LIVE: Jetson Leder-Luis to Present on Fighting Healthcare Fraud

Live News Network
1.4K followers
53 Video Views·Sep 1, 2026
AI Generated·May contain errors
Video Summary
AI GeneratedThis presentation analyzes the effectiveness of civil litigation and whistleblower incentives in combating healthcare fraud. It argues that the primary value of legal action is \"specific deterrence\"—preventing future fraudulent spending—which often far exceeds the value of recovered damages. The speaker emphasizes combining litigation with strategic program design to eliminate fraud incentives.
Whatever weird system, nothing else works like this, right?
This is a private market for antifraud, right?
You can become a whistleblower and get paid for it.
We have created both a private incentive, that is, we pay people for the information,
but also we've created a private market in the sense that good cases are won and the
person gets paid, bad cases are lost and the person gets paid.
When I think about civil litigation, it's very important to think about where the value
is, and this graph goes through that.
So, on the x-axis here, we have time, and on the y-axis, we have dollars spent, and
the horizontal line is how much money would we be spending if there hadn't been fraud
in this space?
And if you want, you can make it slightly upwards, but the point being, normalize that
to, you know, roughly top, that's our baseline, okay?
Spending on fraud is higher than that, right?
So, spending on fraud is our upward sloping line, and at some point, the whistleblower
catches this and the whistleblower sues, that's my vertical line in the middle, and then spending
comes back down.
Well, maybe, I'll show you whether or not it does, but it does, in practice, often come
back down.
The whistleblower's value is not the amount of money they get back from the government.
That is the blue triangle labeled damages, right?
The government says, oh, you stole $10 million from us, maybe double that, $20 million, that's
proportional to the historic billing, right?
That's backward looking.
Whistleblowers are paid for the fraud that they caught that already happened.
However, if the spending went down, the government averted future spending, right?
And that is represented by the lavender trapezoid labeled specific deterrence.
So we don't get to see how much would have been spent in the counterfactual world in
which we had not caught the fraud, but that is the upward sloping line that I say will
probably not upward slope forever, maybe it eventually tapers off, but for the whistleblower,
spending would have potentially continued to grow, probably eventually tapered off,
and the integral between what the government would have spent and what the government did
spend, which I've labeled here as the specific deterrence, that's how much money the government
is actually saving, right?
That's dollars not spent.
And so in this paper, I estimate that dollar figure.
I say how much money did the government not spend because whistleblowers caught that behavior?
And so I pair enforcement data on large whistleblower cases where I had a Freedom of Information
Act request against the Department of Justice as a lowly graduate student with Medicare
claims data graciously provided by the NBER, and I said, let's look at some big categories
where we've had whistleblower lawsuits, and then see how much money we spent or didn't
spend on those categories of care.
And I go through four case studies of the biggest settlements that I'm able to observe
in my data, which are the outlier payments fraud, the unnecessary inpatient kyphoplasty
fraud, off-label promotion of Botox for migraines and headaches when it was only approved for
torticollis and hyperhidrosis, and then unnecessary inpatient admissions.
In every case, I'm going to say how much money did the government save, and I really actually
want it as a ratio, which is if the government got back $100 million, did we save $100 million
on top of that, or a billion dollars?
What's that ratio?
And that's really the goal of this paper.
To do that, I use synthetic controls, which is an advanced econometric tool that referees
do not like, but I fought with them, and I got the paper published.
But synthetic controls are tricky, right?
You have to make some assumptions about how other things that were not treated would have
evolved in the absence of the treatment.
But I think I did so in a relatively convincing way.
So I want to give you a real case study, right?
What does this actually mean in terms of medicine?
Let's talk about kyphoplasty.
So kyphoplasty is a spine surgery for patients with vertebral compression fractures.
So many patients, often older women with osteoporosis, experience a compression fracture, which is
one of their spine discs gets shortened.
It's very painful.
And the surgery, kyphoplasty, involves the injection of a balloon into the affected vertebra,
and then they inject cement into that, and it chores up that vertebra.
But this can be done safely outpatient, pretty notoriously very safe outpatient surgery.
However, outpatient doesn't pay very well.
An outpatient back surgery like this pays like around $1,000.
The maker, allegedly the maker of the spine kit, had promoted that patients should try
to bill, excuse me, that hospitals should try to bill for a one-night inpatient stay
under I believe it was like spine surgeries not otherwise listed, because then you could
receive inpatient coverage for this.
And dozens of hospitals engaged in unnecessary inpatient kyphoplasty when instead they should
have been performing these procedures outpatient.
There was a lawsuit in 2005, then there were a series of lawsuits against dozens of hospitals
and the maker of the spine kit.
There was a settlement and the data speaks for itself.
This is a graph of one-night stays under the DRG that was allegedly promoted by the company.
We see between 1999 and 2005, the number of one-night stays rose from like less than a
hundred per month to about 1,500 per month.
The lawsuit happened, and then the first lawsuit settles in May of 08, and we see this massive
reduction back to baseline.
And basically what this paper does is this, for those who are online, I'm drawing a little
line.
I say, if we hadn't caught it, if it hadn't got back down, how much would we have spent
on it?
Maybe this much, maybe this much, somewhere in there, right?
You can bound it.
So how much money did we save?
The difference, right?
We saved the difference between this, how much ever we would have paid for that.
So I do this in these four case studies, which collectively had received $1.9 billion of
settlement, that is money that was returned to the government.
But I estimate that the government actually saved more like $19 billion, when you look
at just the future spending discounted over five years, over how much I estimate we would
have spent, but for these whistleblowers.
So the deterrence value here is about 10x the settlement value, although there's massive
heterogeneity between the cases.
That's the other thing I have to say.
Some cases are really big, and some cases I actually got a zero, and I have some theories
as well.
The point being, civil litigation works super well here, right?
I'm not saying that civil litigation doesn't work.
I just showed you civil litigation, not only does it get back billions of dollars, it saves
additional tens of billions of dollars.
Large firms, like hospitals, are not judgment-proof, and there are not that many of them.
And so when a couple dozen hospitals engage in a fraud, we can sue them, and they can't
close.
Hospitals do not close very easily to hide from a million-dollar or a $10 million judgment.
Hospitals have extensive physical capital, they have established presences, and they're
not going to close when we go after them for $10 million, they pay us.
And because they know that, they stop, right?
Hospitals, because they're not judgment-proof, understand that if they get caught for this,
they have to pay.
They can't just float into the wind like an ambulance comes in.
And because of that, they're very sensitive to perceived litigation and the threat of
litigation, and so when a hospital gets caught doing this, they stop doing it, which produces
massive deterrence to future value, and other companies stop doing it, right, which produces
this big spillover.
And so my graph was not one-night stays for kypoplasty among the hospitals that got sued,
it was among all hospitals, right?
And so we can produce massive deterrence effects using this civil litigation when it's targeted
against firms who are not judgment-proof.
What about patient health here?
I have a follow-up paper, a fantastic co-author, David Howard at Emory University and I wrote
a paper about unnecessary hospital admissions.
These are not kypoplasty-specific, these are any admission out of the ER.
So dozens of hospitals have faced lawsuits by the Department of Justice for pressuring
physicians to admit from the emergency room because inpatient care is substantially more
expensive, they want to fill the bank.
We examined eight hospital systems, collectively 205 hospitals, that settled to a total of
almost $500 million with the Department of Justice for unnecessary inpatient admissions
from the emergency room.
We're going to do a difference-in-difference design, look at hospitals that have high admission
rates against hospitals that have high admission rates and got in trouble, to estimate what
happens both to the hospital spending and to the patients.
We look at 238 million ER visits, which is the 100% sample of Medicare claims, from 2005
to 2019.
Litigation works against hospitals, right, as we just showed, but here's another example.
90-day Medicare spending fell by about 4% per patient, admission rates fell by about
4% per patient, 30-day mortality fell slightly, that's what we want to be really worried about,
not letting people into the hospital could potentially kill them.
We see that the sign goes the other direction, although it's a little bit small.
There's a slight uptick in readmissions, but that's mechanical, right?
A patient who gets admitted, has like a week in the hospital, they can't be admitted during
that time, and so when you send them home, mechanically there's a greater risk that they
ever get admitted.
Excuse me, these are revisits, not readmissions.
So some of the patients do have to come back, but way less than offsetting, right?
So some patients do not get admitted, some of them will return, but we're able to find
a five-year discounted savings of $1.3 billion here, but importantly, no harmful patient
health effects.
And so much like in the regulatory example, well-targeted anti-fraud efforts can potentially
improve patient health, or at the very least, have no negative consequences.
So the tool must meet the fraud, right?
That's really the point of this talk.
There are a lot of things we could be doing.
We could have tried to use regulatory enforcement on hospitals, that's hard, right?
Do you want to exclude a hospital from the entire Medicare program?
I could imagine lots of concerns with that.
Litigation is actually very valuable in the hospital space, in a way that it is completely
not valuable in the ambulance space.
And so the right tool for the right fraud, I believe, I missed this slide, right?
This is the one I was looking for.
I thought it was the hat.
It's behind, great.
Okay, so what does this look like in practice?
Well, we are currently doing this.
The Department of Justice, which is responsible for civil litigation on behalf of the United
States government, has recently announced a set of large settlements against M.A. plans.
Let's talk about M.A. plans.
Medicare Advantage is a form of private insurance provided where Medicare pays a capitation
rate, basically pays for your insurance, as provided by a third-party company.
And we do so based on what's called risk adjustment, which is to say how sick, on average, the
patient is.
So a patient who has diabetes, it costs us more to insure them than a patient without
diabetes, because diabetes is, on average, more expensive.
So we don't pay for the claims, we pay capitated rates, so we pay for the whole insurance bundle
as provided by a third party.
However, this creates weird incentives for M.A. plans to up-code, that is, to list diagnoses
that a patient might not necessarily have, because if they can get the patient coded
as having diabetes, well, they're going to get more money for the patient.
There has been extensive civil litigation by the Department of Justice, and if you look
at the timing, one of these was last week, two of them were in the last month, hundreds
of millions of dollars of settlement against the M.A. plans, and in some cases against
provider groups that undertake risk-sharing with the M.A. plans, and have therefore become
responsible for the amount of diagnosis.
And these M.A. plans and some of their provider partners have allegedly engaged in manipulation
of this system, but we get these settlements.
These are real dollar values that I believe have probably already been repaid, given the
timeline.
Many of these have like a 30-day, you must pay within 30 days, and we don't allow them
to sign the agreement until we get validation that they have the money.
And so this is real money that is being returned to the U.S. government, and not only returned,
but I believe that this will cause a reduction in this behavior, because, as we've seen,
civil litigation is effective against large firms with assets, and insurance companies
are large firms with assets.
So I've talked a lot today about how to enforce effectively against fraud, and synthesize
from some studies.
I want to highlight that this is the tip of the iceberg for the work that I think could
be done in this space, because there is a bigger and more important set of questions,
which is how do we design our programs to have less susceptibility to fraud?
And we try to do some of this in practice as well.
You guys might have heard about skin substitutes.
Skin substitutes are basically a fancy band-aid, that's a simplification, but it's a medical
product that treats wounds that we were paying too much for, because we were paying a high
dollar value per square centimeter.
And if you've seen Dr. Oz carrying around a poster board sometime in the last few months,
which he does everywhere he goes, he's shown that there's been a massive reduction in spending
on skin substitutes.
And that massive reduction was accomplished by changing the price of skin substitutes.
We changed the reimbursement method for skin substitutes, and that cut out extensive spending.
So one of the questions that I want to now go forward and ask, and is tricky, is how
do the systems that we set up to administer the health program, the way that the prices
are determined, which firms participate, the industrial organization of healthcare, how
does that contribute to fraud?
And are there ways that, rather than saying, okay, there's all this fraud, should we use
regulation or litigation to clean it up, how can we attack fraud on the front end through
changes in program design?
And so I don't have time to go through it today, but I have a great paper with Ryan
McDevitt, Renika Dewan, Jimmy Roberts, Paul Ellison, and Riley League on auctions in durable
medical equipment.
There have been extensive use of auctions.
We find a surprising result that they, at least 10 years ago in the first round, inadvertently
made fraud worse, which is not what we were expecting.
And so all I'll say is that the study of the design of programs is going to be the next
frontier in eliminating fraud in public programs, both in healthcare and beyond.
This is my academic email and my White House email and my website, if you want to read
some of these papers.
Thank you so much for taking the time to come.
I'm an open book, so if you guys have any thoughts or want to reach out, I'll grab coffee
with anyone, and we'll do some Q&A.
I'm Sandra Decker from HR2, we're working.
So I'm an economist, but my question more comes from being a former D.C. foster parent,
where sometimes Medicaid or CFSA would offer transportation for services.
And I would never accept that if I didn't need it, because the companies they contracted
with were late, they were hard to schedule with, there was a major pain in the butt.
So I understand the supplier incentive to provide ambulance taxis for $500 down trip.
I'm a little less understanding of what the patient motivation is if they don't really
need it.
I guess if they're looking for attention or something, but to me, I guess they save the
gas money, I don't know.
I can look, and I have a second part, one or the other, did you look at Medicaid claims
too, like whether, so about half of ESRB patients I think are duals.
And I could look in the JPE paper and see, but some patients might have moved from Medicare
payment of ambulance services to Medicaid, and that would be hard to look at, right,
because you would have only had max data instead of CFSAs, and probably you couldn't see it
if that had happened.
Great question.
So thank you.
When a mafia organization recruits patients to ride in the ambulance at $500 per ride,
and the patient doesn't want to ride, they pay them $20.
And so the patients in many cases were receiving kickbacks.
We know that.
It's still profitable at $480 for the ambulance, right?
The price was so high relative to the cost of providing the service that there was ample
room to compensate the patient.
Four patients ultimately went to jail for this.
Although that's low, it's very difficult to put a dialysis patient in federal prison.
As you can imagine, they have extensive medical need.
The FBI had extensive video footage of people walking out of the ambulances not on a stretcher
despite claiming bedridden and then receiving an envelope of cash.
So we have pretty good evidence for that.
Regarding Medicaid claims, as anyone here knows, basically you can't analyze Medicaid
claims before 2017.
There may have been some offset to non-emergency medical transportation or power transit or
Uber, which is not paid for at that time, but at this time would maybe be paid for by
some Medicaid plans.
We can't see it.
I don't think it biases the result.
I think the massive reduction in Medicare spending holds regardless, even if there is
maybe some offset.
But you can't rule out that a fair amount of it wasn't pushed from Medicare to Medicaid?
I think that Medicaid similarly has requirements.
It depends on the state of proving inability to otherwise travel.
Also, power transit would be substantially cheaper.
So if you're taking power transit to and from a dialysis facility, that's not $500 per round
trip.
And so the savings would go up.
Thanks so much for this.
I really appreciated you talking about the difference between large actors and smaller
actors.
A lot of the populations we think about are served in settings that are regulated by state
level.
Sorry.
Who are you with?
I'm Emily Rosenoff.
I'm with ACT-C.
Oh, great.
Hi.
How do you think about some of those state regulations and your levels in your current
rule?
So, for example, substance abuse disorder places or personal care homes or even assisted
living are all regulated by state level.
How should we think about that in your framework?
Yeah.
So regardless of who owns the regulatory lever, if we think that it is appropriate in the
sense that it both saves money and has the capacity for improving or at least not harming
patient health, we should ask the state to do it.
So through the Medicaid Broad War Room, we are now building working relationships with
states.
And so we ask the states, hey, you know, you seem to have a massive problem in this area.
The federal authority to put a moratorium on our Medicaid service is limited relative
to the state authority to put a Medicaid moratorium on it.
And so many states have come to us.
CMS tells us to approve the moratorium.
And so we work with the states.
Another example would be payment suspensions.
Payment suspension authority lives at the state level.
I mean, it varies by state, but largely lives at the state level.
The state Medicaid director can often put a payment suspension in place for a firm
when there's a credible allegation of fraud.
CMS will communicate and say, hey, here are firms where the data look for.
We don't have the regulatory authority.
But the states, in many cases, have the discretion to exercise.
And so I think federal-state partnership becomes increasingly important when we want
to take that sort of action.
Thank you so much.
First of all, a wonderful, dynamic program.
Jetson, a lot of compliments to you.
You know yourself.
Really appreciate you.
Dr. Raj, I'm a physician.
I oversee a large organization of 200 doctors based out of Florida, but all around the country.
So we've seen issues like this.
My question to you is this.
Inherently, we have a cultural issue with ethics and just downright honesty.
I don't know if that can ever be addressed.
You spoke about the federal and state relationship.
What I was going to ask is, is there a better opportunity?
Because I don't disagree with you about changing some of the policies and maybe legislation.
So moving forward, the scope of fraud may change for the up-and-coming generations.
But can we do any more policing through the state and or even through the private sector,
even like ACOs, perhaps, or any other organizations that could become partners or policing organizations
within this?
I'd like to hear your thoughts.
I can't see a way to make this a new policy.
How do you view this, because coming from Florida, as you know, South Florida is infamous.
My goodness, for being a burden on the country, not just the state for its issues.
If you could speak to those points.
Thank you.
It is an open question in research whether there are lower rates of overbilling among
private insurers.
I've seen great papers by people I trust in both directions.
And I'm not sure that that's a clear avenue for success.
Regardless, I think that because CMS, as a regulatory agency, has purview of oversight
of both the ACOs and the MA plans, which are very similar, that we are increasingly engaging
our partnership with them because, right, like, all hands on deck given the scale of
the national crisis.
I do think that as we discuss program design, there is this bigger question about are there
kind of contractual setups that promote the best version of, like, utilization management
and reductions in fraud.
And it's not clear to me that we're at that frontier, but that there may be opportunities.
I'm being told I can't take any more questions.
Are we out of time?
Okay, great.
And just to finish this off, the Chief Economist of HHS, Keith Miller.
Well, thanks, Seth, and that was informative and sobering.
Now I want to really underscore a lot of us from ASPE over here want to underscore the
important role of rigorous economic analysis plays in these efforts.
I think Seth was a little bit modest.
Dr. Owens gave you a hint that he was going to be modest, so I'm going to brag a little
bit more for him.
Milton Friedman talked about, he had the, and also Rose Friedman, they had the four
ways of spending money framework.
And they talked about, look, an agency like CMS is spending other people's money on other
people.
And so there's not a lot of personal incentive to police things like fraud.
It would be a little different for a private company whose profits are going to go in the
tank if they allow a lot to be stolen.
So how do you deal with that?
There's a big tailwind behind the fraud, let me put it that way.
That's not impossible.
Sunlight is really important.
So not only is your research helping show us how to better decide what scale of solution
do we need in what situation, but it's putting sunlight out there for the taxpayers.
Those are the other people's money.
The taxpayers are being fleeced, and your kind of research and your efforts here in
the government are showing this to people.
Hey, this is happening to your money.
And they get upset, and they talk to people like Dr. Raz, our president and our vice president,
and really light a fire for them to do something about it, which they are.
So thank you for contributing your expertise to this important work.
And I want to thank everyone who joined us here in person, also the livestream folks.
We appreciate your engagement and commitment to protecting program integrity here at HHS.
Thank you.
