HaloMD is either a lifeline or a leech, depending on who you ask.

Providers say the billings middleman is a vital source of help for doctors and medical clinics seeking fair reimbursement for out-of-network services under the No Surprises Act. Insurers contend the company is greedily sucking profits out of the law’s dispute resolution process, driving up U.S. healthcare spending all the while.

The controversy swirling around HaloMD has grown along with the company. HaloMD, formed in 2022 to help providers submit and win surprise billing contests, now represents more than 150 U.S. medical groups, and initiates more disputes in the NSA’s arbitration process, called independent dispute resolution or IDR, than almost any other entity.

That volume, along with HaloMD’s track record — the company boasts a success rate north of 90% and generates median payouts more than nine times above in-network rates for its clients — have put a target on HaloMD’s back, according to Patrick Velliky, the company’s top lobbyist.

HaloMD has been sued by multiple insurers and frequently called out by payer advocacy groups worried that bad actors are hijacking IDR. The arbitration process, which insurers and providers enter when they can’t agree on reimbursement for an out-of-network bill, has proven more expensive than anticipated as providers submit torrents of disputes, prevail in more than 80% of cases and win awards well above in-network rates.

Estimates of runaway spending have fueled investigations and calls for reform on both sides of the aisle, including from the Paragon Health Institute, a conservative think tank with close ties to the Trump administration, Rep. Frank Pallone, the top Democrat on an influential House committee, and a cadre of consumer and employer groups.

But the NSA is actually saving money — according to HaloMD, which has built a lucrative business on the back of IDR.

Earlier this month, Velliky authored a study estimating that the NSA has cut spending on out-of-network emergency medicine care by billions of dollars. Some researchers accused the analysis of relying on shaky data, while pointing out HaloMD’s incentive to defend the status quo. The intermediary earns a percentage of the net payment awards it secures for healthcare providers.

But Velliky stands behind his study — and IDR writ large. In a wide-ranging interview with Healthcare Dive, HaloMD’s chief external affairs officer refuted criticisms of his research, positioned HaloMD as a necessary counterbalance to powerful insurers and outlined why regulators and lawmakers interested in IDR reform should focus on the margins, instead of overhauling arbitration itself.

HEALTHCARE DIVE: Some researchers criticized the methodology of your study, including that you used the qualifying payment amount, a metric representing in-network rates, or some percentage of the QPA as a stand-in for out-of-network reimbursement outside of IDR. How would you respond? 

PATRICK VELLIKY: I’m very, very transparent in the research that there are places where I had to make assumptions. There isn’t good publicly available, verifiable data on what claims that don’t go into arbitration are reimbursed. But we do have good starting points where we can approximate what those likely look like. There are a couple places where you can find evidence that initial allowed amounts from insurers reflect roughly the QPA.  

I am not at all concerned about that assumption. The reason I felt it was necessary to ensure that we had some very conservative alternatives to subbing in the QPAs is that it’s very possible that the QPAs outside of arbitration are materially different from the ones in arbitration. It’s perfectly valid to raise as a criticism.

How do we know, though, that out-of-network claims outside of IDR aren’t just being paid at the same level they were before?

We know from the fact that we bring on new clients. We have physician groups in emergency medicine inquire about having HaloMD manage their process. And when we onboard a new client, we are able to see their initial allowed amounts. But that’s not something I can pull into the research itself because it’s proprietary. 

Is that data something HaloMD would ever make public or analyze in some way?

We actually are. I’m talking with a third-party research firm to more or less recreate the work that I’ve already published with fewer assumptions. My own research is not the last and final say in the cost impact of the IDR process, right? It is imperfect. But I consider it to be a meaningful proof of concept that, more than anything, calls out a couple of really important facts.

One, existing research is missing a comparator. Most of what has been published so far does not have a baseline against which to compare the cost of the NSA and the IDR process. Without that, folks will interpret any costs associated with NSA and IDR as new costs.

If I tell you that the new Ford Escort costs $35,000 and I don’t tell you what it cost last year, all you know is that it costs $35,000. And that’s what you’re seeing with the interpretations of the impact of IDR. Research saying that IDR costs $22 billion — that only matters if you understand what the cost of the same claims would have been without IDR.

We do have a comparator, though, to talk about whether or not the NSA has lowered spending: the QPA, which represents what those claims would cost in network. It’s just not a metric that providers think is fair to use as a baseline.

That’s a different question — how do out-of-network costs compare to in-network costs. And that’s an interesting question. But the only way to know whether or not the NSA and the IDR process has increased, decreased, unchanged the cost of out-of-network care is to compare it to what that same care used to cost. The portrayal of the cost of out-of-network care under the IDR process has invited a misunderstanding as to what it is that's being measured. 

Another criticism of your research is that it focused solely on emergency medicine, which is one of the only specialties where IDR awards are generally in line with pre-NSA out-of-network rates. Other specialties have sky-high outliers that inflate spending — surgeries and neurology, for example, are paid at 1,355% and 2,400% of the QPA. Why exclude these datapoints? 

I would love to do the same analysis across specialties. I am confident that there would be some different results across different specialties. The reason my work focuses on emergency medicine, candidly, is that it’s the only specialty that had all the data points publicly available.

It’s true that emergency medicine tends to cluster around 300% of the QPA, and there are other specialties that cluster at higher amounts. But it’s important that emergency medicine is the largest specialty represented in the IDR process. Given the even smaller proportionality of those other specialties, I'd be very, very surprised if their net effect was anything other than a reduction in out-of-network spending against a pre-NSA baseline. 

Your research also didn’t include costs related to arbitration fees and administration, which other research estimates drive billions of dollars in spending. Why carve those out if your goal is to get a full picture of the NSA’s cost effects?

I didn’t include it for a few reasons. I believe a lot of IDR spending is avoidable, and what I wanted to convey in the findings is the medical spending effect. That’s because insurers should be making networking decisions based on medical spending, right? So my core argument is that administrative fees would be avoidable if insurers agree to contract at rates that would still end up costing them less for those services than they would have paid pre-NSA.

Secondly, it’s hard to figure out where to draw the line as to what administrative costs are attributable to IDR. Some are obviously and clearly a part of IDR that didn’t exist before. But what about fees from third-party administrators to employer sponsors related to shared savings programs? Does that need to be factored in? At that point it becomes difficult.

On networking — if I’m a provider, what incentive do I have to go in network if I can make significantly more in IDR? Can you respond to the worry that the existence of IDR and the lure of generous awards inside the system will keep providers out of network?

Providers would generally prefer to be in network. It gives a timely and reliable cash flow, and you don’t have to worry about a very complicated process.

Now — I should be clear — what I’m not saying is that every physician and every practice utilizing the IDR process is an angel sent from heaven. There are absolutely folks who, if an opportunity presents itself, will try to take advantage.

And there are outliers on both sides. There are outlier outcomes in arbitration that should be looked at, and that if they are allowed to proliferate, could have an adverse impact on in-network rates as well. On the payer side, there are companies that offer $1 or $0, or attempt to calculate the QPA as low as possible, or refuse to make timely payment.

A lot of folks would say that HaloMD is taking advantage. You’ve been sued by a handful of insurers for allegedly exploiting IDR, including by flooding the process with ineligible disputes and using deceptive letters and misleading data to win cases.

Insurers are very, very frequently wrong about eligibility. I have thousands of examples from our clients where an insurer claimed a dispute was ineligible, and we were able to verify it. There’s a very good reason why plans weren’t given the authority to determine eligibility in this process.

The real reason we have become the target of the plans is that we are the largest player here. We initiate about 21% of annual disputes. It’s more about size than anything else. But 80% of the groups we represent are mid-sized to small independent medical groups. Without a partner like us, they wouldn’t be able to access the process and wouldn’t be able to compete with larger entities.

We’re proud of the fact that we provide that access to independent groups because it’s what allows them to stay independent. We believe that we are a hedge against consolidation.

The real reason we are a target for the insurers is that they never anticipated that some of the small groups that they had otherwise been able to offer take-it-or-leave-it rates to would suddenly be able to stand up for themselves and demand fair reimbursement for the services they’re providing.

If legislators and regulators interested in reform shouldn’t be looking at HaloMD or other companies filing a huge volume of disputes, where should they be looking?

The outliers. The problem is a very, very small portion of disputes are obtaining extremely large awards.

You don’t have to throw the baby out with the bathwater. You have to provide some additional scrutiny on some of these outliers. I think there’s a need for some IDRE [arbiter entity] oversight and training — I’d like to know how some of the really large awards end up happening. There are instances where it’s worth examining how the IDRE reached the decision that they did.

There’s a growing focus on arbiters’ financial incentives. IDREs are only paid if a dispute is found eligible, so they’re incentivized to find disputes eligible. Research also shows that arbiters that more frequently rule in providers’ favor give out larger awards and are also given the most cases, which throws into question whether they’re serving as neutral referees.

Whether or not they’re actually motivated by it, there is a perverse incentive around eligibility. That’s undeniable. It would make a lot of sense to delink payment from the eligibility decision. You could absolutely devise an improved framework where if a provider has the correct information from the payer and they still submit an ineligible dispute, they should probably be responsible for paying an IDRE fee.

To the incentive around outcomes, look — if IDREs are making decisions because they want more business, that is concerning. And that should be something CMS looks at during the recertification process.

Does HaloMD prefer certain arbiters where it’s had more success?

We have very intentionally built a process that removes itself from potential direct biases. It’s one of the few parts of our process where we actually want less human intervention, not more. So we’ve got several criteria that feed in autonomously to a selection process, and that includes the speed at which arbiters process refunds of fees — that’s a big cash flow issue for a lot of clients — and how frequently the IDRE makes an eligibility determination that we disagree with or that is later found to be incorrect.

Obviously our rate of success is a part of that metric. It would be insulting to your intelligence to pretend like we don’t care about win rate. But it is far from the only factor.