Synthea: Massive FHIR Data
summary
Synthea generates realistic but non‑real patient records by simulating life courses with state‑machine JSON modules, coupling disease progression to US census demographics, real provider locations and CMS cost data. It already models over 540 clinical concepts across ~35 disease modules, outputs per‑patient FHIR bundles (transaction bundles by default) and claims, and can post to a HAPI server or export STU2/STU3. Setup uses Java 8 and Gradle; a million‑patient SyntheticMass snapshot is publicly available.
AI-written summary · about
chapters (28)
- 0:00Introduction
- 1:07What is Synthea
- 2:15Why are we doing this
- 3:55The way the project works
- 7:02Statistics
- 9:32State Machines
- 9:49State Machine Visualization
- 10:06State Types
- 10:53Clinical States
- 11:14Symptoms Drive Care
- 13:07Initial State
- 15:50Distribution
- 16:04Last Days
- 16:22Delay State
- 16:51Scheduling
- 17:29Terminal State
- 18:05Diabetes
- 18:46Opiate Addiction
- 19:30Generating Modules
- 20:30Generating Data
- 21:32Output Types
- 23:16Post to FHIR
- 24:15FHIR Configuration
- 26:47Who is using FHIR
- 28:19Algar exhale
- 29:44Tutorials
- 30:31Roadmap
- 35:53Contact information
related talks
39mA Synthetic FHIR Dataset
21mClinical and Claims Data Transformation utilizing FHIR Bulk Data
19mFrom Zero to Production FHIR Server in 6 Months at a Large Northeast Healthsystem
37mTransducing Healthcare: Transducers as Clinical Features
19mTurning data into better healthcare: data for hemophilia
32mMake EHR Data Available for Research and Public Health (MedMorph)
46mHEDIS on FHIR: NCQA’s digital transformation
33mBridging large-scale FHIR to AI in production