Community health organizations and SDOH data quality: the trust gap
Community health organizations collect more social determinants of health data than ever, yet most of it fails basic integrity checks. The trust gap between what gets recorded and what can actually inform clinical or policy decisions costs the U.S. healthcare system billions annually and leaves the most vulnerable populations invisible.
Community health organizations screen for food insecurity, housing instability, transportation barriers, and interpersonal violence at unprecedented scale. Yet fewer than 25% of those SDOH screenings make it into a structured, usable format that downstream systems can act on. The data exists. The trust does not.
This is not a collection problem. It is a data integrity problem. And it sits at the center of nearly every failed attempt to connect social needs with clinical interventions.
The scale of the SDOH data quality gap
The Centers for Medicare and Medicaid Services now require SDOH screening through multiple programs, including the CMS ACCESS program and accountable care models. AHRQ publishes social determinants of health survey instruments. The AMA has stated that social determinants drive 80% of patient outcomes.
But the infrastructure to verify, standardize, and score that data barely exists.
Community health workers collect information on paper forms, tablets, and through verbal interviews. That information lands in dozens of incompatible EHR systems, community information exchanges, and flat files. By the time a health plan or clinical team tries to use it, critical fields are missing, outdated, or contradicted by other records.
Why community health data trust breaks down
The trust gap has specific, measurable causes. First, provenance is unclear. When a screening result appears in a health information exchange, there is rarely a chain of custody showing who collected it, when, under what conditions, and whether the patient consented to its secondary use.
Second, recency degrades fast. A housing status recorded six months ago may bear no resemblance to a patient's current situation. SDOH data has a shorter shelf life than most clinical data, yet it is rarely timestamped with refresh expectations.
Third, concordance fails across sources. One system codes a patient as "food insecure" based on a two-question screener. Another codes the same patient as "no food insecurity" based on a different instrument administered three weeks later. Without a reconciliation layer, both records persist as equally valid.
Fourth, consent is ambiguous. Patients who answer screening questions in a community clinic may not understand that their responses could flow to insurers, researchers, or government agencies. The consent chain is often incomplete or undocumented, which creates legal risk and erodes patient trust.
What an SDOH index should actually measure
Most SDOH indices aggregate population-level data from the Census Bureau, USDA food desert maps, and EPA environmental databases. These are useful for geographic targeting. They are not useful for individual-level care decisions.
A meaningful SDOH data quality framework needs to score individual records, not just populations. SuperTruth's Data Trust Index scores every health data record from 0 to 100 across eight dimensions: Provenance (25%), Consent (20%), Recency (15%), Quality (10%), Concordance (10%), Validation (10%), Breadth (5%), and Stability (5%). Think of it as a FICO score for health data.
When applied to SDOH records specifically, this scoring model exposes which records are actionable and which are noise. A housing instability screening from a verified community health worker, collected within 30 days, with documented consent, scores fundamentally differently than an unattributed Z-code in a claims file from 2022.
Key statistics
Four numbers frame the urgency of SDOH data integrity:
How geographic SDOH scoring changes the equation
SuperTruth's DataSpine product applies trust scoring to geographic SDOH data, connecting census-tract-level indicators with individual patient records. This creates a bridge between population-level indices (the kind AHRQ and CDC publish) and record-level integrity scoring.
For community health organizations, this means their screening data can be validated against geographic baselines. If a patient in a census tract with a 40% food insecurity rate screens as food secure, that is not necessarily wrong. But it is a concordance signal worth flagging. The reverse is equally true.
This approach does not replace clinical judgment. It gives clinicians and care coordinators a confidence score for the data they are using to make referrals, allocate resources, and report to payers.
The documentation problem no one talks about
Social determinants of health documentation remains one of the most inconsistent practices in American healthcare. ICD-10 Z-codes exist for housing, food, employment, education, and social environment. Most providers do not use them. When they do, the codes are often applied without corresponding screening instrument data.
This creates a phantom dataset: millions of Z-codes floating in claims systems with no underlying source documentation, no consent trail, and no recency indicator. Health plans building risk models on this data are building on sand.
Trust scoring does not fix documentation habits overnight. But it makes the quality gap visible and measurable, which is the prerequisite for any systemic improvement.
What community health organizations can do now
Three immediate steps close the trust gap faster than waiting for national standards:
These are not theoretical recommendations. They are operational changes that DataSpine and the DTI Engine support today.
To explore how trust scoring applies to your community health data, contact Louis Simeonidis, SVP Commercial Operations, at louis@supertruth.ai or (215) 918-4140.
Further reading:
Jason Alan Snyder
Co-founder of SuperTruth and Artists & Robots, and an inventor on the Data Trust Index patents. Twenty-plus years building technology inside Interpublic Group. He writes here nearly every day on data trust, provenance, and what AI should be allowed to act on, and publishes essays on his Substack.
About SuperTruth · LinkedIn · Substack · jasonalansnyder.com
See it in practice
The FICO score for health data.
8 dimensions. 0–100. Travels with every record permanently.