Jeremy Schifberg
We’re intervening with patients who are navigating any one of a number of really complex health and social, just life challenges. Sometimes, we’re intervening in a moment of real personal or acute health crisis, and we really need to earn their trust. And so being local, being able to meet in person, being able to offer a resource, a referral that isn’t something they could just Google on their own, goes a long way to being able to earn that trust and ultimately, earn the right to be able to deliver services and deliver outcomes.
Narrator
That was Jeremy Schifberg talking about the importance of hiring local care teams to earn patient trust. Schifberg is a vice president at Waymark, a public benefit corporation that works with Medicaid-managed care organizations to improve access to care and health outcomes for beneficiaries. The company has teams of community health workers, therapists, pharmacists, and care coordinators. It also works very closely with local clinicians.
In this podcast with Oliver Wyman’s Dan Shellenbarger, Schifberg and Hannah Ratcliffe, Waymark’s Vice President of Care Model and Quality, dive into the opportunities and challenges meeting the needs of Medicaid patients. Ratcliffe shares insights from a peer-reviewed study that published last year in the New England Journal of Medicine showing that Waymark’s interventions reduced emergency department visits and hospitalizations. It also improved access to social services.
The Oliver Wyman Health Podcast is brought to you by the global management consulting firm Oliver Wyman. For more insights on the business of transforming healthcare, visit our online publication, health.oliverwyman.com. Now, we pick things up with Ratcliffe, talking about Waymark’s founding and its mission.
Hannah Ratcliffe
Waymark was founded really to improve access to care and the health outcomes for patients who are receiving Medicaid benefits. And the impetus for this was twofold. There’s direct experience of our founding physicians who have been providing care to low-income patients in FQHCs through street medicine and really seeing the systemic challenges that both patients and providers face to providing and accessing that high-quality care. And at the same time, there’s this huge body of evidence that demonstrates how to reduce acute care utilization, how to improve quality in Medicaid. And yet, so many Medicaid programs are struggling to achieve these outcomes. A lot of that is due to a limited community health workforce, a lack of technology that enables that workforce and really, fundamentally, a lack of sustainable financing to build that workforce and the technology. So many of these programs are grant funded and just can’t sustain momentum over time.
That’s where Waymark is designed to come in. We combine our predictive technology and our multidisciplinary community-based care teams to help Medicaid patients reduce their acute care utilization, improve quality, and, ultimately, improve their health and well-being. The thing that makes our model unique is this combination of community-based care teams who are on the ground, who understand the communities they serve, they’re from those communities, and the fact that they’re guided by a technology that helps them understand who to outreach and when to do that, to ensure that patients aren’t ending up in the hospital or the emergency department. It’s really people enabled by our technology, not just a technology or not just product. And we’re currently serving communities across Washington state and Virginia and have achieved some really strong outcomes over our first two years of service, which we’re excited to dig into.
Dan Shellenbarger
All right, so let’s dig into it. How do your care teams actually operate on the ground? Where are you enhancing and finding great leverage? And I’m curious how that interplay happens with existing providers who are, in some cases, seeing patients and in other cases just trying to keep tabs on patients.
Jeremy
There are a few pieces that make up our care delivery model on the ground. So one is, like Hannah mentioned, we have local multidisciplinary teams. Those teams are made up of community health workers, care coordinators, licensed therapists, pharmacy technicians, and clinical pharmacists. Those local teams wrap around primary care providers and their patients to give additional clinical and social support outside of the clinic. We partner closely with the primary care practices to essentially extend their reach and provide extra capacity to support the Medicaid population. We are focused exclusively on serving the Medicaid population in the local areas in which our teams operate. That’s been a key enabler for our success. It means that our workflows, our technology can all be custom tailor-made for the populations in the communities that we’re serving.
The second piece is we have a hybrid value-based care delivery model leveraging evidence-based care pathways. We use a mix of in-person and virtual services. Our community health worker teams are in community, meeting patients at a hospital, in public settings at the home, accompanying them to visits. And then is supported also by virtual care that the rest of our care delivery team is delivering. That enables us to meet patients where they are. And we find a lot of patients prefer phone or video. We also find that we have higher success rates when we’re able to meet folks in the field, at the home, at the hospital, or in person, in one place or another. And it allows us to do whatever it takes to reach hard-to-reach patients for whom phone or text may not be viable. We operate in value-based contracts.
The last piece is technology that’s tailor-made for our population. One thing that sets our model apart a bit is that we have a custom risk algorithm we call Signal. It’s focused on the rising risk patient population. A lot of off-the-shelf risk stratification algorithms in the market are focused on high-cost claimants, patients who have already been identified as high cost or high need. Our model is focused upstream of that. That allows us to ideally target those patients who maybe are yet to have the avoidable hospitalization or ED visit but may be likely to have one of those kinds of healthcare exacerbations in the near future. Most critically, it means that we’re able to intervene before there’s an exacerbation, to really improve healthcare outcomes. It also means that models focus on high-cost claimants. There’s a natural regression to the mean as it relates to cost savings. By intervening a little bit upstream, we’re able to deliver a higher impact, both clinically and in terms of health outcomes, but also in terms of cost savings for MCO and provider partners.
Dan
Jeremy, talk to me a little bit about how that works. Getting upstream is like the Golden Gleece in healthcare, so the ability to do that is huge. I’m sure it requires EMR integration, data sharing, perhaps even engagement of those members before they become patients, if you will. Talk to me about where you guys find those attachment points and how you’re able to get upstream.
Jeremy
A lot of that comes back to Signal, our rising risk algorithm. The algorithm is trained on a national Medicaid dataset and incorporating social determinants of health data and other community health data. That means that most of the outreach that we’re doing proactively to the patient population is because we’ve identified a patient as someone who’s likely to benefit from our care delivery model. That means we’re not waiting for a referral, although we welcome them, but we’re not waiting for a referral. We’re not waiting for a hospitalization. And so, the vast majority of the patients that we serve and the outreach that we’re doing is focused on that rising risk population. It means that we’re often reaching patients at a moment where they’re not sure why they’re being reached, right? That makes the job a little bit more difficult sometimes for our care delivery teams, in that they have to explain who we are or why we do what we do, how we can help.
Dan
I’m curious about the mix of the interventions you then provide. How much of that is clinical versus how much of that might be more in the realm of social determinants and helping them get ahead of either behaviors or other needs that they might have that are going to lead to those exacerbations? Any thoughts on the balance of clinical and non-clinical needs?
Hannah
What we try to do is be patient centered and so have patients help guide us to what’s most important to them, what they think is the biggest barrier to their health. That often starts out as a more socially related need, whether that’s access to food or housing and security, transportation needs to be able to get around, childcare needs, those types of things. We also do have a set of core clinical pathways focused on common chronic conditions like diabetes, hypertension, or heart failure, that many of these patients… those are conditions many of them are struggling with. And we’re focused on making sure that they’re able to manage their care, advocate for their care, connect into their PCP and the right specialists for ensuring their ongoing support and stability, once we’re no longer working actively with them. We also have therapy teams that can provide behavioral and mental support for patients who are interested in that. And our pharmacy team is obviously doing a lot around medication access.
Dan
Because of the care teams being local, there’s a familiarity with the services, the community context, the organizations that might be able to help contribute to avoiding some of these exacerbations. Maybe just speak for a minute to the role that having that local awareness… what we think that adds to the model?
Jeremy
A big part of what our care teams do, especially upfront, as they establish a relationship with a patient, is try to establish trust and rapport. We’re intervening with patients who are navigating any one of a number of complex health and social, just life challenges. Sometimes, we’re intervening in a moment of real personal or acute health crisis, and we need to earn their trust. And so, being local, being able to meet in person, being able to offer a resource or referral that isn’t something they could just Google on their own, goes a long way to being able to earn that trust and ultimately, earn the right to be able to deliver services and deliver outcomes.
Dan
Let’s get to the results. As you mentioned at the outset, you have been at this early stage, but have enough data now to look back and see what the impact has been. Walk us through what you found and then the results from the New England Journal of Medicine study.
Hannah
Last fall, we published a peer-reviewed study in the New England Journal of Medicine Catalyst journal, evaluating the impact of our model in 2023, which was our first year of service. The study included about 64,000 patients who were covered by the two Medicaid health plans that we were serving in 2023, and those patients were assigned to about 2,300 PCPs across Washington and Virginia. And our study found a couple of key highlight results. The first headline finding was that we achieved a 22.9% reduction in all cause emergency department and hospital visits for patients who were receiving our services compared to a matched non-activated control group. And within that reduction, we saw a nearly 48, 49% reduction in avoidable hospital visits and a 20-ish percent reduction in avoidable ED visits. So those are just really huge results. We’re excited about those.
Another focus of our program is around improving HEDIS quality metrics that we’re contractually responsible for. So not all of them, but a select few that MCOs are really focused on. And we found that we were able to improve seven out of our nine combined HEDIS quality measures by an average of about 12 absolute percentage points for our population in that year. And then I think another thing that speaks to what Jeremy was just talking about in terms of improving trust and meeting patients where they are, we also found that we were able to help patients achieve about 63% of their chosen clinical and social goals.
Dan
I’m also curious about maybe some of the more subjective feedback that you’ve received from patients and the providers with whom they either have been working with or will continue to work with in addition to Waymark. Just the achievement of particular health goals you’ve cited from the patients is awesome, but any other subjective feedback that you’ve gotten around the impact that you’re having on how patients are approaching this?
Hannah
One of the things we’re super proud of, and Jeremy, you can give more color commentary to this, is that our CSAT score customer satisfaction among patients has been in the high eighties to 90% from the very beginning. And Jeremy, maybe you can chip in a little bit more around the provider feedback, but I think in addition to these other outcomes that we talked about, doing that all while patients are pleased with their experience, getting value from their experiences is really important to us too.
Jeremy
On the provider side, I think similar feedback has been positive. Candidly, I think it took us a little bit longer to get there and earn trust. We ran into some natural skepticism. We’re an external company. Everything we’re doing is free to patients and providers. It seems like it should be such a clear win-win. But the reality is that we partner with providers who are passionate about the work they’re doing. They’re serving Medicaid patients for a reason, and we had to earn their trust by showing that we were going to advocate for and support their patients in all the ways that they would themselves. It took a little bit longer, but once they started hearing feedback from their patients about the work that we were doing or just saw the fruits of our labor in our communications with them, that’s what turned the corner. It’s what started to generate a little bit more of a stream of referrals from our provider partners. And so, now, the feedback that we get is really positive, and we’ve gone out of our way to try to flex our model where we can to meet the needs of our provider partners.
I do think that’s a challenge looking forward because the results that Hannah shared were from year one. We had one MCO partner in Washington, one MCO partner in Virginia, and a handful of provider partners in each of those markets. We’re now operating with multiple MCO partners, more provider partners expanding to other markets. We’re figuring out how do we make sure that we’re still flexing to the needs of our provider partners, but within a bounded menu of options so that we’re able to scale in a way that isn’t crazy making for our teams.
Dan
You mentioned some quick wins. Give us some examples of what that means and how that’s worked out for your patients.
Hannah
This is such a critical component to how we work to build trust. And so, again, I think the quick win can vary depending on the patient. We find also that it depends on the care team member and their expertise as they’re doing that outreach, both because the original reason they’re outreaching is changing, but where they’re comfortable making that commitment. And so, what we try to do with a quick win is just find something that seems relevant to the patient that they’re identifying as a high priority, where we can either check that box quickly or at least show a very tangible bit of progress by the next time we encounter them. So if that’s a week later after their intake appointment with us, we want to be able to show that we hear them, and when we say we’re going to do something, we mean it and we’re delivering on those results. That can be something simple like setting up a ride to an upcoming appointment for somebody and making sure that that follow-through happens. It can be something that might be a component of a longer time horizon goal. A lot of our patients have, say, housing goals. That’s a really long-term goal. It’s complicated, but helping them might be, “We’re going to help you parse through this paperwork. We’re going to bring you the things that you need to focus on so that we can get this paperwork filed with the appropriate local housing authority to get you on the right list and start this process off.” And sometimes, it’s, I would say, clinical as well. Making a quick referral to our pharmacy team to address this urgent medication access issue that the patient can’t access something because there’s a prior authorization issue. Our pharmacist team can help resolve that.
Jeremy
And one other place where I think this concept pops up, to build off the last piece that Hannah shared, is having our pharmacy technicians initiate outreach based off a trigger they may see in the claims data or that our risk algorithm may pop up. A patient isn’t filling a critical medication, for example, and using that to get a foot in the door and then introducing the broader care model. So, hey, can I help you overcome this immediate term barrier that we’re seeing around accessing a critical medication? Oh, by the way, we also have community health workers who can help you with X, Y, and Z. We have therapists who can provide free therapy and support you on X, Y, and Z things. I think that’s another way in which this concept of an early win or building trust early shows up in our model as well.
Dan
What’s it going to take for a model like this to scale? It sounds like the results are there, the proof points are starting to show, but as you said, it gets more complicated the more we get involved. What do you guys see as the keys to scaling the model?
Jeremy
I think of, what are the facts on the ground that have to be in place, and then what’s the enabling infrastructure nationally? On the former, we are optimistic because we don’t need any hyper specific policy environment or legislative change. We rely, most critically, on just willing and mission-aligned partners, particularly on the provider side. What that means operationally, is we work closely with our provider partners to set up things like supplemental quality data feeds appointment and EHR access for our teams. In some cases, collaborative practice agreements to make sure that folks like our clinical pharmacist can do things like titrate medicines between visits, to get the most out of our clinical care delivery teams.
The other piece in terms of facts on the ground is just, and this is something we’re actively continuing to explore, ,what are the contracting structures that we need to have in place to support a broader set of provider partners than when we launched? We found that oftentimes, we would run into minimalist, maybe, value-based contracts, upside only, maybe a given provider some years would do well, some years wouldn’t. But there wasn’t necessarily resourcing or staffing tied to these value-based contracts. Maybe it wasn’t moving the needle meaningfully, but it was boxing out the ability for other value-based arrangements to be structured for that population, for that provider. And so, one of the things that we’ve been exploring in the last year plus, as we contemplate much more significant growth here and in the years ahead, is how do we figure out other contracting structures that allow us to fit in with existing value-based arrangements, but still make sure that our incentives are aligned and our model is able to be delivered to a broader swath of provider partner.
And then the other huge piece is our technology infrastructure. The model is again, fundamentally and necessarily local and human. We think about technology as an enabler of humans delivering care to other humans. Specifically, we have custom-built technology that improves the targeting. So who are we outreaching to? Are we outreaching to the right folks? That’s where our risk algorithm, Signal comes in. You can imagine if a standard and a risk algorithm might identify, say, 30% of the patients that we think would benefit from our model. We published in a peer review journal earlier this year on Signal, our rising risk model. And it has a 90% predictive value in terms of identifying the rising risk patients that are a good fit for our model. That difference, every 10 calls that our team is making, if seven out of 10 of them in a traditional risk ratification model aren’t reaching the folks who we think we can benefit, whereas with our model, only one out of 10 are. That is a massive improvement in terms of efficiency. Especially in a population that’s really hard to reach. Every one of those phone calls is critical that it’s focused on the right patient.
Another area that our technology is a critical enabler of scale is in reducing the time that our teams spend on administrative tasks that takes them away from investing their time in what they do best, which is working with patients. In our case, that looks like automating some of the initial outreach attempts, automating scheduling where we can, having a custom-built CRM that’s tailor-made for our population so that our teams aren’t running upstream and an EHR that isn’t designed for the population that we serve. So that’s both a driver of efficiency and of team satisfaction, frankly.
And then the third and final piece is supporting high-quality care delivery. So all the work that Hannah and her team have done around mapping out evidence-based clinical pathways for our teams to follow, making sure that our technology is nudging our teams in the right direction. So maybe it’s suggesting a clinical pathway based off of what we’re seeing in the claims data or suggesting a CBO or a referral destination based off of the feedback that our team has given in the past about where they’ve been successful referring patients, those kinds of things. I think that national infrastructure on the technology side is also something that we envision helping us scale this to be able to serve more and more patients.
Narrator
The Oliver Wyman Health Podcast is brought to you by the global management consulting firm Oliver Wyman. For more insights on the business of transforming healthcare, visit our online publication, health.oliverwyman.com.
This transcript has been edited for clarity.
