Integrating social determinants of health into health data exchanges

When housing, transportation and other nonclinical data are shared electronically, clinicians can better understand their patients’ needs.

Doctors and health care organizations know the importance of social factors in determining patients’ health outcomes, but they often are frustrated by their inability to securely exchange this information in electronic form and integrate it into patients’ electronic health records.

CyncHealth, which comprises the state health information exchanges (HIEs) for Nebraska and Iowa, began taking steps in 2020 to address that problem. CyncHealth partnered with Unite Us, a technology company that builds networks of health and social service providers, to provide behavioral and social health information on the HIEs. Medical Economics® sat down recently with Jaime Bland, D.N.P., RN, president and CEO of CyncHealth, to learn more about the program. The following transcript was edited for length and clarity.

Medical Economics® (ME): How did this initiative get started? What was the impetus behind it?

Jaime Bland: I started my career as (a registered nurse), so I have a background in care coordination. When I came to the HIE I really wanted to incorporate the data that nurses and care coordinators working in the community need to leverage for better patient outcomes. That led me to onboarding various social care strategies, which is where we began to partner with Unite Us, to really facilitate that whole person picture of health.

ME: Why is it important to include social health data in information exchanged among clinicians?

Bland: We’ve long known that the social care category — housing, transportation, food, utilities — has an impact on health care outcomes. Knowing this information about an individual enables us to inform our clinical decisions when patients are ready to leave the hospital or to make recommendations on nutrition or even look at access to care. These components around social determinants influence how we care for patients as clinicians as well as make decisions about how we discharge patients from hospitals.

ME: How’s it going? Is there any evidence that it’s helped individual patients or on a population level?

Bland: A lot of these platforms are immature. The recent entry into the market of social care platforms, vertical integration, brings that data into the EHR; it really just entered into the United States Core Data for Interoperability, which is what codifies those data standards in the EHR. That was just finalized in 2021in the interoperability rules to include those social care components.

As we’re working through the technology and integration components, I would say that we’re early. This is very similar to how we rolled out computerized order entry
20 years ago, and it’s how we began to understand the wor flows and data standards and the need for extracting information once we enter the information.

I think it’s very much still in the early stages of integration into health care. Merging social care with health care is challenging. We know this just from the manual processes we have in place today. The approach to codify the data standards and build out work flows that meet both health care and community needs is a huge task. We’re two years into it. We’ve learned a lot from it, and we’re adapting and modifying. We brought in United Way and 211 as partners in community engagement, and one of the things we’ve taken away is the community convening piece, which I think fits very nicely with HIEs. We’re a neutral convenor. We’re trying to facilitate interoperability as well as building a longitudinal record that follows the person, so it really made sense for us to take a lead role in convening the community, agnostic to health system and agnostic to vendor, and leverage other community data collection points into one work group.

ME: Who identifies the need for addressing a patient’s behavioral or social health needs in the first place? Is it the primary care provider (PCP)?

Bland: It can come from a number of sources. It can be the PCP or the care manager or even the person checking in a patient at the front desk — for instance, the patient mentions they had a difficult time getting to the appointment because they didn’t have transportation. Those are things we don’t think about as relating to the clinical encounter, but they definitely affect whether someone has access to care or can get from the doctor’s office to the pharmacy or back home. Those are things that we’re bringing in as automated ways to create those referrals and then have that closed loop. Knowing that you referred someone for transportation or utility assistance is one step, but if you know that need was met, the next time you see that individual you have a better sense of how they’re managing their health. That’s what really drew us to a closed loop referral system.

I remember being a nurse and somebody identified a patient need at 4 p.m. on a Friday and thinking, “Who am I going to get to answer the phone at 4 p.m. on a Friday to ensure that this need is met over the weekend?” I would have loved to have a system that sent me an alert that the need was met and I didn’t have to track down someone on Monday morning to check.

So I think that’s very appealing to doctors and care coordinators when sending patients back into the community.

ME: It sounds like what you’re describing is compiling a 360-degree picture of a patient, whereas up to now we’ve only been seeing the clinical aspect.

Bland: Absolutely. That’s where the more expanded longitudinal view of health really comes into play by connecting those episodic touchpoints with health entities and bringing in those social care and community aspects. We all know the studies showing 80% of a person’s health outcomes are based on their social determinants. And that’s why we’re trying to connect the (social determinants) piece for a true health picture and connect the health care components to that longitudinal record as well. This will provide a new picture of what health looks like in a digital way, which is the goal of the longitudinal health record and what we felt was our place as the HIE.

ME: I want to talk more about those small practices in rural areas. How do they find out about services their patients may need?

Bland: That’s one way we are democratizing these applications and where we saw our role as an HIE to provide those applications to providers. In those one- and two-doctor practices, being able to use a system such as this is probably not financially feasible. But at scale we’re able to provide these applications at nominal cost, working with our largest health systems and our state partners and payers to bring not only the social care platform but event notifications and other applications that normally wouldn’t be available financially to providers in small practices and organizations.

ME: Many small communities lack the network of services to help these people. What do doctors in those places do?

Bland: One thing about Unite Nebraska is that myriad resources are available in that repository, so they can look outside of their own communities if they’re lacking resources patients could be referred to.

Additionally, it provides us insight as to where those gaps are. If needs are going unmet, those are data points we can now bring to our state legislators and say these are resources that are constantly being referred and where the need goes unmet, and we have a solid data. It’s really where a lot of these conversations started because we know those gaps exist in behavioral health and housing insecurity and food resources. So this is also a way for us to bring those data to light so we can have those policy conversations and have data-driven information to give policy makers and appropriators so they can direct resources to the appropriate places.

ME: Is securing and transmitting social data any different than how you handle clinical data?

Bland: Who needs to know about social care is different than who needs to know about clinical data. How we present that information in our longitudinal record is “a housing referral was made.” Some of the details about the encounter will be stored differently and will be visible differently to a provider than, say, the person who was providing the service. So those two things are differently represented in our system and that’s where it’s important to have some understanding of clinical protections and privacy laws versus social care. The more that comes into data standards, the more we can apply the same levels of protection as HIPAA (Health Insurance Portability and Accountability Act).

Privacy and security are top of mind. We have many points of security, layers of security, as well as governance. In Nebraska we have the HIT board, which is ensuring we’re meeting compliance standards for security, and it’s all open meetings. The other thing is we have broad participation in compliance and governance committees, and the data requests that come to us are made visible to our data suppliers.

I think there’s a robust governance in place that meets industry standards for sharing of information, HIPAA requirements, looking at the changing policy considerations from the 21st Century Cares Act into the interoperability rules. All those things are part of our organization and our mission around governance and transparency.

ME: What’s been the response from providers?

Bland: This is a new technology, so we have typical patterns of adoption — early adopters, intermediate adopters and the laggards. There are some health care systems with robust value-based care programs. They are the early adopters. They want to codify the work flow, bringing those data back from the community, having those alerts and flags for their care managers.

We’ve had some good uptake in rural counties because they do a lot of phone calls, so this is an automated way for them to send out a broad notification that they have an issue that needs to be addressed or referral that needs to be made without tying up a clinical person making 30 phone calls to find a resource to meet a patient’s needs. But for the most part we’re still working on adoption to get those intermediate adopters and laggards onto the platform.

ME: Did you encounter any resistance from clinicians or institutions? If so, how did you overcome that?

Bland: There’s a lot of persuasion that needs to happen. We’re not replacing anything currently in place. This is a technical enabler. But because it’s new, it’s seen as a threat to established turfs. And that’s where the community convening component is important. It’s a lot of meetings and socializations around how we’re trying to link services together.

There’s always comfort in doing things the way they’ve always been done. But I think if we’re going to make data-driven decisions around human services and truly address cost and quality issues from a state-wide perspective, these are the kinds of systems we need to implement.

Most of what we know about referrals and outcomes is anecdotal because many locations don’t have the ability to link the outcome to the individual. Until we do, we have to rely on aggregate information from claims data and we don’t know whether social and human services in the area actually affect outcomes in a data-based way.