For SDOH Standardization, Gravity Project’s Pull Creates Hope
Social determinants of health (SDOH) has become one of the most buzzworthy phrases in healthcare these days, and for good reason: researchers believe that around 80 percent of a person’s health outcomes is driven by SDOH factors such as food insecurity, housing instability, and transportation. As such, it’s now become common to see collaborations between traditional healthcare players and community-based organizations, with the goal being to piece together patients’ daily experiences with their healthcare system interactions.
At the same time, there are still a slew of challenges that remain around integrating SDOH data into care delivery, and successfully merging that information with other important data points for population health success. One of the biggest barriers in this area is the lack of standardization in what variables define the social determinants of health.
From the 40,000-foot view, the issue is this: before healthcare providers’ workflows can be changed, there must be agreed-upon methods for screening questions, tools used, and identifying the elements that are included in SDOH. From the 1,000-foot view, meanwhile, documenting the effect that social determinants of health have on an individual can be quite difficult to quantify since traditionally, healthcare organizations have not focused on collecting structured, standardized data about socioeconomic factors. Without this level of standardization, providers’ ability to truly identify and assess SDOH needs, share information with other stakeholders, and connect patients with community-based resources, is greatly hampered.
The Centers for Medicare and Medicaid Services (CMS) did create “Z codes”—a subset of ICD-10-CM codes to capture social determinants—in 2016, but in the first two years that diagnostic codes specific to social determinants of health were available in Medicare fee-for-service claims, they were used for only 1.4 percent of the total beneficiary population, according to a January 2020 report released by the federal agency. The most frequently used codes were related to homelessness, social isolation, and troubled domestic relationships, the data showed.
In the private sector, organizations have taken a stab, too. In 2019, a major collaboration was announced between UnitedHealthcare, the nation’s largest private payer, and the American Medical Association (AMA) that would support the creation of nearly two dozen new ICD-10 codes related to SDOH. On the clinical side, of course, physicians use a system of ICD-10 codes to classify and record all diagnoses, symptoms, and medical treatments and procedures. The data model developed by UnitedHealthcare focused on standardizing the capture and processing of SDOH-related information, and officials said the codes will combine clinical data and self-reported SDOH data, which will then trigger referrals to social and government services to address social needs.
Then there’s the work being done by the Gravity Project, a community-led HL7 Fast Healthcare Interoperability Resources (FHIR) Accelerator that grew out of efforts to identify data elements for sharing information about social determinants of health. Founded by the University of California San Francisco (UCSF) Social Interventions Research and Evaluation Network (SIREN) in 2018, Gravity Project consists of over 1,000 healthcare stakeholder participants. In December, the coalition published an implementation and recommendation guide for SDOH data and terminology, with a focus on food insecurity, housing instability and homelessness, and transportation access. As a collective, the Gravity Project created recommendations for the necessary context, and for each domain, some are already existing in the U.S. terminology and some are not, explains Sarah DeSilvey, SDOH clinical informatics director at the Gravity Project. She notes that while food insecurity, housing instability and transportation access were the original focus areas, the next domains to tackle will be financial strain, education, unemployment and Veteran status. Once those domains and recommendations are complete, the next phase is working with terminology standards organizations to build them into code, and that process can take six months up to a year or more, says DeSilvey.
She goes on to explain that the Gravity Project also has small terminology development teams that have subject matter expertise for each domain. “Part of the reason why we’re trusted is we ground everything that we’re doing regarding that domain in peer-reviewed literature. We employ subject matter expertise, and terminology and clinician experts; I am a clinical informaticist myself and I’m a family nurse practitioner. We also use the community to cross-check. So through peer literature review, through analysis of white papers, and through community submissions, we create a representative set that we adjudicate to the subject matter experts,” DeSilvey says.
It’s the experts who are there to clarify if something truly is or isn’t housing instability, for example. This comprehensive process, says DeSilvey, is “necessary in the social determinants space because it’s kind of like the Wild Wild West. There’s lots of hope and goodwill, but it’s not yet sorted. That all happens within Gravity.” Since it’s a little bit of the Wild Wild West, the Gravity Project is actually responsible for defining the domains themselves as well, DeSilvey notes. “So, for food insecurity, housing instability, and homelessness, we have to create our own definitions, separate from the public health literature, for clinical and community applications in order to ground the terminology that we’re creating.”
Everything that’s done at the Gravity Project is open, public and consensus-based. To become a member, that individual or entity just has to essentially say, “I want to become a member,” DeSilvey says. Current participants include patients, payers, community-based organizations, clinicians, federal partners, and health IT vendors. One recently joined member is the Pittsburgh-based integrated healthcare company Highmark Health, which operates a health system as well as a health plan. Speaking to the motivation for joining the initiative, Deborah Donovan, Highmark Health’s vice president, SDOH strategy and operations, says, “For us to be able to truly start to look at the social determinants of our populations, both at the population level as well as at the patient level, the work that Gravity is leading is critical to actually creating the national infrastructure to do that.”
What’s more, Donovan affirms what CMS’ report uncovered about the ICD-10 Z codes: while the ones available are being used sparingly by some, they don’t reflect the full breadth of social issues that are impacting populations. For example, she offers, Highmark Health’s clinicians will often say that transportation services and resources is the one social determinant they would most want to fix, as that’s what their patients need most. “But we don’t have an ICD-10 Z code to capture that. So, without it being built through the work that Gravity is doing, we really have a gap in our ability to truly understand the social risks of our populations,” Donovan asserts.
As leaders of the Gravity Project and others surge forward in their work around SDOH data standardization, they’re also keeping their eyes on the eventual goal of implementing this data into organizations’ respective electronic health records (EHRs). When could that become a reality? “That’s the million-dollar question,” DeSilvey acknowledges, again noting the timelines for building codes.
In the interim, providers can ramp-up efforts around developing SDOH screening assessments and getting those into EHR systems. Donovan notes that one of Highmark Health’s plans, Highmark Blue Cross Blue Shield, and its partners have come to consensus on a social determinants screening assessment that’s now live at the Pittsburgh-based Allegheny Health Network and is part of the organization’s Epic EHR platform. Across the nation, however, most healthcare professionals still aren’t screening for SDOH at all; a 2019 JAMA Open Network study found that just one-quarter of U.S. hospitals and 16 percent of physician practices self-report screening patients for social determinants of health such as food, housing, transportation, utilities, and interpersonal violence needs.
Ultimately, both Donovan and DeSilvey emphasize that the work they and others connected to the Gravity Project are doing will create momentum in the healthcare sector through value-based program design that “considers social risk as well as clinical risk when we model care coordination payments to providers,” Donovan says. She adds, “There are many ways we can tackle this, but we need the infrastructure first, and Gravity is really being looked at as the source that we need to funnel all this through.”