Podcast Episode Transcript: Reducing health care costs: You can’t improve what you don’t measure

With L. Gordon Moore, MD

Dr. Gordon Moore: Welcome to the 3M Inside Angle podcast. This is your host Gordon Moore, and with me today is Christopher Koller, who is president of the Milbank Memorial Fund, a 114-year operating foundation that improves population health by connecting leaders with the best information and experience. Before joining the fund, he served the state of Rhode Island as the country’s first Health Insurance Commissioner, an appointment he held between 2005 and 2013.

Under Chris Koller’s leadership, the Rhode Island offices of the Health Insurance Commissioner was nationally recognized for its rate review process and its efforts to use insurance regulation to promote payment reform, primary care revitalization, and delivery system transformation. Welcome, Mr. Koller.

Chris Koller: Thanks, Gordon. It’s great to be here.

Gordon: I would like to talk a little bit about health care costs, and before that, though, could you give me a thumbnail sketch of what is the Milbank Memorial Fund?

Chris: Sure. We’re an operating foundation. That means we spend money on our own projects, as opposed to giving money away in grants. As you know, we’re almost 115-years old, and our origins are really in New York City, where we’re still based. When our founders started to be concerned about—stop me if you’ve heard this before—the living conditions of people who’d recently immigrated to the United States, in this case, the lower east side, who are living in settings of poor sanitation, inadequate running water, livestock in the street who were causing poor hygiene and, in some cases, dying.

And our founders started this rather radical idea of a philanthropy, a foundation—at the time, it was radical—to start adjusting those conditions. That led it to think about access to health care and starting what became what were precursors to community health centers. They got into the public health movement, upstate in Syracuse, and fast forward 80 years and you have the Fund here, which is really focused on working with public officials, primarily, at the state level, to improve the health of all populations.

We try to give them the information, the experience, the relationships, the focus to work on some of these long-term issues. We also do a lot of communications. We publish written work, and we publish a peer-reviewed journal, The Milbank Quarterly, that’s been around for quite a while. So it’s a privilege to work with something with such a long history working on these issues.

Gordon: I’m glad you brought up The Milbank Quarterly. Still one of the most impactful articles I’ve ever read was Barbara Starfield and her colleagues who published a compendium on how primary care is a foundation for high performing health systems, and that’s a strong focus of your work as well?

Chris: Yes. It is. I’ve always been attracted to that work professionally. I cut my teeth actually in a staffed model HMO, so I got the primary care jones very early. I had the privilege of working with community health centers within a Medicaid health plan in Rhode Island. So I’ve been a long believer in the importance of primary care and a balanced delivery system. I’m very much influenced, like you, by Professor Starfield’s work and her findings, conclusive, that comprehensive primary care internationally is the only part of a health care system where an increased presence is associated with improved population health and improved health equity, in particularly, this time with declining life spans, and increased attention on social justice, and health care disparities, and primary care getting a declining share of the pie, as our work at the Fund has documented. We’ve really got to stay focused on it. So I’ve been much inspired by that work as well.

Gordon: I’m impressed by some of the statistics I’ve been seeing come out of your publication and from work that you’re doing around. Four to seven percent of U.S. health care spending goes to primary care, when in OECD countries, it’s more like 15 to 20 percent. And it makes me think that if there were some way that we could spend more in primary care, we could begin to enable them to do the full scope of work and engage in the kind of care and services that people need, that may move care upstream, as opposed to waiting for problems, and catastrophes, and missed opportunities downstream.

Chris: Yeah. We think of primary care as the bridge, if you will, to public health, to prevention, to community-based care, and we’ve worked really hard to publicize that five to seven percent number and work with states, in particular, who are taking steps to increase it. We did some of that work in Rhode Island. Oregon is a leader in this work. So that number can be changed, and we think it has to be changed if we’re going to make a dent in some of the health issues, the inequities that really haunt us in health care.

Gordon: I think about the $3-plus trillion budget right now for health care spending in the U.S. I think about adding dollars to that bucket to increase primary care spend, that it makes my knees weak.

Chris: Knees weak in what way? How do you mean? At the potential or at the difficulty of it?

Gordon: Well, it’s the difficulty. Where does it come from? Are we going to ask people to shell out more? Are we going to have employers tap in again to say we need to bump, almost triple, the primary care spend? “Could you open your checkbook?”

Chris: No. I don’t think you ask employers to do that. I will tell the Rhode Island story, since I’m fundamentally involved in it. When I became Health Insurance Commissioner, we discovered that not only do we have rate review—so we reviewed all the premiums of the health plans as they came in—but we had this mandate to direct health plans to improve the affordability of the health care system. And as my lawyer told me, “Chris, that’s a shall not a may. You have to figure out how to do that.”

And so we literally told the health plans “Put a greater proportion of your spending into primary care and less into the other services.” And the health plan said, “That means we have to say no to some of the specialists.” And we said, “Yes.” And they said, “Well, what should we do?” And I said, “Well, point to the fact that your regulator is telling you that you have to, and I’m telling you that you have to because we have an advisory commission of a bunch of employers,” who, just to your point, Gordon, said, “Look, we’re not going to fund this. This is about a transfer.” And low and behold, they did it.

The specialists didn’t leave town, they learned to live with three percent as opposed to five percent, and Rhode Island’s been able to move that five or six percent of spend number up to something closer to 10, 11—put in place behavior health integration, community health teams, help fund some primary care driven ACOs. You can do this stuff, but it’s a redistribution. We are not going to compete our way to better primary care for everybody.

We might be able to compete our way to a similar innovation at the margins for those who can’t afford it, but if we’re talking about having a primary care-based delivery system, that’s not going to happen by pitting primary care docs against one another. Public policy has to recognize that primary care is a common good that everyone should have access to and work to make it happen, whether it’s at the state level or at the Medicare level, in terms of how Medicare sets rates. Okay. End of sermon.

Gordon: No. I think about the models that I see that are being proposed. Often, it’s “Here’s 10 million on primary care spend. Let’s just turn that into a capitation rate so it’s zero-summit primary care.” And, gee, you have a potential benefit by somehow cutting yourself off at the knees and shared savings from that?

Chris: Yeah. In our experience—and we think there’s some lessons from Medicare’s comprehensive primary care initiative—when you get stuck in this “We need the money; give it to us, and we’ll give you the results,” and the payers say, “Show us the results, and we’ll give you the money,” you have to build a bridge. You have to pay primary care for some structure and process improvements, the elements of a patient-centered medical home.

You move them to a performance-based payments, but you don’t expect them to implement 24-hour access nurse care, coordination, electronic health records, team-based care on the paltry E&M codes that we’re giving them. You’ve got to give them some upfront money and some clear expectations for what they have to do to maintain that money.

And then over time, you move them to more performance-based accountability on a population basis. And the innovative commercial health plans have done this. Medicare has done this with CPC. We know how to do this. We just have to have the backbone to scale it. But it’s important that payers, whether they’re self-insured, or insurers, understand that it starts with investing more in primary care and articulating really clear expectations for what you want from them.

Gordon: When I think about, what do I get in return if you do this well—and I understand that I need to front the investment, and I should get this downstream—how would I measure that and know that it’s working?

Chris: I think the first thing you do is—and this is a tough conversation because this is where Medicare has gotten a little bit, I think, wrapped around the axle of CMMI. We don’t think you can expect short-term reductions in costs, and we counsel primary care advocates all the time. “Don’t talk about the return on investment in primary care in a year or two.”

The places that have been doing multi-payer primary care transformation do eventually see returns in costs. I’m talking about Oregon, Ohio, Arkansas, but they see it after 5 to 10 years. It takes a long time to unlearn the bad habits of fee-for-service medicine. So I think early on, you’re looking for process and structure improvements. You’re looking for 24-hour access. You’re looking for care coordination nurses. You’re looking for referrals management, disease management registries.

We’ve seen you start to see changes in the quality measures for chronic conditions, and then you start to see some reduced ED utilization. That’s the first place it starts to happen. A big driver of our health care costs is consolidated health systems and the prices they’re charging. Don’t expect primary care to move on that. And we’re talking about pennies. We’re talking about one or two points on the dollar, and we can find that from someplace else. And so you have the structure and process expectations, and then you start looking at quality measures, ED, and over time, you’re going to get to the kind of return on investment that you want.

And the other thing I would say, by the way, when we think about usual source of care, I think employers— because I know there are a lot of those who are listening here—are part of the bargain for health benefits and for value-based benefit design—should be that you pick a usual source of care.

It’s a perfectly reasonable expectation to say that I’m going to get not only zero cost sharing on wellness and preventative, but zero cost sharing on routine visits, but only for my usual source of care for this place that I pick. Because central to primary care is a good relationship with a team. How can you have a good relationship if I think I want to reserve the right to go anywhere in the health care system I want?

Now, you’re the doc. If you don’t give me 24-hour access, yeah, I’ll take my medallion, and I’ll go someplace else that does serve me better, and maybe, it is a CVS or an urgent care. But the payer should expect the beneficiary to designate a usual source of care. That’s part of the bargain for this. I can’t be allowed to just go anywhere I want.

Gordon: So that’s getting at a significant book of work around continuity of care and making sure people show up at the same practice, with the same teams, or they get to know them over time. And the sense I have is that there is something very subtle in the interaction. Because I remember taking my own call for my practice and recognizing voices in the middle of the night and thinking “Oh, if I’m hearing from this guy, that’s bad.” Or from that person, it’s like, “Okay. No, it’s all right. You’re going to do okay. We’ve been here before.” In your data, your research, is that borne out?

Chris: Well, I would hesitate to call it research, but some of the elements of high-quality primary care—and I’ve been privileged to work on the National Academy of Sciences, Engineering, and Medicine Committee on primary care. And we’re looking at, this is a long-term relationship between the patient and a team of caregivers. Well, what can payers, what can employers do to facilitate that relationship? Because relationship’s a two-way street. And if I’m allowed—moral hazard. That’s what I teach my students. If I’m shopping with somebody else’s money, and I’ve been given a free card that I can go anywhere in the system, that’s not the basis of a healthy relationship.

I should have the right to switch. You’ve got to do some things to keep me happy. Things like, give me access to answer my phone calls. Let me reach a nurse, and if you don’t, then I should be able to go someplace else. And that’s a problem because supply is limited right now. We want a mutual relationship based on a partnership between you and your patient. What are the tools in our payment, in our benefit design to encourage that relationship, not to discourage it?

Gordon: So make a choice upfront that I’m going to have this relationship, and I could switch it over time, but make that a force up top. That’s one piece. What about auto assignment for the 20, 30 percent of, let’s say, Medicaid enrollees who aren’t picking a PCP?

Chris: I think that’s important. I worked in Medicaid. We did RO assignment. I think you have to have in place a mechanism so that when the patient wakes up and says, “Oh, no, I don’t want to get my primary care; I want to get it over there,” you have to make it really easy for them to switch. Their voice is the rule because you don’t’ want to create bad financial incentives. Your mechanisms should always put the patient in the driver’s seat.

On the other hand—you know this—health insurance functions on a lot of nonusers, and that’s not a bad thing. The many healthy subsidize the few sick. So we have to have some attribution mechanism because there are going to be some people who are just healthy. They’re not going to touch the system, but they’re probably not going to use a lot of care either. So they’re not going to need much.

Gordon: One thing that’s explicit in what you’re describing is a relationship, and I wonder about the measures we use to understand quality of care. We have a handful of process-of-care measures, which you’ve identified as being important, and we can look at certain structural elements, present or not present in practices.

But I’m intrigued by evidence I see in some studies that say, if the trust I have in my provider, my clinician, improves the probability that I’m going to follow through on their recommendations—and I’ve seen this for colorectal cancer screening, for following through on treatment plans for chronic conditions. And trust in one of these studies was defined as “You don’t keep me waiting, and you listen to me when I speak.” And I wonder, are we missing a major element of understanding when we don’t have patient experience measures in the mix and driving of a core piece of understanding quality?

Chris: Yeah. I think to define quality, only in clinical terms, it’s really to do a disservice to patients and to perpetuate the doctor-knows-best paternalism. A lot of people have spilled a lot of ink trying to understand that stuff, and I’m not particularly an expert, but I think we can’t do patients a disservice by just dismissing them and saying, “Well, it’s about how clean the waiting room is,” or “It’s about how many and what are the diplomas on the wall?” I think patients, by and large, are smarter than that.

Think about the relationship between a parent, a child and a pediatrician, or a son, or daughter, and an elder parent in a geriatrician; there’s a deep, deep relationship there that is driven by the patient’s perception of trust, and I think we have to honor that, and build on it, and not just blow past it with a whole raft of clinical measures.

Gordon: So I want to come back to a comment you made, early on, when you were talking about health care cost growth targets. I want you to unpack that for me if you will.

Chris: Sure. The evidence, it’s very clear. I have to keep sight of it. We spend twice as much as a country per capita than any other country, OECD country, on health care, and what we get for that is shorter lifespans. That’s a pretty lousy return. And I know we’ve said for years, “Oh, my goodness, health care is eight percent of the U.S. GDP. This cannot go on.” And now here and we keep stumbling along.

I do wonder if we’re going to run into a wall with Medicare, in particular. Medicare Trust Fund is now once again almost insolvent. We’re going to have a hard conversation around that. We’ve got a whole bunch of folks my age and younger who are going to be aging into Medicare. And so we’re going to have a reckoning around our health care costs, and we think it’s important to learn some of the lessons from other countries on this.

Two things they do to keep their health care costs in check: one is they spend a greater ratio of their expenses on social care rather than on health care, so very deliberate public policy to focus on retirement, on income security, on unemployment, on daycare, education and other things that reduce downstream health care costs.

And the other thing they do is they have some form of global budgeting. They accomplish it in very different ways. Some do it with single payer health care. Some do it with a government-owned delivery system. Some do it with a series of private insurers. Some do employer-based; some are non-employer based, but they have some form of a diet that they put their health care industry on.

And we, at the Fund, are working very hard to try to apply those lessons, particularly, in the state setting. We’re not going to go to single payer in the next year or two. We’re dealing with Medicare, and Medicaid, and commercial payers. But there are states now that are looking at saying, how do we look at all of our health care expenses, all in, how do we focus on our underlying health care costs, and not on my ability to pay less by making you pay more, which is our time-tested trick here, and how do we measure the rate of growth and set targets for what we want?

And so we, at the Fund, have been working hard on that for the last few years. The project really started this year with states on what we call our sustainable health care costs growth project. And I’d be happy to talk with you about it. We’re working with the Peterson Center on health care. They’re great partners. We’re stealing some lessons from Massachusetts, which was in early on this.

And we think it’s absolutely essential. Otherwise, two things are going to happen. We’re going to spend still more money on health care, which is not the best use for our dollars, socially, and we’re going to become more inequitable because we’re going to keep playing out what we do now, which is that employers unable to afford it are going shift more costs to employees, who are going to withhold and forego important care or give up completely and either be uninsured or go on Medicaid.

Gordon: It just makes me think that to do this you have to get everything onto one ledger. I’m thinking about all the different ways health care costs show up in all these different books. It’s the commercial insurer, and then there’s Workers’ Comp, and then there’s Auto—just on, and on, and on. And I remember hearing some recent pieces about the cost-shifting between all those different players. It’s almost like ping pong. And then unfortunately somebody ends up in medical bankruptcy because they fall through the cracks. How do you get that ledger? How does that work?

Chris: So we think you have to do it at the state, rather than the federal level because a lot of the expenses are at the state level—Medicaid, commercial insurance. We think that you’ve got to bring in data sources from a bunch of different places. We think it’s a public function—not necessarily government perse, but some sort of a public-private commission. We are drawing a lot of our lessons from Massachusetts.

Remember Massachusetts was first out of the gate with what became the Affordable Care Act when they established their MassHealth connector, and a few years after that, they went after costs. And they really pioneered this idea of measuring costs, all in, and then setting a targeted cost growth. I’ll get you data question in a moment.

It’s important to note that they did that as opposed to full on rate setting. They said, “Look, let’s, at least, measure; let’s set targets. Let’s try to hold ourselves accountable, and see if we can socialize the idea of this target, and see what happens, and keep the threat of provider rate setting in our back pocket.” And that’s what they’ve done so far with pretty good success, although, the pressure’s been a little bit harder in the last couple of years.

How does Massachusetts do it? How does anyone do it? You can do this. We track economic growth statistics. We track unemployment. We can track our health care costs. You’ve got to get Medicare data, Medicaid. You’ve got to get commercial. Turns out the biggest problem is self-insured because they are not subject to state regulation, and the administrators won’t hand it over, so you’ve got to negotiate with a whole bunch of individual employers for that.

The Workers’ Comp ends up being pretty small. Those are the big three, if you will: Medicaid, Medicare, and commercial, and then you can split commercial into fully insured and self-insured. If you can get those three together—you need a summary reports—you can get a pretty good measure of what’s going on. The process ends up being really the important. You don’t just hand this to a person sitting around who pops up a number.

It turns out you got to get a whole bunch of people sitting around the table. The process is important to agree on “What do we do with this group? How do we think about the VA? How do we think about people who get their health care across the state lines?” So you have to go through all these structural definitional things. So the process ends up being as important as the technical pieces.

Gordon: And luckily, we can look to Milbank website to actually see some examples of states coming and doing that work. You’ve put out some reports recently about this?

Chris: Yeah. We’re very pleased, as I said, to work in partnership with the Peterson Center on Healthcare. So now we are now working with eight states—if you will—three pioneer states: Massachusetts, Rhode Island, and Delaware, who got into this early. And then with the generous support of the Peterson Center, we have identified five other states that they had to have several ingredients. They had to have statutory or gubernatorial commitment to these health care cost growth targets.

They had to establish a public-private commission that would be convened to look at the baseline measures and to set targets, and they had to commit to keep convening the commission, going forward, to get the data and understand what they would do with it, and so we’re working with Washington, Oregon, Connecticut, New Jersey and Nevada.

So those are the five. It’s a great mix. They’re at different stages. They’re geographically diverse, politically diverse, but they’re committed to this process. We don’t know any other way. We’re not advocating for a particular solution, but as quality-improvement gurus tell us all the time, you can’t improve what you don’t measure. And we measure unemployment; we measure economic growth; we should be measuring our health care costs. So that’s what this effort is about.

Gordon: That’s fantastic. With a few minutes left, if you’ve got the ear of a managed care organization (MCO), or a state Medicaid, or insurance commissioner, what’s your recommendation on moving forward?

Chris: I think you have to educate people on the underlying costs. First thing I did when I was Health Insurance Commissioner is I tried to help people understand that increased insure competition is music to the ears of a dominant health system because they’re dictating prices anyway. So you have to help, what I call, the 80 percent—meaning the 80 percent of the economy that isn’t health care—understand what’s inside the 20 percent.

You have to educate them on price, on increases, and utilization increases, and we’ve made remarkable progress in this country at unpacking the black box, particularly, of commercial provider pricing and understanding how much—as the public payers have started to squeeze down, they’re shifting cost to the commercial sector and how systems are consolidating and using that leverage to extract from employers what they can’t get from the public. And that’s not right. That’s using monopoly power. So regulators who can educate folks on that dynamic I think becomes very important.

And then I think we want to see these cost growth targets that tend to be around GDP. We’re not asking for the sun, the moon, and the stars for health care costs to actually go down for everybody. I think that’s really hard because then you’re laying tons of people off, and we’re not politically ready for that, but if we can just grow at GDP.

And a part of how that happens is that we socialize this. We expect that in our growth rates. We expect that in our commercial growth rates. We’re talking about overall per capita growth rates, and we need to—we know what good unemployment is. We know what a good rate is. We know what a good general rate of inflation. We know what a good GDP growth rate is. We’ve got to start focusing on health care the same way, so it’s something that we watch and that we monitor.

Gordon: Chris Koller, from the Milbank Memorial Fund, thank you so much for your time today.

Chris: It’s my pleasure, Gordon. 

Go to Podcast