Podcast Episode Transcript: How payment reforms address social determinants of health

With L. Gordon Moore, MD

Gordon Moore: Welcome to 3M’s Inside Angle podcast. This is your host, Gordon Moore. With me today is Dr. William Bleser, PhD He’s Managing Associate Payment Reform and Population Health at the Duke Margolis Center for Health Policy at Duke University in Washington, DC. Welcome Dr. Bleser.

William Bleser: Thanks, Dr. Moore. I appreciate it. I’m looking forward to chatting today about this super important topic.

Gordon: Thank you for coming in. The reason I thought it would be interesting is that you and a number of coauthors have put out an issue brief this last February from the Duke Margolis Center for Health Policy and the Milbank Memorial Fund talking about how payment reforms are addressing social determinants of health, policy implications, and next steps.

That to me is interesting. It’s a hot topic; a lot of people are talking about it; some health plans and states are acting on it, and you and coauthors looked at what they were doing, what’s working, what are the challenges and I’d love to dig into that.

Tell me, how did this come about and where did you guys go with your study?

William: When you think about the future of the health system and the future of health delivery in the US and how we pay for health, there’s a couple of topics that come up frequently as being the next frontier, per se.

One of those topics, which comes up in the context of value-based payment, which has greatly expanded over the past ten years, largely thanks to the Affordable Care Act and a lot of good work by public and private payers. One of the topics that comes up through value-based payment a lot is these models are fairly flexible in the services you can provide, because they’re less tied to reimbursing specific services. So, there’s an interest in—we know that upstream social needs and social determinants of health have a huge impact on people’s care.

There’s interest in using the flexibility of those models to bring in new types of clinicians and new types of providers that aren’t clinicians and do new things like address food insecurity; address nonemergency medical transportation; address housing and security and the like.

There’s been some significant work in this space over the past several years and some really big models that are getting ready to begin, so we thought, “Why don’t we do a little bit of the deeper dive into the landscape out there in terms of what’s been done and what’s coming up,” but also, because we like to be practically focused, we wanted to understand, “Okay, so there’s the landscape and there’s new models about to begin, but how do you do this? How do you do it from the payer perspective, from the clinical perspective, from the human services perspective? How do the health sector and the social services sector work together to actually do this on the ground?”

We wanted to dig into the major challenges and the major strategies and lessons learned in order to do that. That was the general impetus behind this work.

Gordon: Where did you go for information?

William: It was a few sources. We did conduct a systematic literature review—not just peer reviewed, but we also looked at—of course we did look in the peer reviewed literature, but we looked at reports; other issue briefs; so called grey literature.

We also supplemented that with a scan of state based websites, since often those are the most up to date and they’ll have things that aren’t published in the literature, as well as some proposed payment models out there that we could find on the web. A variety of sources.

Gordon: In that you were looking for certain things like what were you doing and how did you do it? What problems did you run into and what was the impact? That kind of thing.

William: Exactly. We were focusing on programs that centrally went a step beyond simply screening for social needs, which the vast majority of the clinical side does. Value-based payment models that were delivery models that were more directly tied to going down screening, whether it was requiring that they work with certain community based organizations to address the social need indirectly. Maybe they said, “You can do this and we’ll give you a bonus for it.”

Or maybe models like what’s about to roll out of North Carolina, where it’s explicitly part of what they have to do. We were looking at those types of things and the information we were trying to understand were what you hinted at. What are some of the challenges in doing this, in implementing it? For example, challenges around cross sectional collaboration; challenges around data and technology; challenges around adjusting for social risk factors, in addition to how to finance these things. How do you move to a more sustainable model?

I will say that we did—a lot of our information is drawn from the Medicaid world. There’s a couple of reasons for that. One would be because a lot of the information is publicly reported, we have a better sense of it, but also the Medicaid population, just by definition, has a significantly larger number of marginalized individuals with higher unmet social need, so there’s also a greater demand for it in that space. We did draw our findings from some commercial efforts and Medicare Advantage, for example, but a lot of it was from the Medicaid world.

Gordon: Give me an example of the kinds of programs you saw. You hinted at North Carolina. Are they an exemplar of what’s happening across the country or are they an outlier?

William: That’s a good point and a good question. I would say that North Carolina is the exemplar moving forward. There has been a lot of really good work in this space, for example, in Oregon. They have this coordinated care organization model and they have for a number of years.

Essentially, what they can do is provide what they define as health related services for their enrollees, which could be something like food and housing supports. The second phase of that model, which they’re getting ready to roll out, will actually require these coordinated care organizations to spend part of their annual surplus on addressing health inequities and they outlaid some specific domains of what that means.

The North Carolina model, I would say, is the exemplar—perhaps a positive outlier—looking forward. The basic idea there is North Carolina has an 1115 Medicaid waiver program, so the general goal of that program is to transition the state to a managed Medicaid model. Most states have that model, but North Carolina hadn’t yet.

As part of that model, they’re trying to also shift to more value-based approaches. Part of the waiver program also includes creating some Medicaid issues, creating some advanced medical homes, and one specific pilot, which is really the exemplar program that we’re focusing on. It’s what they call, “The health opportunities pilot.”

They have not yet announced who exactly these pilot entities are going to be, but they did say that they were going to do it this month. Hopefully, it means in the next week or so. Essentially, they’re going to, in two to three regions of the state—they put out an RFP soliciting basically the Medicaid MCOs, which they called “prepaid health plans” to contract with clinical groups—what they’re referring to as lead pilot entities, who then will manage networks of community based organizations.

These lead pilot entities, which is what the applicants are called, they’re going to select two to three of them and they will be required, during the remainder of the waiver period, to specifically address four areas of social determinants of health. It’s fairly specific. They have a long list of a fee schedule, essentially, that will allow them to pay for things like healthy meal boxes; certain legal supports like legal counseling; certain housing supports like helping with first month’s rent. There’s a whole long list, so in that sense, it’s pretty impressive in that they’re requiring that and they’re going to evaluate it, as well.

The primary mechanism through which that model works is the managed Medicaid care organizations, which again, are called “prepaid health plans” in North Carolina. Too many acronyms always a problem. As part of their negotiated fee with the state, they’re going to have to work within that to address all these social needs and they’re going to be contracting with networks of community based organizations to provide those services.

It’s possible the clinical side of the entity may do a lot of direct provision or they may work with one of these community based organizations to pay for it and have the community based organization provide it.

Anyway, long story short to say that it is unprecedented. It’s slated to begin this summer. There’s going to be a capacity building period, as well, essentially as part of the waiver program. The CMS, the federal government, and the state have committed $650 million to this program for reference. So, fairly significant and roughly $100 million of that is going to be dedicated to simply building capacity. That could be standing some of these new legal entities up; hiring new staff; managing and creating these networks of community based organizations, etcetera.

The remaining $500 plus million will go toward service provision for the remainder of this waiver period, which is another four years or so.

Lots to watch there and I’ll stop there for right now, but a very exciting—we actually have a separate grant to study the implementation of that, which we’re excited about. Lots to look out for in North Carolina.

Gordon: You were saying all the different labels we have on these. I’m wishing we had some standardization of labels. There’re so many—each state chooses its own nomenclature for this thing. I’m looking at programs across states, I’m often wondering, “What do you mean by that? Are these hospital systems? Are these managed care plans? Or just some amalgam.” I guess it’s all over the map.

It makes me think about standardization. One of the things I’ve heard about North Carolina is standardization of what data’s going to flow, what screening tools are being used is part of the conversation, at least, which reduces some of the work burden maybe for community based organizations, providers groups and MCOs that are using standard tools. Do you have any insight into that?

William: That’s a good question. One thing we talk about in this issue brief is the challenges of standardized screening tool, standardized evaluation and simply even just how do you coordinate the referrals? Can that process be standardized?

I can speak from the North Carolina perspective again, because I think it’s a pretty interesting example. North Carolina develops as a public private partnership, the state worked with some tech innovators essentially to develop a platform, which is called “NC Care 360.” That program is going to be used for the Healthy Opportunities pilots to handle all the referrals and the accountability of tracking what happens when you’re having a patient who is being referred between the clinical setting and the human services setting.

The general idea of that program is that it’s bidirectional. If a Medicaid patient, for example, is in one of these pilot regions and therefore they could be a Healthy Opportunities pilot participant, during screening, whether it’s from the clinician side or from a human services organization side that’s part of one of these pilot entities—the network of CBOs that I was referring to—if there is a medical or a social need, either side of that can initiate it.

They can log into the system. Perhaps it’s a patient who was already being seen by a community based organization regarding, say, food insecurity and the patient said, “I have a medical problem that’s not managed and I’m having trouble with diabetes,” or whatever it happens to be. They could go into the system and create a referral to one of the clinicians that’s part of this pilot entity.

For example, that could be like, since I’m from Duke and Duke did Duke Health, which we are not affiliated with, in terms of the application, Duke Health put in a proposal for the Healthy Opportunities pilot, so it could be, for example, a patient who is working with a food insecurity community based organization said, “Yeah, I’d like to get more follow-up on this issue. I haven’t initiated contact with the doctor yet.”

The CBO side could refer them to Duke and then, on the other side, which is probably a bit more common, if a Medicaid provider—say, it’s a Duke Health provider—if they end up being selected, if they are screening the patient’s social needs and the patient says, “I’m worried about my housing situation. My current lease is about to expire in a few months and I haven’t found a spot yet.” Or they say, “I’m having trouble securing funds for the first month’s rent and security deposit,” the clinician could go into the system and secure 360 and they would be able to see some of the local CBOs that are part of this big network, find one that addresses housing insecurity, and refer them to that organization.

Then, of course, the Healthy Opportunities pilot is that $550-ish million over the next four years or so is going to be going to help pay for those types of services. The community based organization would be getting money to help with that housing security need, the first month’s rent, etcetera.

That is the tool that they’re using, but you also asked about standardized screening tools. That’s something that the field is actively working on. There is not a standardized screening tool that’s been necessarily recommended by any major source. There are some national efforts worth noting, though, that aim to compare the streamline and to standardize some tools. That can be a starting place, for example, for a payer or a plan that’s interested in delving into a value-based approach to address some social needs.

For example, CMMI, the Centers for Medicare and Medicaid Innovation, otherwise known as the Innovation Center at CMS, has what they call the “Accountable Health Community Screening Tool,” which is used by organizations participating in the accountable health communities model. It is one that you could go to.

There’s also the National Association of Community Health Centers. They actually took the core questions from that CMMI Accountable Health Community Screening Tool and they added in a more expansive set of the social screening measures. That’s one that builds off of the more basic one and adds more. There are efforts out there from which to start, so I wouldn’t think that anyone would be starting from scratch on this, but there’s no consistent tool.

Gordon: One of the things that occurs to me is hearing that screening at the person level is critical if I’m going to know what to do with on behalf of that person’s needs. On the other hand, some of the most marginalized people or communities may not respond or may not be receiving surveys and so, we may be blind to them unless we’re using some sort of geographic tool like the American Community Survey data that identifies social deprivation scores for census tracks, for instance so we know where to target screening outreach. Are you hearing anything about that as you were surveying the country and have any lessons learned from that kind of dual geographic plus person survey approach?

William: There is the CDC developed something called “The Essentially, that uses census variables to try to help local officials identify a geographic assessment of social vulnerability.

I would say that—North Carolina has a—I keep going back to North Carolina, but honestly, I work in Washington, DC, of course, I work for Duke in North Carolina. It’s because there actually is fairly unique things going on there.

One thing I want to mention that’s relevant to your question is, during COVID-19, North Carolina created a program that used their CARES funding that was called the “COVID-19 Social Support” program.” What that program—it’s been colloquially thought of as a preview of the Healthy Opportunities pilots because, what they did was—again, they asked entities to apply for this.

They said, “We have this bucket of CARES funding. We want to use it to help you address a whole bunch of nonmedical things, but it has to be for people who are experiencing social isolation due to COVID-19.” That was the CARES angle. They asked the applicants to essentially put together a network of community based organizations that could address a pretty broad set of things.

For example, this program said, “We will pay for rent assistance, sometimes even back pay for a number of months. We will pay for meal delivery. We will pay for medication delivery, for delivery of cleaning supplies and masks and whatnot.” That is a pretty expansive set of services and one way that they approached, to answer your question, trying to make sure that they were able to screen people, because you have the natural challenges you were suggesting.

If someone is isolated due to COVID-19 or just in the context of the context of the Healthy Opportunities pilots, if they’re a marginalized population with a lot of unmet social needs, they may be harder to find. The one thing that the COVID-19 support services program did in North Carolina was, they used a community health worker approach.

They have a community health worker program, but they said, “For the support services program, you have to work with one of these community health worker networks and the community health workers will help identify patients in the local community that would qualify for and/or isolated for COVID-19 and need social supports.”

That’s one thing that we’re thinking about. I mentioned we had this other grant to help study the implementation of the North Carolina Healthy Opportunities pilots. We’re thinking about that too and we’re going to be asking some of our interviewees, since it’s primarily a qualitative approach, what strategies that they’re using to do some advanced screening.

I agree with you, if you’re just relying on your normal patient base, you’re probably going to miss some eligible folks. Even within your own patient base, it may require some advanced screening and persistent reach outs to try to capture some of these populations. Hopefully, that answered your question. I thought it was a really relevant and timely example.

Gordon: Yes, and now, I want to pivot to the money part, which I think about health care in stepping into a public health space and adding a lot of expense in the way we do things typically in the health care delivery system in the US. Does this pay for itself? Are all interventions of equal value? Are there certain population sub segments based on illness or anything like that more demonstrating better outcomes with these types of interventions? What did you guys see when you were surveying?

William: It’s a good question. Financial sustainability, if you look at the current evidence, for example, of what is the return on investment for addressing X social need, a lot of the evidence—like a specific intervention; right? Like a grant funded project and it may be like a quasi-randomized trial, in a very specific population. Let’s say, for example, people experiencing homelessness.

Those types of things, it’s great to generate that evidence and we need that evidence, but if it’s something that’s time limited if it’s a grant, there are concerns of sustainability and oftentimes, those are smaller populations anyway.

I think in terms of financial sustainability, this is one thing we talked about in the brief. Going through a Medicaid program, for example, could help make things more sustainable. With the 1115 waiver process, which again, states can apply to essentially change their Medicaid program. They work with the federal government—the waiver terminology means they may be waiving certain requirements, but it’s only in the sense that they might be adding a whole bunch of new things and the federal government will only sign off on it if they feel that it’s achieving the core goals of Medicaid and goes beyond that.

The waiver process, depending on how it’s designed, will often come along with some extra support. That’s the case with North Carolina. If you go through a Medicaid approach, you are, in a sense, making it more sustainable because then, your whole entire state is saying, “We’re going to work with—“ they may be a state that directly feeds providers or they maybe work with managed Medicaid entities. That’s sort of a way to make this a little more sustainable, especially if you tie it to performance in some way so then, the organizations have to, in order to pass some sort of benchmark so that they don’t generate financial losses, they’re strongly incentivized to think through it and to do well and to build the infrastructure to do well, which helps influence future sustainability.

I think my comment on financing goes back to the next step here is for major payers to start working through these accountable models to do this with all of their groups instead of these one-off interventions.

The other bonus to doing it through a Medicaid program, for example, or through a larger commercial payer program, is the evidence generation then becomes not just a one-off intervention, but it becomes, “Our three large regions in our state did this for five years,” let’s say, instead of a 12-month intervention, “And it was all Medicaid patients in those regions. It wasn’t just 100 people experiencing homelessness. Here’s what the cost saving looks like.”

We can have a whole separate conversation on evaluation and certainly trying to capture—we learned from the Camden Coalition, for example, if you are too narrow with how you define success or if you’re just focusing on health sector savings or hospitalization reduction, for example, you may not capture the vast majority of the good impacts of these programs. There’s a whole conversation, which we talk about in the paper and which we’re talking with our stakeholders about how to evaluate this.

The CMS evaluation of this North Carolina program, they have their metrics built in and they’re going to be looking at how much money did it save to the Medicaid program after benchmarks are taken account of, but a lot of our stakeholders were saying, “That’s great, however, what about, let’s just say we make large gains in reducing homelessness, because we introduced all these housing insecurity supports like paying for security deposits.

A lot of the savings then that some of these CBOs are experiencing or that the local housing authority—they’re probably generating a fair amount of savings or other very positive outcomes that aren’t necessarily going to be captured in this health system side or CMS side or state Medicaid side evaluation.”

Even beyond that, even for a program like the Healthy Opportunities pilots, which will be going on for the remainder of this waiver for another four years or so, that’s a little bit longer time period than some of these shorter interventions, but it’s still not that long of a time period, when you’re talking about trying to capture savings in upstream social needs and how that translates downstream.

I think there’s a lot of creative thought to be done on the short term like should we be capturing healthy days, asking patients in the program throughout, what percentage of their days or what was the number of days in the last month that they felt they were healthy. Or asking about their quality of life and, of course, there’s many process measures that could be helpful. Making sure the CBOs that you’re working with, let’s just say a food insecurity community based organization, make sure they’re tracking, “Oh, yeah, we have seen X percent drop in reported food insecurity on a weekly basis.” Or “We’ve been able to enroll 25 percent more people because of the funding that came in from this program and we’ve been able to deliver this many meals and satisfy this much unmet food insecurity needs.”

There’s so much capturing to be done beyond just what the financial performance that the payers will care about and that’s definitely worth talking about.

Gordon: Makes me think I was talking to the Patients Care Intervention Center in Houston, Texas. It’s a group that works with unhoused individuals usually with significant and persistent mental illness and substance abuse disorders and their interventions working on housing and connecting with other community based resources also show a reduction in arrest rates and court dates and jail days.

It makes me think we need to step back and look at a much broader lens of the benefit to society of doing this work well. It gives me some hope that, at the societal level or some grander level than just unnecessary emergency room visits, we’re doing a good thing for all of us through this work. That’s a much bigger view of return on investment.

William: Absolutely, and I think it’s crucial to capture that sort of information. As you said, perhaps even if you’re just looking at financial return on investment, a lot of that is going to be outside of the health system. Then, there’s certainly all these other process measures or quality of life measures or unmet social need measures that aren’t related to financial success, but that are equally as important.

Gordon: I think we need some sort of score not just of healthy days and quality of life at an individual level. I’d love to see a score of community health and how is our community doing. That would be fascinating, as well.

William: Absolutely. I should mention too that, in terms of setting up for success, one thing that we talk about in this paper, which is super important, this is important just for value-based payment in the health system, so nonetheless than when you’re trying to work with networks of community based and human service organizations. There’s a startup cost to setting this up.  

If you want to succeed, you need to make sure that those organizations feel supported to establish the infrastructure and that they have the competencies and the technology, etcetera, to do this. I mentioned in North Carolina and I’m glad that this was part of their plan, what is that? One hundred out of six hundred fifty—roughly a sixth or so of the dollars allocated to this are for capacity building.

That’s really important, especially when you think about a lot of these community based and human service organizations—there’s literature on the major financial solvency challenges that local CBOs face. It’s a sad reality perhaps because our social safety net infrastructure in the US is so chronically underfunded, it’s a reality for a lot of these local CBOs to be facing financial insolvency on a month to month basis.

I remember reading a study that was talking about the percentage of CBOs that fail after one year, for example. It was not a very encouraging number. What happens then, and this is something, yes, you want infrastructure building and you want some upfront funding, you also have to consider though that you’re bringing a lot of money and power to the table—the health system is and the payers—that’s not normally there.

Yes, of course, we want to build more capacity, but we want that to be sustainable. We also have to keep in mind that the ultra-competitive, almost cutthroat competitive financial environment that a lot of these CBOs function in oftentimes, they have to undercut themselves in what they’re proposing so that they can be competitive enough to get funded.

You can imagine how that’s a vicious cycle. They’ve got, for example, paying every staff member at minimum wage because that’s all that they could afford and underestimating how much FTE time you need, how many hours you need to actually accomplish something.

When you start to bring to the table, when the health system and the health payers show up with a ton of money, they need to remember that there is a power imbalance and you really need to try to work with the human services and community based organization side to understand what they actually need and to make sure that they don’t undercut themselves. Acknowledge the power dynamics and the tensions as you’re walking in.

In the literature, for example, we even saw on sustaining community health collaboratives, which is essentially, you could use the term, “community health collaborative,” to describe this North Carolina Healthy Opportunities program. One of the things that’s recommended in that literature for sustainability, a technique is potentially using a third party convener. Somebody who can come in and broker the conversation that’s not part of the health system or part of the community based organization human services side so that they can make sure the power dynamics don’t influence proper messages and realities being expressed to the degree that they need to be.

I wanted to mention that there’s a lot of learnings from that space too on how to do that. But the big point is, yes, capacity building is important; upfront cash and infrastructure building are super important, but make sure that you’re acknowledging power dynamics as you approach this space that’s used to chronic underfunding and open those lines of communication and have shared governance and do that early. Then, you can set yourself up for a more sustainable path forward.

Gordon: That is a perfect recitation of the critical variables for success in doing this work. I want to thank you, Dr. Bleser, for your time and for your insights.

William: And thanks to the 3M and to you, Dr. Moore, for having me. As a closing message, I want to emphasize that this is a new frontier and an exciting frontier for health policy and for the health system, but also remember that the human services side and the social services side, this is their whole life and has been since the beginning of time. It’s working together to try to sustainably improve everyone’s capacity so we can address social needs and health needs at the same time to improve population health. It’s a great moment and there’s a lot of future potential here, but a lot of careful attention needs to be paid.

Gordon: Thank you, so much.

William: Thank you.

Go to Podcast