What We’re Seeing From Wiser, and Why CMS Should Stay the Course

Early experience in Texas and New Jersey suggests Congress should allow the six-year Medicare model to continue. What we’re seeing from WISeR and why CMS should stay the course. Early experience in Texas and New Jersey.

By Marri Brackman, DO

Congress is debating whether to end a Medicare experiment before CMS has enough evidence to determine whether it works.

The Senate has rejected one effort to terminate the Wasteful and Inappropriate Service Reduction (WISeR) Model, while House lawmakers have advanced a separate effort to block its funding. Critics warn that the model will allow artificial intelligence to deny care and import the worst features of Medicare Advantage into traditional Medicare.

Those concerns deserve scrutiny. But the debate has focused heavily on reported problems in Washington state, even though WISeR is being implemented by different contractors across six states.

At ilumed, we work with physicians and health systems in two of those states where WISeR is being tested—Texas and New Jersey—as well as in states outside the model, predominantly Florida. Because we help practices manage the quality and total cost of care for Medicare beneficiaries, we can see how utilization and spending patterns are changing. Our experience has not aligned with the prevailing public narrative.

We are not hearing complaints from doctors or patients that WISeR is delaying or denying needed care, as has been reported in multiple news outlets. In fact, our internal data suggest the model is working as intended: applying existing coverage rules before payment is made, encouraging evidence-based care, and reducing inappropriate services.

WISeR introduces prior authorization into traditional Medicare, testing the approval process for a mere 17 highly abused medical services in six states. These include certain spinal procedures, skin substitutes, nerve stimulation devices, and other services for which Medicare already has coverage criteria.

The model is testing whether AI can make the review of existing Medicare coverage rules faster and more consistent. But technology does not have the final word when a request is rejected. Every denial must be reviewed by a human clinician with relevant expertise before it is issued.

In 2026, per-member spending on WISeR-reviewed services in our practices, setting skin substitutes aside, has fallen 75% from the prior year among our Texas patients and 84% in New Jersey. In Florida, it rose a marginal 3%. In pure dollars, that is roughly $820 less per beneficiary per year in Texas and $480 less in New Jersey.

These figures show a meaningful divergence between our practices operating under WISeR and those outside the model. At a minimum, that pattern deserves further evaluation before policymakers reach a verdict on the entire six-state test.

The debate over WISeR has become political. A Senate Finance Committee report leveraged data from Washington hospitals to claim that WISeR was delaying care by up to four times longer. But that data cannot be independently verified. Medicare’s claims processing contractors are barred from publicly speaking on their work on broadly and cannot refute false or misleading claims.

WISeR merely ensures coverage determinations are followed before it pays for certain services, which is simple and common sense. These coverage determinations have been set by clinical experts in that field, and AI is used to ensure the request matches the pre-set determination.

Traditional Medicare has many strengths, but its default approach has too often been to pay claims first and try to recover improper payments later. The Government Accountability Office estimates that improper payments in traditional Medicare were $32 billion in 2024. For perspective, the NIH’s budget that year was $47 billion.

That is where a prior authorization can play a role. Prior authorization is a necessary tool, and problems only arise when it is overused. In a fee-for-service model, on which our health system is built, fraud and overuse drive unnecessary spending. Patients are forced to pay more in cost-sharing and subjected to unnecessary care for the sake of providers making more money.

Medicare Advantage plans broadly use prior authorization, and coverage rules are opaque. This creates an environment that’s too difficult for patients and clinicians to navigate. But WISeR is not importing Medicare Advantage’s prior-authorization practices into traditional Medicare. It is limited to a small set of services with documented vulnerabilities to waste or inappropriate use.

It’s important to note that WISeR is a new model in its first six months and won’t work perfectly right out of the gate. The point of CMS Innovation Center models is to test different payment models at a smaller scale and then expand over time if successful. WISeR is testing prior authorization in six states in a small number of highly abused services. It’s slated to last six years.

No model is perfect in its first months. Accountable care models, bundled payments, and primary care models have all required course corrections. The right response to early implementation concerns is oversight and correction, not cancellation.

Additionally, WISeR is being tested with different contractors in its six test states. A problem in one state may reflect a contractor issue or a localized problem. It may not be fair to judge the experience in Washington state with those in Texas, for example. Washington has been singled out by CMS and ordered to submit a corrective action plan.

The skepticism surrounding WISeR is understandable. Physicians and patients have experienced prior authorization processes that are burdensome, opaque, and disconnected from clinical judgment. WISeR should not receive a pass simply because its stated goals are worthwhile.

Implementation concerns should be addressed through active oversight, corrective action, and continued evaluation. That is how Innovation Center demonstrations are supposed to work: test an approach, identify problems, make adjustments, and determine whether the model produces better results over time.

The alternative to WISeR is a fee-for-service system that too often pays for questionable care, identifies abuse only after damage is done, and leaves patients to absorb the consequences.

WISeR is still in its first year. From what we are seeing in Texas and New Jersey, the model is beginning to do what CMS designed it to do, which is apply existing standards before payment, discourage questionable services, and protect patients and taxpayers from inappropriate spending.

Congress should continue to oversee the model, but it should not terminate it before the evidence is in. CMS should be allowed to correct implementation problems, evaluate results across all six states, and finish the test it began.