in 2024. CMS has said the change is expected to cut spending on those products by close to 90 percent. That is not an agency looking for fewer ways to scrutinize claims.
• Second, Medicare Advantage is not going anywhere. Whatever happens to WISeR in Original Medicare, MA plans have run prior authorization at scale for years, and MA penetration keeps climbing. The documentation discipline WISeR demands is the documentation discipline MA already demanded. Practices that build it now are building it for the payer mix they will actually have.
• Third — and this is the one that costs practices money quietly — repeal does not refund the A/R you already lost. Six months of first-pass denials in six states have already aged. Those claims are sitting somewhere, and some of them will die of old age if nobody works them.

What we would do about it

At VOSKPO LLC, we do not treat prior authorization as a form-filling function. We treat it as the front edge of the revenue cycle, because that is what it has become. Here is the operating posture we would put in place, and it costs nothing to start this week.

• Build the packet before the request, not after the denial. For every service line subject to prior authorization, define in advance exactly which clinical elements the payer needs to see, and make assembling that packet part of scheduling — not part of billing. The 62-to-84 gap exists because information that already lived in the chart did not make it into the request. Fixing that is a workflow change, not a technology purchase.
• Standardize across payers instead of maintaining parallel workflows. The AHA's complaint about running one process for commercial and another for Original Medicare is real, and the answer is not to wait for the portals to converge. Build one internal standard that satisfies the strictest payer you deal with, and use it everywhere. You will over-document for some payers. That is cheaper than under-documenting for one.
• Appeal the non-authorizations. All of them. When a physician review turns 62 percent into 84 percent, the message is unambiguous: the first answer is not the final answer, and practices that accept it are leaving collectable revenue on the table. Track your first-pass rate as a named metric with an owner. If nobody owns it, it does not improve.
• Age your prior-auth-related A/R separately. Blend it into the general bucket and it disappears. Break it out and you can see, in dollars, what the process is costing you — which is the only number that will ever get the workflow fixed.
• Watch the policy, but do not wait on it. If WISeR is repealed, you will have built a denial-prevention discipline you needed anyway. If it survives and expands beyond six states, you will already be running it. There is no version of the next three years where tighter documentation hurts you.

The uncomfortable summary

This is the work VOSKPO does every day for independent practices: engineering the revenue cycle so that the money that was earned actually arrives, on time, without a fight. Practices that tighten first-pass documentation typically see measurable movement in days-in-A/R within a quarter — a typical outcome, not a guarantee, and one that depends heavily on payer mix and starting baseline.

A machine looked at Medicare claims in six states and said no to roughly four in ten of them. A doctor looked at the same claims and said yes to most of those. The claims did not change. The paperwork did.

That is not an argument about artificial intelligence in healthcare. It is an argument about whether your practice is submitting complete work. VOSKPO exists because for most independent practices, under real staffing pressure, the honest answer is not yet.

The GAO determination may end WISeR. It will not end the scrutiny. Build for the scrutiny.

Ready to see where your first-pass denials are actually coming from? VOSKPO offers a free revenue review for independent practices — a straight look at your denial patterns, your A/R aging, and the specific documentation gaps costing you cash. Start at voskpo.com.