AI will say "No" in Medicare

[WendyBG]

https://www.nytimes.com/2026/01/17/health/nursing-homes-health-aides-medicare.html

3 Policy Moves Likely to Change Health Care for Older People

Two regulatory rollbacks, along with a new A.I. experiment in Medicare, raise some worrisome questions.
By Paula Span, The New York Times, Jan. 17, 2026

…

Medicare’s A.I. Referee

Beyond rolling back policies for care of the aged, the Trump administration has established a pilot program to introduce one to traditional Medicare: prior authorization, using artificial intelligence and machine learning technologies.

Touting it as a boon to taxpayers, Medicare calls it WISeR — “Wasteful and Inappropriate Service Reduction.”

Prior authorization, in which private insurers review proposed treatments before agreeing to pay for them, is widely used in Medicare Advantage plans despite its unpopularity with patients, doctors and health care organizations. It has rarely been used in traditional Medicare.

This month, however, WISeR debuts in six states (Arizona, New Jersey, Ohio, Oklahoma, Texas, Washington) in a six-year trial to determine whether review by tech companies can reduce costs and improve efficiency, while maintaining or improving quality of care.

Initially, WISeR targets 17 items and services that C.M.S. said “historically have had a higher risk of waste, fraud and abuse.” The list includes knee arthroscopy for arthritis, electrical nerve stimulation devices for several conditions, and treatment for impotence… [end quote]

Of course denial will reduce costs! The whole point of paying for Medicare instead of Medicare Advantage is getting health care when the doctor recommends it without the denial process.

I’m currently reading a book called “The Great Nerve,” about the vagus nerve which controls many physical processes including inflammation. The author helped invent a vagus nerve stimulator which is similar to a heart pacemaker. This could be one of the electrical nerve stimulation devices that will be denied. The vagus nerve stimulator directly competes with several high-profit drugs.

Wendy

[Leap1]

Not to worry, he may default on the bonds. Then he can cut Medicare entirely to save on taxes. As we go bankrupt as a nation, he can sell off the assets on the cheap.

We’ll all be ok. Particularly the ric…well by golly…won’t they be bankrupt?

Excellence in governance.

Only stupid people don’t make money in bankruptcy.

[MataroPete]

@WendyBG

This month, however, WISeR debuts…in a six-year trial to determine whether review by tech companies can reduce costs and improve efficiency, while maintaining or improving quality of care.

Yes, we’ll see about the “while maintaining or improving quality of care”. One suspects that part may go by the wayside.

Pete

[pauleckler]

I don’t think we can expect Medicare to pay for every treatment someone can dream up. Coming up with coverage for treatments offering worthwhile benefit makes sense. And AI can probably do a better job deciding which treatments deliver benefits.

Regardless of sales pitch, giving preference to treatments with a history of success makes sense to me.

[Goofyhoofy]

@pauleckler

I don’t think we can expect Medicare to pay for every treatment someone can dream up.

This is likely true.

@pauleckler

Regardless of sales pitch, giving preference to treatments with a history of success makes sense to me.

How do you get a history of success without a history? Or is that only available to people with money?

[DrBob2]

@Goofyhoofy

How do you get a history of success without a history?

There are long histories of the various medical procedures as well as data on fraud. Unless you think there hasn’t been any fraud, you target those with the highest rates.

DB2

[WendyBG]

@pauleckler

Regardless of sales pitch, giving preference to treatments with a history of success makes sense to me.

I agree with this. There is a book, “Ending Medical Reversal,” which is about how evidence-based medicine (large, multi-centric, double-blinded random studies) proved the uselessness of hundreds of drugs, surgeries, procedures and medical devices.

Ending Medical Reversal: Improving Outcomes, Saving Lives (Johns Hopkins Press...

The FDA requires approval of drugs and medical devices. FDA also has a post-market adverse events system to discover whether use in a wide population over a long time will cause adverse effects.

https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/overview-device-regulation

My concern is that AI might cancel the use of FDA-approved drugs and devices that have already been approved.

Wendy

[Goofyhoofy]

@WendyBG

My concern is that AI might cancel the use of FDA-approved drugs and devices that have already been approved.

My concern is that AI won’t approve a procedure that medical people (you know, humans) might think necessary because it doesn’t have a history of being approved. How does it get a history of being approved? By being approved, of course.

Clearly there have to be some guardrails to stop any quack doctor from doing things which are dangerous, or merely charging for placebo treatments, but then there are already guard rails in place for that aren’t there? Should they bge strengthened? OK. Relying on a technology which can only look backward doesn’t seem to be the best use, although I’ll say that AI might be able to root out fraud more quickly, so I’ll give it that.

[Leap1]

@WendyBG

@pauleckler

Regardless of sales pitch, giving preference to treatments with a history of success makes sense to me.

I agree with this

One of the major tenets of the ACA.

As we all get older, if you don’t remember something that does not mean it is new.

[intercst]

@WendyBG

My concern is that AI might cancel the use of FDA-approved drugs and devices that have already been approved.

I’d be more worried about a for-profit insurer denying an FDA-approved drug in the pursuit of continually expanding excessive Executive Compensation.

intercst