FHIR data flowing through EHRs supports a lot more than the aspirational "unified patient record" narrative suggests. Seven specific use cases show up in production 2026 deployments.
1. Population health analytics via bulk export. Bulk Data IG $export feeds Spark/dbt/Databricks pipelines for quality reporting and risk stratification. This is now table-stakes for CMS-0057 attestation.
2. Real-time clinical decision support via CDS Hooks. CDS Hooks fires context-aware prompts at patient-view, order-select, and order-sign workflow points. FHIR resources ride in the hook payload; recommendations return as cards.
3. Patient-facing FHIR apps via SMART launch. SMART on FHIR apps launched from the patient portal give patients read access to their own records. The USCDI data set is the minimum expected coverage.
4. Provider-to-provider clinical data exchange. FHIR-based referrals, transitions of care, and consult requests happen via Bundle transactions between institutions.
5. Payer-to-provider data exchange under CMS-0057. Da Vinci PDex IG defines the request-response flow. Providers submit patient reference, payers return coverage + clinical data.
6. Machine learning feature engineering. CQL expresses clinical logic on FHIR data declaratively. Feature stores draw from FHIR-backed pipelines.
7. Public health case reporting via eCR. eCase Reports use FHIR Bundles to submit reportable conditions to public health agencies.
Adoption maturity, mid-2026
| Use case | Maturity | Vendor support |
|---|---|---|
| Bulk export analytics | Widespread | All top 4 servers |
| CDS Hooks | Growing | Epic, Cerner, Aidbox |
| SMART patient apps | Widespread | All top 4 |
| Provider-provider exchange | Regional | Direct + FHIR bridges |
| CMS-0057 Provider Access | Table-stakes 2026 | Payer-driven |
| ML feature stores | Emerging | Custom pipelines |
| eCR | Widespread | State/CDC-driven |
FHIR is well past the "pilot" stage in 2026. The use cases above ship real value; teams should pick which they need and design accordingly.