Is this real patient data? Do I need a BAA or DUA?

No real patient data is involved — all records are 100% AI-generated synthetic data. HIPAA does not apply, no Business Associate Agreement is needed, and no Data Use Agreement is required. You can share the data with your team, use it in staging environments, load it into cloud systems, and publish results based on it without any compliance concern.

EHR Implementers Health Informatics FHIR Developers Interface Engineers SMART on FHIR

FHIR R4 Synthetic Patient Data —
Conformant Bundles, Instant Download

Q: Are the FHIR R4 bundles conformant to the specification?

Yes. Every FHIR export passes the HAPI FHIR validator against the R4 (4.0.1) specification. Resources use the correct R4 resource types, proper code system URLs (http://loinc.org, http://snomed.info/sct, http://www.nlm.nih.gov/research/umls/rxnorm), and valid reference chaining between Patient, Encounter, Condition, Observation, MedicationRequest, and AllergyIntolerance resources.

Q: What FHIR resources are included in each bundle?

Each FHIR R4 Bundle includes: Patient (36+ demographic fields, gender identity, preferred language, address), Encounter (facility type, class, status, period), Condition resources (ICD-10-CM coded, clinical status, verification status), Observation resources (LOINC-coded vitals and labs with reference ranges and abnormal flags), MedicationRequest resources (RxNorm coded with dosage and indication), and AllergyIntolerance resources. All resources are properly reference-chained so subject, encounter, and recorder references resolve within the bundle.

Q: Can I use these datasets for SMART on FHIR app development?

Yes. The FHIR R4 bundles are conformant and use proper resource references, which is exactly what SMART on FHIR apps need for development and testing. Load the bundles into a HAPI FHIR server or Azure Health Data Services, point your SMART app at it, and you have a complete test environment without touching real patient data.

Q: Are the LOINC and SNOMED-CT codes valid and current?

Yes. Observations use LOINC 2.76 codes with the correct code system URL (http://loinc.org). Conditions and procedures use SNOMED-CT International edition codes with the correct URL (http://snomed.info/sct). Medications use current RxNorm codes. ICD-10-CM codes are from the 2026 fiscal year release. All code systems are referenced by their canonical URLs as specified in the FHIR R4 specification.

Q: Can I load these records into an Epic or Cerner sandbox environment?

Yes. You have two options: use the FHIR R4 bulk import endpoint if your Epic or Cerner sandbox supports it, or use the HL7 v2.4 ADT^A01 messages via an interface engine to feed the ADT feed. Both formats are included in the Architect tier and above. The data has been used successfully for Epic Sandbox, Cerner Millennium staging, and Athena Health integration testing.

Conformant FHIR R4 Bundles with LOINC and SNOMED-CT coding. HL7 v2.4 ADT messages. C-CDA documents. Test your EHR, FHIR server, or interface engine without touching real patient data.

Download Free Sample → View Pricing

HAPI FHIR validator passes 0 PHI — no BAA required Instant download Commercial use cleared

patient-bundle-example.json — FHIR R4 Patient Resource

// FHIR R4 Bundle entry — Patient resource (excerpt)
{
  "resourceType": "Patient",
  "id": "pt-7f3a9c2e-1d84-4b6f-a231-8e0d5c9f4a71",
  "meta": {
    "profile": ["http://hl7.org/fhir/us/core/StructureDefinition/us-core-patient"]
  },
  "identifier": [{
    "system": "urn:oid:2.16.840.1.113883.4.6",
    "value": "MRN-00847291"
  }],
  "name": [{
    "use": "official",
    "family": "Delacroix",
    "given": ["Marcus", "T."]
  }],
  "gender": "male",
  "birthDate": "1974-03-18",
  "address": [{
    "use": "home",
    "line": ["4821 Northfield Drive"],
    "city": "Columbus",
    "state": "OH",
    "postalCode": "43215",
    "country": "US"
  }],
  "communication": [{
    "language": {
      "coding": [{
        "system": "urn:ietf:bcp:47",
        "code": "en",
        "display": "English"
      }]
    },
    "preferred": true
  }],
  "extension": [{
    "url": "http://hl7.org/fhir/us/core/StructureDefinition/us-core-race",
    "extension": [{
      "url": "ombCategory",
      "valueCoding": {
        "system": "urn:oid:2.16.840.1.113883.6.238",
        "code": "2054-5",
        "display": "Black or African American"
      }
    }]
  }]
}

FHIR R4 Conformant

LOINC + SNOMED-CT

HL7 v2.4 ADT

C-CDA 2.1

19,900+ Records

0 PHI

FHIR R4 Resource Detail

What's in each FHIR R4 bundle

Every synthetic patient export includes six properly reference-chained FHIR R4 resources — enough to power server validation, app development, and migration testing without stub data.

Patient

Patient Resource

36+ demographic fields (name, DOB, gender, address, phone, email)
Gender identity extension (US Core StructureDefinition)
Race and ethnicity extensions (OMB categories)
Preferred language via communication with BCP-47 coding
MRN identifier with proper system OID
US Core Patient profile conformance

Encounter

Encounter Resource

Facility type, service type, and encounter class (AMB, IMP, EMER)
Status lifecycle (planned → in-progress → finished)
Period with precise start and end timestamps
Reason code referencing associated Condition resources
Participant (attending provider) with NPI reference
Hospitalization discharge disposition where applicable

Condition

Condition Resources

ICD-10-CM 2026 coded with SNOMED-CT clinical concept equivalent
Clinical status (active / resolved / recurrence)
Verification status (confirmed / provisional / differential)
Category (encounter-diagnosis vs. problem-list-item)
Onset date or age at diagnosis
Subject and encounter references properly chained

Observation

Observation Resources

LOINC 2.76 coded vitals: BP, HR, RR, temp, SpO2, BMI, weight
LOINC-coded lab results: CBC, BMP, LFTs, HbA1c, lipids
Reference ranges (referenceRange.low/high) per LOINC NORMAL values
Interpretation codes (H/L/A/N abnormal flags per v2 table 0078)
Units of measure using UCUM codes
Subject and encounter references chained to parent resources

MedicationRequest

MedicationRequest Resources

RxNorm-coded medications (current release, canonical URLs)
Dosage instructions: route, frequency, dose quantity
Reason reference linking to corresponding Condition
Requester reference (prescribing practitioner)
Status and intent fields (active / completed / draft)
Days supply and dispense request where applicable

AllergyIntolerance

AllergyIntolerance Resources

SNOMED-CT coded allergens (drugs, foods, environmental)
Clinical status (active / inactive / resolved)
Verification status (confirmed / unconfirmed)
Reaction manifestation with SNOMED-CT coding
Criticality (low / high / unable-to-assess)
Recorded date and recorder reference

Interoperability Formats

All three healthcare interop formats — one dataset

The Architect tier and above delivers FHIR R4, HL7 v2.4, and C-CDA 2.1 from the same underlying patient records — so your team can test every interface without maintaining separate data sets.

FHIR R4

Architect+

R4 (4.0.1) conformant Bundle resources with LOINC, SNOMED-CT, RxNorm, and ICD-10-CM coding. Proper reference chaining throughout. Passes HAPI FHIR validator out of the box.

R4 conformant Bundle resource type
LOINC / SNOMED-CT coded observations and conditions
Proper subject, encounter, recorder references chained
US Core Patient profile extensions
Compatible with HAPI FHIR, Azure Health Data Services, Smile CDR

HL7 v2.x

Architect+

Version 2.4 ADT^A01 messages with all required segments. Special characters properly escaped per HL7 encoding rules. Drop these directly into Mirth Connect or Rhapsody.

v2.4 ADT^A01 message type
MSH, EVN, PID, PV1, IN1, AL1, DG1, OBX segments
Special characters escaped (pipe, caret, tilde, ampersand)
Mirth Connect and Rhapsody compatible
Compatible with Azure Health Data Services HL7 ingestor

C-CDA 2.1

Architect+

Continuity of Care Document (CCD) Release 2.1 with all coded sections and human-readable narrative. Required for Meaningful Use, USCDI, and 21st Century Cures Act compliance testing.

CCD Release 2.1 document structure
Coded problem, medication, allergy, and results sections
Human-readable narrative text in every section
LOINC section codes per C-CDA 2.1 templates
Suitable for Direct messaging and EHR transition testing

Use Cases

What teams are building with this data

From FHIR server validation to Da Vinci implementation testing — these datasets remove the bottleneck of finding safe test data.

🔬

FHIR Server Validation & Testing

Load conformant R4 bundles into HAPI FHIR, Smile CDR, or Azure Health Data Services and run your validation suite against known-good data.

📱

SMART on FHIR App Development

Seed a FHIR sandbox with realistic patient data so your SMART app has real resources to query — patients, encounters, meds, labs — from day one.

⚙️

Interface Engine Testing

Replay HL7 v2.4 ADT^A01 messages through Mirth Connect, Rhapsody, or Azure Health Data Services to validate transformations, mappings, and routing.

🏥

EHR Migration Load Testing

Populate Epic, Cerner, or Athena staging environments using FHIR R4 bulk import or HL7 ADT feeds — without risking production patient data.

🤖

CDS Hooks Development

Build and test Clinical Decision Support Hooks against realistic patient contexts — conditions, meds, labs — with proper FHIR resource references.

📋

Da Vinci Project Implementation

Test Da Vinci PDex, ATR, CDEX, and other IG implementations against synthetic patient populations that conform to US Core profiles.

💳

EDI 837/835 Workflow Testing

CSV and JSON exports contain all required fields for EDI claim generation — member ID, payer, diagnosis codes, procedure codes, and rendering provider.

📊

Quality Measure & HEDIS Testing

Build synthetic cohorts with specific condition mixes to test HEDIS measure logic, quality reporting pipelines, and CMS eCQM specifications.

Technical Specifications

Technical specifications

Every version number, code system, and validator reference your integration team will ask for — before they ask.

Specification	Value
FHIR Version	R4 (4.0.1) — HL7 FHIR Release 4
US Core Profile	US Core IG v6.1.0 (Patient, Observation, Condition)
Terminology — Labs & Vitals	LOINC 2.76 (`http://loinc.org`)
Terminology — Clinical	SNOMED-CT International (`http://snomed.info/sct`)
Terminology — Medications	RxNorm current release (`http://www.nlm.nih.gov/research/umls/rxnorm`)
Terminology — Diagnoses	ICD-10-CM FY2026 (`http://hl7.org/fhir/sid/icd-10-cm`)
Units of Measure	UCUM (`http://unitsofmeasure.org`)
FHIR Validation	All bundles pass HAPI FHIR Validator (v6.x)
HL7 Version	v2.4 — ADT^A01 message type
HL7 Segments	MSH, EVN, PID, PV1, IN1, AL1, DG1, OBX
C-CDA Version	CCD Release 2.1 (HL7 C-CDA 2.1)
Average Bundle Size	~45 KB per patient (varies by encounter complexity)
PHI Content	Zero — 100% AI-generated synthetic data
License	Commercial use — no BAA, no DUA, no IRB required

Pricing

Unlock FHIR R4, HL7, and C-CDA

All three interoperability formats are included in the Architect tier and above. Try 5 records free — no credit card required.

Free

5 complete records in CSV. Full demographic and clinical fields — every column that appears in paid tiers.

5 synthetic patient records
CSV format
Full demographic + clinical fields
Instant download — no credit card

Download Free Sample

Architect

$349

1,000 records · all 7 formats

The complete interoperability kit. FHIR R4, HL7 v2.4, C-CDA 2.1, CSV, JSON, SQLite, and Parquet from the same patient cohort.

1,000 synthetic patient records
FHIR R4 bundles — all 6 resource types
HL7 v2.4 ADT^A01 messages
C-CDA 2.1 CCD documents
CSV, JSON, SQLite, Parquet
60+ clinical specialties

Get Architect Tier →

Premium

$599

2,500 records · all 7 formats

Scale up for ML training, HEDIS abstraction, and population-level interface load testing. Same formats, 2.5× the cohort size.

2,500 synthetic patient records
All 7 formats including FHIR R4, HL7, C-CDA
Larger specialty distribution
Ideal for ML model training sets
HEDIS abstraction at scale

Get Premium Tier

Need 10,000+ records, custom specialty mix, or white-label rights? Contact us for Enterprise pricing →

FAQ

Frequently asked questions

Everything your integration team will ask before approving the purchase.

Are the FHIR R4 bundles conformant to the specification?

Yes. Every FHIR export passes the HAPI FHIR Validator against the R4 (4.0.1) specification. Resources use the correct R4 resource types, proper canonical code system URLs (http://loinc.org, http://snomed.info/sct, http://www.nlm.nih.gov/research/umls/rxnorm), and valid reference chaining between all six resource types. US Core Patient profile extensions are included.

What FHIR resources are included in each bundle?

Each FHIR R4 Bundle includes: Patient (36+ demographic fields, gender identity, race/ethnicity extensions, preferred language), Encounter (facility type, class, status, period), Condition resources (ICD-10-CM + SNOMED-CT coded, clinical status, verification), Observation resources (LOINC-coded vitals and labs with reference ranges and abnormal flags), MedicationRequest resources (RxNorm coded, dosage, indication), and AllergyIntolerance resources. All resources are properly reference-chained — subject, encounter, and recorder references resolve within the bundle.

Does the HL7 v2.x output work with Mirth Connect or Rhapsody?

Yes. The HL7 output is v2.4 ADT^A01 format with all standard segments: MSH, EVN, PID, PV1, IN1, AL1, DG1, and OBX. Special characters are properly escaped per the HL7 2.4 encoding rules (pipe, caret, tilde, ampersand). Messages have been tested with Mirth Connect channel configurations and are compatible with Rhapsody, Azure Health Data Services HL7v2 ingestor, and other common integration engines.

Can I use these datasets for SMART on FHIR app development?

Yes. The FHIR R4 bundles are conformant with proper resource references — exactly what SMART on FHIR apps need for development and testing. Load the bundles into a HAPI FHIR server or Azure Health Data Services, configure your SMART app's FHIR endpoint, and you have a complete test environment with realistic patients, encounters, labs, and medications — without touching real patient data.

Are the LOINC and SNOMED-CT codes valid and current?

Yes. All code systems use current releases and correct canonical URLs as specified in the FHIR R4 specification. Observations use LOINC 2.76 codes (http://loinc.org). Conditions and clinical concepts use SNOMED-CT International edition codes (http://snomed.info/sct). Medications use current RxNorm codes. Diagnoses use ICD-10-CM FY2026 codes. Units use UCUM codes (http://unitsofmeasure.org).

Can I load these records into an Epic or Cerner sandbox environment?

Yes — two paths work well. For Epic: use the FHIR R4 bulk data import endpoint in your sandbox instance, or feed the HL7 v2.4 ADT messages through your interface engine. For Cerner Millennium: use the ADT^A01 HL7 feed via an integration engine like Mirth Connect. For Athena Health: the HL7 ADT feed is the most direct path. All three formats (FHIR R4, HL7 v2.4, C-CDA 2.1) are included in the Architect tier and above.

What is a C-CDA document and what's included in each one?

C-CDA (Consolidated Clinical Document Architecture) is the XML-based clinical document standard required for Meaningful Use, USCDI, and 21st Century Cures Act compliance. Each synthetic C-CDA is a Continuity of Care Document (CCD) Release 2.1 with coded sections: problems (active diagnoses), medications (with RxNorm codes), allergies, vital signs (LOINC-coded), laboratory results, and social history. Every section includes both machine-readable coded entries and human-readable narrative text, meeting the dual-use requirement of the C-CDA 2.1 specification.

Is this real patient data? Do I need a BAA or Data Use Agreement?

No real patient data is involved — all records are 100% AI-generated synthetic data. HIPAA does not apply. No Business Associate Agreement (BAA) is needed, no Data Use Agreement (DUA) is required, and no IRB approval is necessary. You can load this data into cloud systems, share it with your team, use it in staging and dev environments, and publish results based on it — with zero compliance overhead.

Ready to test?

FHIR R4 synthetic data. Download in minutes.

Conformant bundles. LOINC + SNOMED-CT. HL7 v2.4. C-CDA 2.1.
No BAA. No real patients. No waiting.

Download Free Sample → View Pricing

HAPI FHIR validator passes 19,900+ records available 14-day refund on Architect tier

FHIR R4 Synthetic Patient Data —Conformant Bundles, Instant Download

What's in each FHIR R4 bundle

Patient Resource

Encounter Resource

Condition Resources

Observation Resources

MedicationRequest Resources

AllergyIntolerance Resources

All three healthcare interop formats — one dataset

FHIR R4

HL7 v2.x

C-CDA 2.1

What teams are building with this data

Technical specifications

Unlock FHIR R4, HL7, and C-CDA

Frequently asked questions

Are the FHIR R4 bundles conformant to the specification?

What FHIR resources are included in each bundle?

Does the HL7 v2.x output work with Mirth Connect or Rhapsody?

Can I use these datasets for SMART on FHIR app development?

Are the LOINC and SNOMED-CT codes valid and current?

Can I load these records into an Epic or Cerner sandbox environment?

What is a C-CDA document and what's included in each one?

Is this real patient data? Do I need a BAA or Data Use Agreement?

FHIR R4 synthetic data. Download in minutes.

FHIR R4 Synthetic Patient Data —
Conformant Bundles, Instant Download