Incentivizing Cardiologists to Do the Right Thing

A Discussion of Payment and Quality Metrics

; Karen E. Joynt Maddox, MD, MPH


April 23, 2018

Robert A. Harrington, MD: Hi. I'm Bob Harrington from Stanford University. I am here in Orlando, Florida, at the American College of Cardiology (ACC) meeting, and it is always a great time to catch up with colleagues about various topics that may be of interest to the general cardiologist.

Today my friend and colleague will talk with us about some of the current thinking on health policy and incentives. Karen Joynt Maddox is assistant professor of medicine from Washington University, and she also has an appointment in the department of social work, which is where policy lives in public health at WashU.

In an era of value-driven care, how do we incentivize health systems, and how do we incentivize practitioners to do the right thing from a patient perspective and from a society perspective? What are some of the incentive models that work from a policy perspective? What are some of the implications of adverse unintended consequences that incentives might lead to?

Karen, thanks for joining us.

People will do what they are paid to do. You can argue that perhaps in medicine we are somehow immune to that. But in reality, we are not.

Karen E. Joynt Maddox, MD, MPH: Thanks for having me.

Learning the Ropes of Value-Based Payment

Harrington: You are in a nice intersection of clinical medicine and policy. You spent some time in Washington working with the Centers for Medicare & Medicaid Services (CMS) on this topic. Why don't you reference what you did there?

Joynt Maddox: I had the opportunity to spend 2 years working in the main office of Health and Human Services at the Office of Health Policy, which is an advisory think tank research group within the Office of the Secretary that advises the Secretary on all matters of health policy, such as financing, incentives, and access.

I had the opportunity to work on a report for Congress focused on value-based payment and social risk factors. Lots of folks around the country, including members of Congress, recognized that as we move towards value-based payment, the issue of social risk is going to be a big one. No one really knew what to do about it. To be fair, no one still knows what to do about it. Our job was to put a report together focusing on that issue.

Why Use Incentives?

Harrington: Let's peel some of that back. From a health policy perspective, talk conceptually about the idea of incentives. What are incentives intended to do from a policy perspective?

Joynt Maddox: Payment reform, in general, is the attempt to use financial incentives to signal people where to put their effort and their energies.

Harrington: Follow the money.

Joynt Maddox: Follow the money. People will do what they are paid to do. You can argue that perhaps in medicine we are somehow immune to that. But in reality, we are not.

Harrington: That has been well-studied and documented. With volume-based care, we definitely followed that tranche of money.

Joynt Maddox: Quite effectively, actually. The intent of financial incentives turning toward quality is to get us to start focusing on quality. It is to say that we are not just going to continue to pay you to do more, we are going to start paying you to do better. Ultimately, I think that moves us in a direction towards focusing on outcomes; focusing on what is the outcome of what we are doing as opposed to just the elements of care that we provide. Those two things are obviously a long way from each other, but the intent of financial incentives is to start moving us more towards thinking about producing the best possible outcomes for our patients.

Harrington: From a cardiology community perspective, although we may not understand everything about MACRA (Medicare Access and CHIP Reauthorization Act of 2015), MIPS (Merit-Based Incentive Payment System), and alternative payment models (APMs), we are pretty far along in the quality arena as a community. We understand performance measures. We understand how it is a process measure, not necessarily an outcome. We are also starting to understand that we have done a good job on the performance measure, but now we really need to focus on the outcome.

Joynt Maddox: In many ways, cardiology is absolutely poised to lead this forward. Another thing I learned from my time in government is that, with their best intentions, people who are not clinicians can make policies that may have a lot of unintended good or bad consequences. Obviously we worry much more about the latter.

As clinicians, we understand that the point of these things is to drive better care for our patients. We know we can do better. We know that we do not meet the best rates of anticoagulation for atrial fibrillation or for stroke care when people show up to the hospital. Our system is not very patient-centered. As soon as you have been on the other side or have had a family member on the other side, you know that we can do better. As cardiologists, particularly in the space where we have a lot of evidence, we really should be able to drive this forward to make sure that all of our patients have access to high-quality care.

Is Evidence Being Used for Policy-Making?

Harrington: Let's unravel some of what you talked about. You said that sometimes nonclinical policymakers can go awry—not by malice but because they may not understand the clinical context and consequences of certain decisions. As a cardiology community, we depend heavily on evidence. What is the state of using evidence to make policy?

People who are not clinicians can make policies that may have a lot of unintended good or bad consequences.

Joynt Maddox: It's in its infancy. When a new drug or device is introduced into a broad population, we have a very prescribed set of ways that we test that element. Even before randomization, we want to get some basic efficacy and safety information. We have animal models. We move things in a stepwise prescribed way. You would never put a new drug out into the population without understanding both its good and bad consequences. We do not do that in policy. The Center for Medicare & Medicaid Innovation (CMMI; "The Innovation Center") was made in part to create the potential to do that and to have some demonstration projects.

Harrington: It is supposed to be the "hub" of innovation for payment models, if you will.

Joynt Maddox: Right. They have put out a number of experimental payment models in limited test settings so that you can learn, evolve, iterate, and put out better programs if you want to go big. Most of the programs we are currently paid under have not gone through that process.

Harrington: It is interesting. Now that I oversee a department of medicine, the example I often use is residency work hours. The concept is the right one, but years ago why didn't we actually study it in a formal way before changing policy? We made many leaps of information, from the airline industry or other industries, rather than saying, "How are you going to deal with things like handoffs and unfamiliarity with patients?" You have to be careful in the policy realm.

Joynt Maddox: You do, and I think what we are seeing now is the collision of the best intentions with the reality of the health system. We do not have the infrastructure we need to be able to do the things that are required to make good policy. If we could perfectly judge quality, use outcome measures, and feel like they are reflective largely of quality as opposed to underlying patient risk, we could do a lot of things. We lack the data infrastructure to be able to do this on the outpatient setting in many cases. As a community, we can see that we need to get better, and we understand that we should be moving in that direction. The methods to get there on the policy side are very blunt instruments right now.

MACRA and 'Not so Modest' Penalties

Harrington: Let's talk about some of those instruments. I want you to define MACRA, MIPS, and APMs. Then, we'll discuss a controversial policy initiative that is heating up the cardiology community right now.

Joynt Maddox: MACRA was the law that replaced the sustainable growth rate (SGR) formula. I think that cardiologists were looking at something like a 30% pay cut if that ever went through.

Harrington: Cardiologists understand SGR.

Joynt Maddox: MACRA was probably a good thing. It created two separate payment tracks for outpatient payment. Inpatient payment is separate. The default track is MIPS. MIPS is basically just pay for performance. You turn in some basic quality measures, and you get your pay for any given part B billable service adjusted up or down by a certain percent. The things you have to do to succeed in MIPS are sort of a "Frankenstein" of technology, practice improvement, initiatives, cost, and quality. Those things all get lumped together, you get a score, and then you get a bonus or a penalty.

Harrington: Let's talk about the size of those penalties. They started off very modest, and now they are not going to be so modest.

Joynt Maddox: They are not going to be so modest. The precursor to MIPS—the value-based payment modifier, which is what we are under now—was essentially a percent. Starting fiscal year 2019, which starts October 2018, there is a 4% up or down penalty. It is 5% the year after that, and it eventually goes up to 9%. That is real dollars.

Harrington: Remind people that it is too late to affect the 4%. Those data are in.

Joynt Maddox: 2019 penalties and bonuses are based on 2017 performance. We are currently in the 2018 performance period, which will affect payment in 2020.

Harrington: Unless the law changes, ultimately we are going to get up to the 9% mark, and that is real money.

Joynt Maddox: It is real money, especially for practices, which generally can bear a lot less risk than hospitals and health systems.


Harrington: Let's talk about the APMs, which I thought when initially introduced might be more interesting than MIPS.

Joynt Maddox: I think they will be eventually.

Harrington: They are clearly not there.

Joynt Maddox: If you are looking at a fork in the road, there is the well-traveled fairly straightforward pay-for-performance track. Then there is this murky, "What can we do for alternative payment?" model track, which is where CMS wants people to go down. The carrot is a 5% flat bonus if you go there. The problem is: There is not much to go to at this point.

Harrington: They are seeing people go there because the 5% is not going to take a lot. The challenge is that once you are down there, then what are you going to do?

Joynt Maddox: I would think that the most common way that cardiologists are getting into that track is through accountable care organizations (ACOs). If your organization is one of those, you are down that track. You can stop worrying about it. For a practice or a hospital that is not currently part of an ACO, we do not have much there. The Bundled Payments for Care Improvement-Advanced (BPCI-Advanced) initiative, which closes enrollment in 2 or 3 days for the first year, is really the first feasible specialty track for cardiologists.

Harrington: Cardiologists do bundles all the time.

Joynt Maddox: Exactly. Things like percutaneous coronary intervention, implantable cardioverter defibrillators, and even a couple of outpatient bundles. It is a little early to know what is going to happen there. CMMI has been quite thoughtful by taking what they learned from the initial BPCI initiative and making a few improvements to BPCI-Advanced. They have not released full details, but hopefully it will be a feasible opportunity for some.

Harrington: The great thing about cardiology is that we are an evidence-driven field. We are also a data-driven field, more than other areas of medicine. We have data infrastructures. We have Get With The Guidelines, the National Cardiovascular Data Registry (NCDR®), and now the PINNACLE Registry through ACC. People can participate by contributing their data and by having their data benchmarked.

Joynt Maddox: MIPS is a tough program. MIPS was the government's effort to throw urologists, cardiologists, and endocrinologists into one bucket and try to make something that was one size fits all. I am not sure that makes a ton of clinical sense. A different way to go about it would be if we, as cardiologists, got together and said, "What are some high-impact measures we care about? Can we collectively move this ball down the field?"

Thinking about underlying goals and how we can use MIPS to help us move quality forward is probably better than trying to take on that "Frankenstein" umbrella program.

Incentives for Reducing 30-Day HF Readmissions: The Positives

Harrington: That is well stated. Let's take on one of the incentives which has garnered a lot of controversy in the community: the 30-day heart failure (HF) readmission. It sounds like a good idea because a lot of people who have HF get readmitted—a bad thing from a patient perspective. Therefore, we ought to improve HF care. Some have argued that we are putting people in a perilous situation by not readmitting them, so the risk for mortality is getting worse. It is a pretty big debate in the community. What is your take on this?

Joynt Maddox: I will go squarely down the middle. Both sides have merit. If we went all the way to either side, we would miss good things for our patients. The argument that reducing readmissions is good has merit. We have all seen patients come back to the hospital because of failures of some sort. Whether or not it is a hospital failure, system failure, community failure, or even an individual failure has been the source of some discussion. Some patients have chaotic unlivable outpatient lives.

Harrington: If you do not pick up your medications and do not take them, you may get readmitted.

Joynt Maddox: Exactly. If you do not have access to care, you may get readmitted. If you are prescribed the wrong medications, you may get readmitted.

Addressing readmissions is a good thing because it requires us to get out of our comfort zone and say, "How are we going to be patient-centered about this? When do they need to see someone in follow-up? How can we make sure they get their medications?" That is different from a hospital quality measure of giving someone an aspirin.

It is fundamentally where we need to go next. The concept of HF readmissions is really important. It has pushed a lot of innovation in the outpatient setting such as infusion centers, home visits, and stuff that is good and patient-centered.

Harrington: It has also made people realize that they have to open up their practices to have acute care visits. I agree with you; it has done a lot of good things. Let's talk about the other side.

Incentives for Reducing 30-Day HF Readmissions: The Negatives

Joynt Maddox: The flip side is: Why do we see these patterns of HF mortality going up at the same time as readmissions are going down? I think it is an open question. There are a couple of potential answers, and I hope someone out there is trying to figure out which is true. They may all be true.

It is probably more financially harmful to your hospital to have a urinary tract infection than an HF death.

One thought is that working on readmissions has essentially distracted us from working on mortality. We are so worried about people coming back into the hospital that we have failed to keep pushing on that inpatient mortality. This is a hard thing to prove, but there is probably some merit. Every sought opportunity is a lost opportunity in some ways.

Another thing that is probably happening is that the population of HF patients in the inpatient setting is really changing. For one thing, there is a lot of HFpEF (heart failure with preserved ejection fraction), which we do not have good therapies for. Those folks are also more likely to die of other things. It is an older, sicker population with a lot more comorbidities. Some of the trends probably reflect differences in admission as we have cut people out of the hospital.

Harrington: It used to be that a lot of HF admission was for acute myocardial infarction (MI) with a large infarct. Now with acute coronary syndrome care, that population is getting more limited.

Joynt Maddox: More people survive ST-elevation MI and go on to have a longer course of very complex multimorbidity sorts of scenarios. There is some merit to that argument, too. On a population basis, we are not doing that much worse with HF. It is a problem with hospitalization, really.

Harrington: As a policymaker, would you say that the answer is not to discard the measure but to figure out why we are seeing these patterns of mortality and determine what contributes to that mechanistically? And then take the best of the readmission learnings?

Joynt Maddox: One thing we need to do is use "right-size" policy, if that makes sense. Right now, mortality is a smidgen of the value-based purchasing program. It is probably more financially harmful to your hospital to have a urinary tract infection than an HF death. I do not know that exactly statistically, but infections are all over the place in payment models, including a whole separate infection payment program. Mortality is one little sliver—that does not feel right. Then we have a program that only focuses on readmission without taking into account competing risks. Do you keep more sick people alive who are then more likely to be readmitted?

We could think more systematically about what as a community we want to incent. I do not think we should throw out the concept of readmissions. We may need to rethink how we approach readmissions in concert with a global approach to how well we are doing for HF. The same is true for acute MI. I do not think we should stop thinking about little pieces of it that we do not like.

Harrington: I remember, in the early days of the time-to-treatment metric, there was a lot of concern that the door-to-balloon time was so stringent that people were just inflating a balloon in the guide catheter to be able to say that they inflated the balloon. Of course, that was not what the intent was. I think we have gotten over that in the acute MI world and the needle has moved. Acute MI mortality clearly is continuing to decrease.

One thing we can do as clinicians is to try to keep informing this policy discussion.

Joynt Maddox: Acute MI mortality is under a readmission program too, and mortality continues to drop. I do not think it is as simple as one program causes something, and one does not. It is to say that these are different conditions, and we need to collectively think about the package of care we deliver and how that is going to get better.

Final Thoughts

Harrington: We think of basic science and clinical science, but in fact there is a lot of overlap. What we are really trying to understand is a biologic phenomenon that has mechanistic underpinnings. As a scientist, it starts to help you think, "Okay, we made an observation. Now we need to unravel it."

Joynt Maddox: Iterate, test, and learn. One thing we can do as clinicians is to try to keep informing this policy discussion. There are not very many clinicians in government, and there are certainly not very many in policy-making positions. Seeing this as a partnership in a way that has us drive our own fate is ultimately probably going to be more useful than just trying to ignore it. It is not going away. Collectively, we can make it better.

Harrington: I agree with you. As clinicians, we need to be engaged. We can be engaged as individuals, as you have, but we can also be engaged through the American Heart Association and ACC.

Karen, thanks for talking with me. I could do this for a while; there are a lot of interesting things to chat about.

Joynt Maddox: Thank you.

Harrington: I want to thank you, our listeners, for joining my discussion with Karen Joynt Maddox from Washington University in St. Louis about some of the implications of health policy incentives.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: