What the AHA Accidentally Told Every Self Funded Employer

On August 28, the American Hospital Association (AHA) filed a 56-page comment letter with CMS about the 2027 outpatient payment rule. Most of it is what you would expect. The payment update is too small. The productivity adjustment is unfair. Please stop cutting 340B. So on and so on….

Then on page five, in a paragraph about administrative costs, the AHA drops this:

“In 2025, hospitals spent nearly $18 billion on overturning claims denials alone. All told, using data from the most recent annual survey, the AHA estimates that hospitals spent a staggering $43 billion in 2025 trying to collect payments insurers owed for care already delivered.”

But the data they reference actually isn’t all from the most recent annual survey.

They go on to note that the average hospital employs about 64 administrative and billing staff dedicated just to this work, which is roughly 6.5% of total hospital employment. And that a federal watchdog study found 75% of care denials were eventually overturned.

This is the hospital industry’s own trade association, in a formal federal filing, putting a dollar figure on the fight between hospitals and insurers over money that was already owed.

Why a plan sponsor should care

Forty three billion dollars of labor, software, and phone calls that produced exactly zero medical care. Think about that.

That money doesn’t just evaporate, it’s real. Hospitals don’t eat it out of civic pride. It shows up in their cost structure, the cost structure informs the chargemaster, the chargemaster anchors the negotiated rate, and the negotiated rate is what shows up in your claims! Employers are paying for both sides of an argument they were never invited to, never a part of.

Sixty four people per hospital, on average…..Nobody staffs 64 people to solve a problem that doesn’t exist. This is the main reason it costs alost thirty cents for a hospital to collect a dollar. That translates into higher costs for self-funded employers.

Now the part your LinkedIn feed will skip

I actually read the letter’s footnotes. You should too, because this number is much softer than the headline suggests, and I would rather you hear that from me than from the carrier rep who looks it up on the way to your renewal.

The $18 billion is not a 2025 number. The AHA cites a Premier report for it. That report was published in February 2025, but the underlying survey covers 2023 claims adjudication, and it covers hospitals, health systems, and post acute providers, not hospitals alone. Premier reached $25.7 billion by taking a self reported 15% denial rate from 280 member hospitals and multiplying it against three billion annual claims from a 2020 industry statistic. Then it estimated that roughly $18 billion of that was potentially wasted. That is a survey result stacked on a five year old claim count, relabeled two years forward.

The Premier report says 70%, not 75%. The AHA swapped in a higher overturn figure from a completely different source.

And that source is the wrong one. The letter footnotes a 2023 HHS Inspector General report for the 75% figure. That report is about Medicaid managed care prior authorization, and its finding was that plans denied one in eight requests. The actual home of the 75% number is a 2018 Inspector General report on Medicare Advantage, using 2014 through 2016 data.

The 75% has a denominator problem. Medicare Advantage plans overturned 75% of their own denials on appeal. But only about 1% of denials were ever appealed. So the honest sentence is this: of the tiny sliver of denials that anyone bothered to fight, three out of four got reversed. That’s actually worse than the headline, because it means the other 99% either got paid by a patient, or turned into care nobody received. It is also not the same claim.

The $43 billion itself has no footnote at all. It is the AHA’s own estimate off its own annual survey. Which is fine, I guess, but where is it backed up? It is also the number everyone is going to quote.

Does that kill the argument

No….I think it sharpens it actually.

The AHA is a trade association asking the federal government for a bigger payment update. Of course they picked the biggest defensible number and rounded in their favor. That is the job. Of course they cherry-picked data from different years and different studies and different payor profiles.

The insurers/carriers will now respond with their own study showing that denials are a necessary defense against upcoding and unnecessary utilization, and they will not be entirely wrong either. What neither side will publish is the missing number: how much hospitals spend generating claims that get denied for good reason. That figure exists, but nobody wants to fund that research. Why? Because both parties benefit as costs go up. Hospitals don’t want to have the discussion about what amount of claims SHOULD be denied, and neither do insurance carriers. Which begs this question: How many IMPROPER claims actually get paid? Do that study.

In fairness, I’m not a neutral party here either. We developed and sell an unbundled self funded plan built on direct contracting. Take the appropriate discount on my enthusiasm.

But here is what survives all of the caveats. Both sides of this fight agree the adjudication/beurocratic layer is enormous, growing, and produces absolutely nothing. They only disagree about who should pay for it. The answer to that question, in every scenario, is you…..Mr. Employer. You pay for it.

What to actually do with this

Stop ignoring the denial and appeal cycle in your medical trend. It is a line item wearing a disguise that very few think they can do anything about. Why? Because assuming an employer can even get this information from the TPA/Carrier, most employers will sit around staring at a denial rate with no benchmark and no idea what “good” looks like.

Ask your TPA or carrier for your own numbers. Denial rate. Appeal rate. Overturn rate. Average days from submission to payment. If they can produce it, you have a baseline to negotiate against. If they cannot produce it, well, you just learned something equally valuable. Ask what your UM or prior auth vendor costs PMPM, and what percentage of its denials get overturned. If a vendor is charging your plan to generate determinations that reverse most of the time anyway, that ROI is auditable and it’s your money. Same question for any network or “repricing” arrangement, including out of network repricing. Anything paid on a “percentage of savings” for the most part. Shared savings compensation on deials and repricing is a real conflict, it sits in the employer’s fee stack, and unlike the AHA’s $43B claim, it shows up on an invoice you’re already getting.

Medical necessity and prior auth denials touch your money, and the direction is counterintuitive. When a denial gets overturned, your plan pays anyway. You bought the UM review, you bought the appeal process, and then you bought the claim. You likely bought an angry employee too.

Then ask the harder question. A traditional network arrangement negotiates a discount off a price that already has all this friction baked into it. A direct contract with a facility, at a fixed price, doesn’t remove the discount. It removes the argument. A direct contract sets the price before a single person walks through the door, and there’s nothing to dispute.

43 billion dollars a year says it’s the argument that is the expensive part.