Medicare has launched a six-year pilot program that could eventually transform access to health care for some of the across the U.S. who rely on it for their health insurance coverage.
is a government-administered insurance plan for people over 65 or with disabilities. About half of the 67 million Americans insured through Medicare have this coverage. The rest have administered by private companies.
The pilot program, dubbed the , is an experimental program that began to affect people enrolled in traditional Medicare from six states in January 2026.
During this pilot, medical providers must apply for permission, or , before giving 14 kinds of health procedures and devices. The program software to identify treatment requests it deems unnecessary or harmful and denies them. This is similar to the way .
As and the , we believe this pilot could save Medicare money, but it should be closely monitored to ensure that it does not harm the health of patients enrolled in the traditional Medicare program.
Prior authorization
The pilot marks a dramatic change.
Unlike other types of health insurance, including Medicare Advantage, traditional Medicare health care providers to submit requests for Medicare to authorize the treatments they recommend to patients.
Requiring for these procedures and devices could reduce wasteful spending and help patients by steering them away from unnecessary treatments. However, there is a risk that it could also delay or interfere with some necessary care and add to the paperwork providers must contend with.
Prior authorization is by Medicare Advantage plans. Many insurance companies hire technology firms to make prior authorization decisions for their Medicare Advantage plans.
that Medicare improves its services. Medicare tests changes on a small number of people or providers to see whether they should be implemented more broadly.
The six states participating are Arizona, New Jersey, Ohio, Oklahoma, Texas and Washington. The that require prior authorization during this pilot include and incontinence-control devices. The pilot ends December 2031.
If the , which administers Medicare, deems the pilot successful, the Department of Health and Human services to include and .
Introducing a hurdle
This pilot isn’t changing the rules for what traditional Medicare covers. Instead, it adds an extra hurdle for medical providers before they can administer, for example, .
If Medicare issues a denial rather than authorizing the service, the patient goes without that treatment unless their provider files an appeal and prevails.
Medicare has to do the work of denying or approving prior authorization requests, with the .
Many of these are the that do prior authorizations for .
The of what Medicare would have spent on the denied treatments. This means companies are paid more when they deny more prior authorization requests.
Medicare for inappropriate denials.
What to watch for
Past research has shown that when insurers require prior authorization, . This pilot is likely to reduce treatments and Medicare spending, though how much remains unknown.
The Centers for Medicare & Medicaid Services by the pilot because there is evidence they are in many cases.
If the program denies cases where a health service is inappropriate, or of “” for a patient’s health, people enrolled in traditional Medicare could benefit.
But for each treatment targeted by the pilot, there are some cases where that kind of health care is necessary.
If the program’s AI-based decision method has trouble identifying these necessary cases and denies them, people could lose access to care they need.
The pilot also that medical providers must do. Paperwork is already a major burden for providers and .
AI’s role
No matter how the government evaluates prior authorizations, we think this pilot is .
The impact of using AI to evaluate these prior authorizations is unclear. AI could allow tech companies to automatically approve more cases, which could speed up decisions. However, companies could use time saved by AI to put more effort into having people review cases flagged by AI, which could increase denials.
Many for Medicare Advantage prior authorization decisions, although there has been limited research on these models, and little is known about how accurate AI is for this purpose.
What evidence there is suggests that AI-aided prior authorization and than when insurers make prior authorization decisions without using AI.
The bottom line
Any money the government saves during the pilot will depend on whether and how frequently these treatments are used inappropriately and how aggressively tech companies deny care.
In our view, this pilot will likely create winners and losers. Tech companies may benefit financially, though how much will depend on how big the treatment reductions are. But medical providers will have more paperwork to deal with and will get paid less if some of their Medicare requests are denied.
The impact on patients will depend on how well tech companies identify care that probably would be unnecessary and avoid denying care that is essential.
Taxpayers, who pay into Medicare during their working years, stand to benefit if the pilot can cut long-term Medicare costs, an important goal given Medicare’s .
Like in Medicare Advantage, savings from prior authorization requirements in this pilot are split with private companies. Unlike in Medicare Advantage, however, this split is based on a so that payments to private companies cannot exceed total savings, and the benefits of the program are easier for Medicare to quantify.
In our view, given the potential trade-offs, Medicare will need to evaluate the results of this pilot carefully before expanding it to more states – especially if it also expands the program to include services where unnecessary care is less common.
Original article
Comments