Emory Exec Details Launch of Population Health Collaborative
Emory Healthcare Network, the Georgia-based health system’s clinically integrated network, has announced a new population health collaborative in partnership with Guidehealth, a value-based managed services company. Emory Healthcare Network CEO Patrick Hammond and Guidehealth CEO Sanjay Doddamani, M.D., spoke with Healthcare Innovation about their efforts to improve clinical outcomes.
The new population health collaborative integrates Emory’s primary care service line and affiliate physician network to expand value-based care – a focus on quality and outcomes while containing costs – to more than 350,000 individuals across Georgia.
Healthcare Innovation: Has Emory’s clinically integrated network grown gradually over several years?
Hammond: We started the network back in 2011 and it was very slow growth. But right before COVID, we had about 100,000 attributed lives. In 2019 we were bending the cost trend, and improving quality. Ironically, as healthcare systems were financially struggling during the COVID period, we actually tripled in size. We went from about 100,000 attributed lives to where we are today, about 340,000 attributed lives. So we had massive growth, literally in the middle of COVID, which made it interesting to say the least.
HCI: Well, what do you attribute that rapid growth to?
Hammond: Several things. We went from starting the network with about 146 employed primary care physicians to today we have over 500 primary care physicians, of which about 350 are non-Emory employee primary care physicians that are clinically integrated. We also increased the number of value-based agreements we have. So we have value-based agreements with all the major commercial carriers and in Medicare Advantage product lines.
HCI: Can you talk about what some of the things you will work on that you hope will end up reflected in improvement in value-based care performance?
Hammond: Our health system CEO Dr. Joon Lee basically laid out an objective to us saying he not only wanted us to be maximizing the performance of our current value-based agreements, he really wanted us to accelerate into more two-sided arrangements, especially in the Medicare book of business. So we are looking at how to do that. We thought there were several things that Guidehealth brought to the table. One is to rapidly accelerate. This isn't something that we want to do in three, four, or five years. We wanted to start seeing results of this acceleration in the next 12 months. Guidehealth also brought several resources, including one of the industry-leading population management platforms.
The challenge for most primary care physicians is that they simply don't have the resources to do the outreach and the other things that go with population management. This was another resource that the Guidehealth partnership brought that we were able to really start to roll out is private physicians support.
I was a little concerned that some of our non-employed Emory providers would feel that Emory was sticking its nose in their business, but it actually has been the exact opposite. They have embraced it.
HCI: Sanjay, what are some some strengths and opportunities you saw looking at the Emory clinically integrated network?
Doddamani: Emory has an incredible reputation, not just in Georgia, but nationally, for teaching and research, as well as in its care delivery. But as the network expands, we felt we could be excellent partners in enhancing their technology suite, expanding their reach into the community, especially working with both their independent and employed physicians.
HCI: A lot of times, when we're interviewing folks at clinically integrated networks, they have a mix of say, Epic as the platform of record, but then a plethora of different EHRs among the non-employed physician groups. Does that complicate things?
Doddamani: We work across all payers, and we work across all electronic health records. To overcome some of the interoperability constraints that many of our other enablement competitors confront, we've put our health guides on Arcadia into each of these primary care practices, remotely working to bring not just administrative relief, but coordinate care and, with the right oversight, expand the clinical capabilities to reduce open quality gaps or improve access and network referrals — just managing better coordination, so that's been great to have them working in the EHRs directly. That being said, technology is constantly evolving, so we've also had to stay ahead of that game with integrating into existing technology and at Emory, some of the new technologies that they've adopted.
HCI: Is there an AI or predictive component you’re using to try to get ahead of patients before they get to higher acuity?
Doddamani: Conventionally in population health, it's been an 80/20 rule — focusing on the 20% of the patients who cost over half of all healthcare costs. What we know is that that's only partially true. With very advanced analytics and machine learning algorithms, we really need to find the right interventions for the right patients, and in an omni-channel world, deliver them, not just through conventional means. That means that we're not deploying nurses for 20% of the population. It might be nurses for 5% of the population. It might be medical assistance, supporting coordination for some of the others. But then we can also expand beyond the conventional 20% where we can support, for example, referrals and coordination for over half the patients or quality across the entire population. That's what makes it very exciting for us to look at the overall population across payer mixes, across insurance types, to address those very specific things, whether it's contractual obligations by a payer, or whether it's best practice for primary care.
The first decade in value-based care was really focused on risk adjustment without too much of actual true health outcomes improvement. I'm not talking about Emory; I’m talking about nationally. I think we've come to a moment of reckoning with V 28 exposure that true population health has to be improving quality of care and reducing not just total cost, but actually improving health outcomes. And I think that's what we've uncovered — that collaborating together, what we're seeing is very early movement in quality performance and health outcomes that will continue to evolve as we'll work together.
Hammond: Just to echo what Sanjay says, from Emory's perspective I would have had no support from our physicians and our leadership if it ultimately wasn't about improving the overall quality in care. Yes, there are some things you have to do with coding to make sure it's accurately reflecting the severity. But if I was trying to move this along and within the Emory system, and it wasn't about overall improving the care, it would not have gone anywhere.
Doddamani: We tie any risk adjustment to care plans, and those care plans inform what needs to get done for the patient. So if the patient's acuity is not very severe, that may have far fewer or a different set of clinical best practices, whereas very medically complex patients may have very expanded care plans that need to be followed. When those things are not done, those patients become very tedious to manage within the practice. When they are well performed and the patients well managed, they become a delight to be able to to see, even within the schedules of a primary care physician.
HCI: Are you starting to see alignment on the part of the payers in terms of quality measures, or is it still problematic that each each payer might have its own set of measures that they want your organization to report on?
Hammond: On the Medicare side, you do get a little bit better alignment because of the star ratings. The commercial side is definitely still a challenge. I was meeting with our team the other day, and we had something like 77 different measures. So as we're talking with those payers, we say you've got to work with us on this because, it's too many, so it definitely still exists on the commercial side and that is a focus for us.
