Counting What Counts: Standardizing the Data Behind the Social Determinants of Health

The thing everyone agrees matters and almost no one records well

There is broad agreement that the conditions in which people live, work, and age, the social determinants of health, explain a large share of health outcomes, often more than clinical care itself. Housing, food security, transportation, income, and social connection shape who gets sick and who recovers. The agreement is near-universal. And yet the health system records this information so inconsistently that it can rarely act on it at scale.

This primer is about that gap between recognition and record. The argument is that the social determinants will not be addressed systematically until the data describing them is captured in a standardized, shareable form, that serious work is underway to make that possible, and that the project raises a governance question the technical work alone cannot answer: who holds this deeply personal information, and to what end.

Why the data is so hard to use

The problem is not a shortage of concern. It is a shortage of consistency. Different organizations screen for social needs using different questionnaires, record the answers in different formats, and store them, when they store them at all, as unstructured notes rather than retrievable data. One clinic asks about transportation in a way another never captures. A patient's housing instability is noted in a free-text comment that no analysis can find. The result is that even an institution that genuinely wants to understand the social risks of its population often cannot, because the information was never recorded in a form that adds up.

A telling illustration: clinicians frequently name transportation as the social need they would most want to address, yet for a long time the standard diagnostic code system offered no clean way to capture it. The need was real, recognized, and clinically urgent, and it was effectively invisible to the data. What is not recorded in a usable form cannot be measured, and what cannot be measured cannot be managed across a population.

The work to fix it

A coordinated effort is changing this. The Gravity Project, a community-led initiative working within the widely used HL7 FHIR interoperability standard, has been developing standardized terminology and codes for social determinants, defining how concepts such as food insecurity, housing instability, transportation insecurity, and social isolation should be screened, documented, and exchanged across systems. The aim is to let social-needs data move between health care and social services the way clinical data already moves, and to integrate it into the electronic health record alongside diagnoses and treatments.

Progress is real but uneven, and the honest picture matters. The standardized diagnostic codes that do exist for social needs, the so-called Z codes, remain badly underused; in one recent accounting, only a small fraction of a billion claims included any such code at all. The barriers are practical: clinicians are rarely trained to ask about social circumstances in a way that builds trust, smaller practices lack the staff and tools to document it, and the path from screening to a recorded, reimbursable code is unclear to many providers. Standardization is necessary, and it is far from sufficient on its own.

The governance question underneath

Here the data question becomes a trust question, which is where it gets genuinely hard. Social-determinants data is among the most sensitive information a person can share. Whether someone is housing-insecure, food-insecure, isolated, or under financial strain is intimate, and it can be used to help or to harm. Standardizing and sharing this data multiplies its usefulness, and it multiplies the consequences of mishandling it.

This is why the technical project cannot be separated from a governance project. Capturing social-needs data at scale raises questions that code systems do not answer on their own: who controls the data once it is recorded, who is permitted to see it, how the people it describes are protected from its misuse, and how a community can hold accountable the institutions that hold it. Aggregated and de-identified, this data could guide public health toward the upstream causes of illness. Mishandled, it could expose vulnerable people in the moments they are most exposed. The standard makes the data usable. Governance decides whether it is trustworthy, and the two have to advance together.

What this means for community health infrastructure

Standardizing social-determinants data is not a back-office concern. It is the precondition for treating the social determinants seriously at all, because a health system cannot address what it cannot consistently see. But standardization without governance is an incomplete answer, and arguably an unsafe one. The full task is to capture this information in a shared, usable form and to build the stewardship around it that the people it describes would recognize as fair.

That pairing, usable shared data and accountable governance of it, is the Foundation's central concern applied to its most sensitive case. The social determinants will be managed at the scale of populations only when the data behind them is both standardized enough to act on and governed well enough to trust. Counting what counts is the first step. Holding it responsibly is what makes the counting worth doing.

References

1. Healthcare Innovation. For SDOH Standardization, Gravity Project's Pull Creates Hope. Lack of standardization in screening questions and variables; the transportation Z-code gap as illustration. https://www.hcinnovationgroup.com/population-health-management/social-determinants-of-health/article/21211225/for-sdoh-standardization-gravity-projects-pull-creates-hope

2. Office of the National Coordinator, Interoperability Standards Platform. Social Determinants of Health. Gravity Project submissions to add food insecurity, housing instability, transportation insecurity, social isolation, and stress to U.S. Core Data for Interoperability. https://isp.healthit.gov/uscdi-data-class/social-determinants-health

3. Standardizing social determinants of health data: a proposal for a comprehensive screening tool. Health Affairs Scholar (2024). Heterogeneity in screening challenges interoperability; role of LOINC, SNOMED, FHIR, and Z codes; aggregated de-identified data for public health. https://academic.oup.com/healthaffairsscholar/article/2/12/qxae151/7900047

4. Kodjin. SDOH and the Gravity Project: Standardizing Social Health Data (2026). HL7 FHIR-based standardization; interoperability and privacy challenges in SDOH data sharing. https://kodjin.com/blog/sdoh-and-gravity-project/

5. Radicle Health. Z Codes for Social Determinants of Health (2025). Z codes remain underutilized; only 6.2 million of roughly 1 billion claims included a Z code in 2022; provider training and resource barriers. https://radicle-health.com/blog/z-codes-for-social-determinants-of-health/

6. Civitas Networks for Health. Standardizing Social Determinants of Health Data to Transform Equity. Pilot sites applying Gravity Project tools across diverse care settings with neutral convening support. https://civitasforhealth.org/social-determinants-of-health-data-sharing/