Leveraging social determinants of health data remains challenge
While social determinants—such as socioeconomic status, education, physical environment, and employment—have a significant impact on health outcomes, clinical systems are not currently designed to collect that kind of data.
However, there is a push going on within the healthcare industry to capture that information in electronic health records in order to improve clinical decision support and quality measurement, care coordination, as well as population health management.
On Wednesday, the Office of the National Coordinator for Health IT held a half-day webinar on how to leverage HIT to support the use of social determinants of health (SDOH) in clinical practice.
“Providers serving low-income people and other vulnerable populations for a long time have recognized that factors outside of the traditional bounds of medical care have a large impact on healthcare outcomes,” said Caroline Fichtenberg, managing director of the Social Intervention Research and Evaluation Network (SIREN) at the University of California-San Francisco.
Nonetheless, Fichtenberg contends that little has been done in the longstanding fee-for-service environment to use social determinants data to improve patient care. But, that is beginning to change as the industry transitions to newer payment models based on quality, not quantity, she said.
Toward that end, Fichtenberg noted that earlier this year the Centers for Medicare and Medicaid Services announced that it was funding 32 provider organization sites around the country over five years as part of the Accountable Health Communities Model, which will screen patients for SDOH and refer them to services that address their unmet health-related social needs.
Still, results of a new Deloitte survey of nearly 300 U.S. hospitals and health systems showed that while value-based care is driving better alignment of clinical care and health-related social needs, many healthcare organizations are struggling to fund and build effective capabilities.
“Some hospitals are in the early stages of developing capabilities to address social needs, such as ways to collect relevant data from the community and across the healthcare system, methods to integrate the data and measure performance, and relationships with community organizations that are also addressing social needs in their communities,” states the Deloitte study, which reports that less than one-third of hospitals are integrating social needs into the EHR for most of their target population.
Although EHRs have the potential to provide critical information to providers treating patients with such social needs, until recently SDOH has not been linked to clinical practice. “Further integration of data into the EHR, and strategies for making this data more useful for the care team, may help hospitals in the future,” adds the Deloitte study.
One of the challenges is “access to standardized data that can easily be aggregated across a whole range of systems,” according to Fichtenberg, who made the case that “interoperable social determinants of health data is key” which is being undermined by the lack of a “full complement” of medical terminology codes that capture SDOH data.
“That is a challenge, obviously, for enabling this data to be collected in EHRs,” she added. “Guidance is missing about how to use those codes. Many EHR systems do not yet have the functionality that enables providers to have a place to put this data—although, that is changing day by day. Even once you have a standard set of codes, you still need a lot to make interoperability happen in terms of all of the different standards for the exchange of data.”
Daniel Vreeman, a research scientist at the Regenstrief Institute and the Regenstrief-McDonald Scholar in Data Standards at the Indiana University School of Medicine, emphasized the increasing need and value of representing patient-reported and community-level SDOH data using common clinical vocabulary standards.
“Vocabulary standards can cover that sort of full landscape from things that we’re measuring or observing about genetics to lifestyle variables to environmental factors,” said Vreeman. “And, so, inside a vocabulary standard such as LOINC you have ways to represent each of these different kinds of variables and the assignment of codes and the standard names help you sort out—and helps the computer understand—the important differences between these kinds of variables.”
“Once we get social determinants data represented in data standards, those elements can be shared and understood by diverse IT applications and put to use for the benefit of many in the health ecosystem,” he added.
Yet, the process of identifying and creating the needed structured vocabularies to appropriately represent SDOH domains, questions and other components of screening tools, as well as how to integrate these new data elements into EHR systems, remains a complex task.
However, Michelle Proser, director of research at the National Association of Community Health Centers, pointed to PRAPARE—Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences—an effort to pilot test and promote a national standardized patient risk assessment protocol to assess and address patients’ SDOH.
“PRAPARE is really meant to be more than just a screen tool—it’s a patient engagement tool coupled with a package of implementation and response resources,” said Proser, which includes questions to assess core SDOH domains. “We also intentionally designed it so that it would live in the electronic health record to facilitate assessment and intervention. It’s meant to be a conversation starter and meant to be very patient-centered.”
According to Proser, OCHIN—a nonprofit healthcare innovation center—was one of the pilot sites for PRAPARE which the organization expanded for its own network membership using their Epic EHR system. Rachel Gold, lead research scientist at OCHIN, indicated that it has about 500 community health centers on a single Epic system.
“We took PRAPARE and enriched the EHR functions around the tool, so we’ve been calling our version PRAPARE-plus,” said Gold.
In addition to Epic, Proser said three other EHR templates—eClinicalWorks, GE Centricity, and NextGen—are also available online for free here.
“Together, those four EHRs represent about 60 percent of all health centers nationally—in other words, 60 percent of health centers are using one of those EHRs right now,” added Proser. “We’re also in the process of expanding PRAPARE with other health center network and EHR vendor partners.”
She offered that the other EHR vendors her organization is working with are Allscripts, athenhealth, Cerner, and Greenway Health to develop templates. While health centers form the bulk of PRAPARE users to date, Proser said that interest from hospitals and other provider organizations is growing.
“Making a decision about which screening tool you should use to meet your own needs is an important consideration,” concluded Al Taylor, MD, medical informatics officer in ONC’s Office of Standards and Technology. “Part of it has to do with the fact that some tools are more developed and more capable of being integrated into EHRs.”
At the same time, Taylor cautioned that “there’s more work yet to be done—we’re not really at a point of being interoperable if we only have a collection of codes that are being captured in the EHR.” To achieve interoperability, he added that information must be shared with other providers, payers, and patients.