Wharton's Mark Pauly and Genevieve Kanter, a professor at Penn's medical school, discuss potential changes to physician kick-back rules.

Most older Americans are dealing with multiple health issues that require them to see different doctors or go to different health facilities. But what if these different providers were working together to deliver care to the patient? That’s the idea behind accountable care organizations (ACOs), which deal with Medicare patients. The goal of ACOs is to eliminate inefficiencies, control costs and improve overall health outcomes. But these partnerships often run afoul of anti-kickback legislation known as the Stark Law, which was created to keep physicians from referring patients to services in which they may have a financial interest. The Department of Health and Human Services wants to reform the law, so it’s likely to be on Congress’ agenda this year. Genevieve Kanter is a senior fellow at the Leonard Davis Institute for Health Economics at Wharton and an assistant professor at the Perelman School of Medicine at the University of Pennsylvania. Mark Pauly is a Wharton professor of health care management, and business economics and public policy. They have conducted research on the topic and recently wrote an article for  The New England Journal of Medicine titled “Coordination of Care or Conflict of Interest? Exempting ACOs from the Stark Law.” Kanter and Pauly joined the Knowledge at Wharton show on SiriusXM to talk about what they describe as “a very complicated piece of legislation.”

An edited transcript of the conversation follows.

Knowledge at Wharton: What is the basis of conflict between accountable care organizations and the Stark Law?

Genevieve Kanter: In order to answer that question, we should talk about two things. One is exactly what an ACO is and how it’s organized. An ACO is a group of doctors and hospitals that typically, historically have not worked together, but they are joined together to jointly contract to deliver care for a specified population of patients. This is important because one of the issues with our health care system is fragmented care. For example, as we mentioned in the New England Journal piece, if you’re a typical Medicare patient with cardiovascular disease or coronary artery disease, you will see about 10 doctors in a given year over six different sites. People think that that’s undesirable, so ACOs are designed to help doctors and hospitals coordinate care.

One of the things they do to coordinate care is that doctors can refer patients to hospitals — even hospitals that they are not employed by — and this conflicts with the Stark Law. The Stark Law is, shall we say, a very complicated piece of legislation.

This was legislation introduced by Pete Stark, a former congressman in California in 1989. Over the last 15 years, there have been multiple changes in the legislation as well as in the regulations. It’s known as the physicians’ self-referral law, and it’s meant to prohibit certain kinds of physician referrals. For example, say you have lower back pain and go see a doctor. The doctor says, “Back pain is really difficult to diagnose. With what I have here in my physical exam, I’m not quite able to do it. I’m going to send you to an imaging facility. You need an MRI, and I know of a really good one just down the block. They’ll set you up, and then I’ll be able to see the MRI and really diagnose you.”

“It’s the Dutch boy with the finger in the dike type of law, because there is water pouring out on both sides around it.” –Mark Pauly

What the doctor doesn’t tell you is, first, MRIs are almost completely uninformative in terms of helping doctors diagnose back pain. Secondly, what the doctor doesn’t tell you is he may have a financial stake in the imaging facility. The Stark Law is meant to prohibit these kinds of referrals. Where this maps into ACOs is that ACOs are in these joint arrangements where they can share savings by coordinating care, but they also have referral systems, so the doctors can refer patients to the hospital — and so they conflict with this prohibition on physician self-referral.

Knowledge at Wharton: How much of a concern has there been about potential violations of the Stark Law over time?

Mark Pauly: Before the ACO era, a lot because there’s a lot of jealousy in the health care sector. People don’t want other people profiting, and neither do patients want their physicians being motivated by profit rather than by what’s best for the patient. It has been a make-work program for lawyers — there is no doubt about that. It has influenced the configuration of the traditional fee-for-service sector substantially by saying that, as a physician, you’d be punished if you refer profitable business to some facility in which you have part-ownership.

Knowledge at Wharton: In the business world, this type of referral goes on every day. But there are different considerations for referrals in health care, correct?

Kanter: That’s right. There’s plenty of evidence, at least in the health care sector, that physicians who financially benefit from the referrals they make tend to do a lot more for their patients. They prescribe more drugs if they can benefit from the prescriptions. They order more tests. They order more imaging if they can financially benefit. That results in not-good outcomes necessarily for the patient, certainly in unnecessary care, as well as greater health care spending and just waste.

Pauly: I think the punchline is, you’d like your physician to be giving you advice or making decisions about referrals based on what’s best for you, although maybe a little bit about what’s best for Medicare, if it was cheaper. But the problem has been in the fee-for-service side that the prices that Medicare would pay to this imaging facility [the physicians] would own would be higher than their costs. So, there would be profit. You’d be tempted to refer to the facility you own. Having created the temptation, Medicare or the government then has to put in rules saying, “You shouldn’t see them.”

Knowledge at Wharton: Would it also increase the potential for doctors who would want a financial interest in an MRI facility or some other entity?

Pauly: Yes, that’s right. Many ACOs have a fee-for-service chassis, meaning they get paid a fixed dollar amount per patient, then in terms of contracting for services, they pay doctors and hospitals. Those doctors and hospitals get paid per service, and they’re subject to these same kinds of temptations because they’re being paid fee-for-service. So, it’s a little more complicated because the topline ACO that’s getting the money, if they were just greedy for money, would want patients to get fewer referrals and less care, and they would head off to Mexico with the money.

But on the other hand, the people further down are still subject to the same kind of temptation that has persisted. After Genny turned up this feature, we ask ourselves in this article, is there a better way?

Kanter: One of the weird things about the Stark Law is it really regulates referral. There’s sort of an arms-length relationship between the doctor and the facility they’re referring to. Because of its accumulating and massive number of exceptions, the Stark Law actually gives doctors an incentive to purchase the imaging equipment, get it within the physician group, get it within the physician practice, and then they’re good on the Stark Law.

Pauly: The other extreme is, if I were a salaried physician, I could and probably would refer people for MRIs for lower back pain, if they insist on it, to Perelman (Perelman Center for Advanced Medicine, University of Pennsylvania). People would think I’m salaried, so I don’t benefit financially from that. But I do benefit financially if the overall system makes a lot of money, which it’s doing. And maybe even in some systems other than ours, the compensation for a physician may depend on how much rain they caused to fall in the system as a whole. It’s the Dutch boy with the finger in the dike type of law, because there is water pouring out on both sides around it.

Knowledge at Wharton: We’ve seen a continuous push in the Trump administration to roll back regulations. Cutting regulations usually benefits business. In this case, will it benefit the business of individual doctors and hospitals as well as health businesses that are on the fringe of the medical industry to begin with?

Kanter: That’s right. I saw the deputy director of the Centers for Medicare and Medicaid (CMS) speak about the Stark Law at Brookings, and he did mention that their drive to rethink the Stark Law, in relation to accountable care organizations in particular, was really part of a broader initiative to cut down regulation. I can tell you, if people’s positions are any indication of their interests, the National Association of Accountable Care Organizations supports exempting ACOs from the Stark Law, and it has really been behind this.

“I would be even more enthusiastic about ACOs if it had been shown that they actually improve the coordination of care.”— Genevieve Kanter

Currently, ACOs have a temporary waiver from CMS so that they aren’t prosecuted either by the Stark Law or by the anti-kickback statute. But the ACOs want more. They’re saying, “Look, there’s a lot of regulatory uncertainty. We would like you to codify this exception.” And the government can do it in two ways. One is they can have CMS do it, and CMS has some authority to promulgate some exceptions under certain conditions. We can also have legislation do it. We know that legislation was introduced in November to codify this exception.

Knowledge at Wharton: By allowing the waiver in the current law, it sounds like they have opened the door to take the next step. If the waivers had not been granted in the first place, we may not even have this discussion, correct?

Kanter: That’s a good point. I think because there was so much uncertainty, because the law has been changed so many times, given so many exceptions, given so many exceptions to the exceptions, it’s very complex that both physicians and hospitals have viewed it as sort of this random ticking bomb. They don’t know when they might be prosecuted. But it was really designed as a temporary measure to say, “We really want to push this ACO model forward. Before we really think about it in a regulatory way, we’ll give you a temporary reprieve. While you sort that out, we’ll give you the best chance of succeeding. And we’ll also think about how we want to regulate this and deal with this.”

Pauly: Seema Verma, who’s the head the Medicare, was here for the Wharton Health Care Program business conference. I had a chance to talk to her. What I took away from that is that the administration is in favor of removing regulation and rules that are so complicated you need a whole team of lawyers to figure out how to deal with them. And they’re not particularly fond of the ACO idea. That was an Obama idea. If it would save money or improve outcomes, it was OK, but they’re not ideologically devoted to the idea. In a way, I have to worry about this because when I stop teaching, I have to go on Medicare immediately. And an ACO really is just like a Medicare HMO for Democrats. It really sets up the same internal incentives as are present in a Medicare Advantage HMO. It’s just a little kinder and gentler because it’s not, for example, permitted to deny payment if you use care out of the system or things like that. But it’s basically that model that people have been talking about for years and years and likewise talking about coordinated care for years and years.

Some people have great faith in the ability of these systems to provide better care at lower cost. But research that I’ve done with Rob Burns (Lawton Robert Burns) in my department, who was the reason why Genny got in touch with me about this, suggests “not so fast.” There really isn’t evidence that these ACOs do all that much good to begin with. They may save a little bit of money now that the ones that couldn’t save money have dropped out. But because it’s voluntary, you can’t really tell. I don’t know how you feel, but from a personal point of view, when I want coordinated care, I want it to be coordinated. But then I want them to stop bothering me.

Knowledge at Wharton: Coordination in health care is something we want to achieve. But we also want to achieve it in the process of doing the greatest good for the patients, not having a side angle for doctors to make more money.

Kanter:  Absolutely. In fact, I would be even more enthusiastic about ACOs if it had been shown that they actually improve the coordination of care. What we see is that for many ACOs that are led by hospitals, for example, even though they have money for the shared savings, their revenues come from, frankly, in-patient admissions. So, they have less incentive to send people to primary care docs to avoid hospitalizations. What we have found is there has only been a small decrease among hospital ACOs in hospital admissions.

A lot of the improvement in costs that we’ve seen have really been driven by I would say is a little bit of trickery, in terms of the kind of utilization management that has been going on, and we don’t have any direct evidence that it improves coordination of care.

“It all depends on the internal culture of the hospital and whether they can persuade all those fast horses that they employ to run slow.” –Mark Pauly

Pauly: Yes. Although this is not quite as definitive, the research does suggest that if a hospital organizes an ACO or a hospital system, and the reason they would is because they’ve got all the money and the borrowing power, it doesn’t work nearly as well as if it’s organized by, say, a large physician group practice — because hospitals are good at treating really sick people in an expensive way. Can you make a fast horse run slow or go the other direction? This is the dilemma that they face. Some have been successful, but it all depends on the internal culture of the hospital and whether they can persuade all those fast horses that they employ to run slow.

Knowledge at Wharton: But the difference between a large physician group and a hospital in many cases are the beds. If you have the beds, you’re making your money. Whereas a physicians’ group isn’t concerned about that. They are thinking about treating the patient and where they are going to potentially refer them if they need to go someplace else.

Pauly: Yes. If you’re going to pick an ACO run by a hospital, you should try to find a hospital that’s already got all its beds full so that it wants to go in the right direction, rather than in the wrong direction.

Knowledge at Wharton: What specific things do you think need to be examined in the Stark Law?

Kanter: It is written as a blanket prohibition for self-referrals. Consequently, there may be some cases where I’m a doctor and I also treat patients at the hospital, so I’m getting some compensation. I would like to send my patients there, but Stark prohibits me from doing that. That seems like a reasonable thing to do, to accommodate. Because the law is a blanket prohibition, we start introducing exceptions. What we now see characterizing the Stark Law is multiple exceptions, either done legislatively or through regulation, which essentially weaken the law and regulation.

A second feature, which is interesting and why hospitals are in play, is that even though it looks like we’re regulating doctors, part of the Stark Law imposes a penalty for anyone who submits a claim for a referral that was prohibited. That means if a doctor refers a patient to the hospital and shouldn’t have done, and the hospital tries to bill for it, they are on the hook for that claim. I think the penalty is $10,000 per claim, as well as three times the amount they were trying to bill for. Hospitals have more at stake. Perhaps that’s why with ACOs, they’re really feeling strongly about this.

Knowledge at Wharton: What happens if the Stark Law is changed in the coming months?

Kanter: In terms of the general population, I would say probably diddly squat. Patients are largely unaware of participation in ACOs and how that might affect them and affect their providers.

“What we now see characterizing the Stark Law is multiple exceptions, either done legislatively or through regulation, which essentially weaken the law and regulation.” –Genevieve Kanter

Pauly: I think the administration is probably moving in the direction of removing restrictions on ACOs. The original regulations went on page after page and created this kind of safe harbor from pre-existing regulations that we’re talking about. Ideally, its vision would be that they should just compete, and the management of the ACO should be given freedom as to how they want to contract with various sources that are going to provide care for the patients, including whether they want to do the policing of excess referrals to a physician-owned facility, rather than have the federal government say, “We’re here to help you,” and step in and do that.

But the danger is one we’ve already talked about because the topline incentive to the ACO, if it’s interested in profits, is to incur as little cost as they can, which may mean that they will try to not provide services that might be worth more than their costs.

The solution to that is to measure outcomes so that I can look up stats or some measure of the outcome from the different ACOs that I might want to choose among. But we’re not used to thinking of health care that way. We need a Michelin Guide for hospitals and health systems, but it’s not the way people are used to picking their doctors and hospitals.

Knowledge at Wharton: What’s the reaction of the insurance industry to this?

Kanter: I suppose they’re waiting with bated breath. The Stark Law primarily governs Medicare transactions. But there’s certainly spillover, and there’s certainly leadership value to CMS taking a stand.

Pauly: Yes, there are some private ACOs above and beyond Medicare. They’ve been kind of a dud. They haven’t really taken off. But private insurers were, in some sense, protected by the Stark Law from having to police referrals in physicians treating their patients because Medicare was doing it, and they’d like to continue to piggy-back on Medicare.