A Huge Number of Employers Are Clueless About Where Their PBM Rebates Go

If you want to search Cost Curve back issues or link to anything you read here, the web links and archive are online at costcurve.beehiiv.com. You can subscribe there, too.

There is a new CBO report that estimates what it would cost Medicare to fully cover obesity medicines. (There hasn’t been a ton of attention -- Bloomberg had the story yesterday, and Politico has coverage today -- which is a little weird.) 

The topline here is that it would cost Medicare around $35 billion over 10 years to cover the meds, which includes a small contribution from cost savings elsewhere in the health care system. 

Three other quick takeaways on the data: 

• The analysis assumes that, by 2034, about 14% of eligible Medicare beneficiaries will be on one of the drugs. That’s about 1.6 million people. 

• The pricing calculus is interesting. If the policy went into effect in 2026, CBO estimates that the medicines would run about $500 a month. In 2027, after price controls for Wegovy kick in, the price would drop to $333 a month.

• Annual savings from treatment ends up being about $650 per patient by 2034. This is the place where I’m most curious to dig into the methodology … CBO estimates the per-patient impact as $50 in 2026, growing each year. 

My gut reaction here is that $35 billion over 10 years … should be doable for the richest nation in the history of humanity? That’s not a statement of optimism about the politics here. We’re in an environment where any increase in spending is going to be treated with great suspicion. 

But in the grand scheme of things, the question is whether we can squeeze out a few billion in waste and abuse and put it toward a revolutionary, cost-effective intervention. I keep coming back to this WSJ analysis that suggests that Medicare Advantage overdiagnosis cost the system $50 billion over three years. 

Feels like “minimize MA abuse, get universal coverage of obesity meds” would be a good tradeoff. I know that the health care system doesn’t work like that, but please permit me my daydreams.

There are a very small number of reports each year for which you should drop everything and just luxuriate in the data. The IQVIA “Use of Medicines” publication is one. The NHE data drop is another. 

The KFF Employer Health Benefits Survey -- released today -- is a part of that august group. It is the defining assessment of what is going on in the commercial market, and it is bleak. (In fairness: it’s almost always bleak.)

Premium costs are going up, and the rate of that growth is increasing. Covering a family now runs $25,572 a year, up 7% from 2023. The good news (maybe) is that the worker portion of the premium fell last year, but the flip side of that is that businesses are picking up more of the ever-growing tab. Not good news for employers.

There are a lot of nuggets in the report. CNN noticed the GLP-1 discussion. USA Today offered some thoughts about health care inflation versus general inflation. The Wall Street Journal turned the data in the report into some easy-to-play with-charts. 

But my attention went to one particular analysis. The survey asked employers how much of the rebate dollar they received from their PBM. 

There is an argument that the answer here should have been easy: PBMs have been shouting from the rooftops that they’re passing through nearly all of the rebates to the employers. 

The reality, it turns out, is a lot more complicated. Fully 37% of those surveyed said they didn’t know. Only 19% said they were receiving “most” of the rebates. 

Those are stunning numbers. 

The data suggests a handful of overlapping explanations. It’s possible that employers are simply ignorant. This is how KFF Health News positioned it, including this memorable headline: “Employers Haven’t a Clue How Their Drug Benefits Are Managed”. It’s possible that contract arrangements make a specific number hard to come by. (This argument was raised in a LinkedIn post discussing the finding.) Or perhaps there is a quirk in the KFF data-gathering process that makes answers here less reliable. 

Regardless of what explanation(s) you favor, there is an education/transparency issue. Mark Cuban has been dragging employers up and down the court on this for a while, but he clearly needs some support from the bench here.

Kamala Harris wants to add long-term care to Medicare, and she is proposing extended price controls as the pay-for.

I’m going to invoke my usual stance of not overly engaging on plainly unserious policy proposals. It’s hard to get the numbers to work, given that Harris’ campaign suggested we’re looking at $40 billion a year in additional spending. (Details from Harris are here.) And at some point, patience for the idea that we’re going to get lower drug prices for the government -- but not pass those savings along to the patients themselves -- is going to wear thin. 

All that said, the math underlying all of this is scary, and I’ll get into that later this week. 

n.b. Harris’ plan includes a line about sticking it to the PBMs, too. Not much detail there, but STAT is crediting Mark Cuban with influencing the campaign’s thinking. So it’s not all bad!

ELSEWHERE: 

• I do not think Cost Curve readers will gain anything from reading this long NYT piece about the drug-pricing policies of Trump and Harris, but it’s probably important for you all to know that there is a long NYT piece about the drug-pricing policies of Trump and Harris.

• A couple of PBMs have filed a legal brief trying to kick FTC Chair Lina Khan and two others from the FTC’s PBM lawsuit because of bias, per Reuters. No idea how that works or what effect it will have. The whole thing feels kind of performative on all sides but, ngl, it feels like anything that keeps this issue in the public eye benefits the pro-reform movement. 

• Axios has the scoop on the Biden administration’s effort to deliver certain generic drugs to Medicare patients for a $2 copay via a pilot program.

Thanks for reading this far. I’m always flattered when folks share all or part of Cost Curve. All I ask is for a mention or tag. Bonus points if you can direct someone to the subscription page.