The previous article in this series gave some history of social determinants of health (SDoH) and the importance of addressing them in medicine. This article delves deeper into various aspects of data handling, and who is working in these fields.
Defining SDoH
Lyd Paull-Flores, senior director, healthcare at GCOM, summarizes the social determinants of health in an article:
- Economic stability
- Economic access and quality
- Social community and context
- Healthcare access and quality
- Neighborhood and built environment
A few different organizations have undertaken to categorize SDoH. The most comprehensive initiative, cited by several of my interviewees, is the Gravity project. Dr. Marc Rabner, chief medical officer of the nonprofit CRISP Shared Services (CSS), has served on the technical advisory committee for Gravity. He told me that the standard is mature, but that organizations are adopting it slowly. He also pointed out that there is no requirement or incentive for using the standard.
Jason Rose, CEO of AdhereHealth, also pointed to the Social Vulnerability Index (SVI) measured by the Centers for Disease Control at the ZIP code level.
The ICD standard, used universally to assign diagnostic code for billing purposes, also includes some SDoH-related codes that begin with a Z. Rabner estimates that these codes appear in only 3-4% of patient records, but that their use is growing. Rahul Sharma, CEO of HSBlox, points out that doctors haven’t been trained to use the Z codes, but that training will improve. Dr. Victor Lee, vice president of clinical informatics at Clinical Architecture, says that LOINC also has a smattering of codes related to SDoH.
The Agency for Healthcare Research and Quality (AHRQ) defines SDoH in five categories: social context (a very broad category that includes, for instance discrimination and neighborhood safety), economic context, education, physical infrastructure, and healthcare context (such as access).
A roundtable cohosted by Department of Health and Human services issued a report in December 2019 listing about two dozen data elements (pages 43-44), such as eligibility for social programs, air quality, and exposure to violence. It’s notable that the committee included the climate and environment as factors (page 8). The report also includes a listing of organizations working in this area.
Clara Lambert, director of financial navigation at TailorMed, points out the SDoH issues in rural areas include help with transportation, the need for a hotel room when visiting a medical center in a major urban area, and assistance for gig workers.
Gender and sex identity, long assumed to be a fixed, binary trait, has become bewilderingly complicated over the past decade or two. The modern health information API, FHIR, includes the Gender Harmony taxonomy developed by HL7.
Sharing SDoH
CRISP Shared Services is a nonprofit health data utility (HDU) that integrates data between health information exchanges (HIEs). According to Dr. Marc Rabner, chief medical officer, they are trying to improve on the current situation where a patient fills out a questionnaire about SDoH at their clinician, but the results stay in that institution. Patients get frustrated as different clinicians ask them the same questions over and over.
Community-based organizations that provide social services also have valuable SDoH data about their clients, but don’t get reimbursed for sharing it and therefore lack the incentive to do so.
Rahul Sharma, CEO of HSBlox, says that EHRs have few fields dedicated to SDoH, and that these are usually peripheral issues such as tobacco use. Instead, SDoH data such as food and transportation needs are normally found in free-text notes. The HSBlox platform helps to digitize and use data for healthcare administration, billing, and engagement. They serve both payers and providers, providing each client with a customized platform that incorporates 10 different ontologies (with 8 more on the way), and they capture SDoH data.
There are scattered attempts to share data, such as a platform based on GCOM technology to help the state of Virginia get SDoH data, such as “employment, broadband access, and education level,” to specialists working on opioid addiction. GCOM has also publicized another data sharing initiative in the state of New York.
Gerry Miller, founder and CEO of Cloudticity, tells me that healthcare data can’t cross from one province to another in Canada.
Diana Zuskov, associate vice president for healthcare strategy at LexisNexis Risk Solutions, says that the ONC’s United States Core Data for Interoperability (USCDI) recently published standards for sharing data from healthcare records.
Lee says that these standards are still in draft stage and are not accompanied by incentives for adoption. However, Clinical Architecture is prepared to support them in its Pivot platform, which collects and harmonizes health data.
Clinical Architecture is currently working with the National Association of Community Health Centers (NACHC), which co-developed and co-owns PRAPARE, a national standardized patient risk assessment protocol which is designed to be built into an EHR and to engage patients in assessing and addressing SDoH information. NACHC, a ground level member of The Gravity Project, is a key contributor to the Office of the National Coordinator for Health Information Technology (ONC) on the development and use of USCDI SDoH standards.
Clinical Architecture created value sets for NACHC to map standard terminology to the answers patients give to PRAPARE. Clinical Architecture also did a cross-reference and gap analysis of those value sets to the value sets authored by the Gravity Project, and now make the Gravity Project value sets available through their Symedical subscription portal.
The next article in this series turns to the task of collecting SDoH, which has multiple dimensions.
