Secondary Use FHIR Server Implementation Guide
0.1.0 - ci-build

Secondary Use FHIR Server Implementation Guide - Local Development build (v0.1.0) built by the FHIR (HL7® FHIR® Standard) Build Tools. See the Directory of published versions

Synthetic Dataset

Synthetic Dataset

Dataset Types

The server generates named fake datasets:

Dataset Type Planned people Source Patient resources Notes
hospital hospital-focused IBD fixture 240 241 richer procedure and hospital-care pattern
gp huisarts / GP view 270 271 primary-care view with overlap against hospital
default alias 240 241 resolves to hospital

Use --patient-count=<n> to override the configured size for a named dataset.

The default hospital and GP fixtures add one deliberate duplicate Patient row for an existing planned person. An override below the configured duplicate index may omit that row, so the number of Patient resources is then equal to the requested planned-person count.

Persistence

Generated data is persisted below the configured data directory:

data/generated/instance-seed.txt
data/generated/hospital.json
data/generated/gp.json
data/generated/hospital-1000-patients.json

The seed file is created once per data directory. Within one directory, hospital and GP datasets share the seed so their synthetic BSN overlap is deterministic.

A new data directory creates a different synthetic site.

Scenario Model

Every generated patient is based on a deterministic patient scenario.

A scenario includes:

  • IBD type
  • disease severity
  • flare state
  • comorbidity burden
  • medication plan
  • care setting
  • procedure pattern

The generator then emits a patient-centered record set. The data is fake but deliberately less uniform than a hand-written fixture.

Resource Set

The generated record set includes:

  • Patient with fake BSN, name, gender, birth date, telecom, address, GP, and managing organization
  • one IBD Condition
  • optional related and unrelated Condition records
  • Procedure records for hospital patients and a subset of GP patients
  • functional status Observation
  • fecal calprotectin Observation
  • laboratory Observation records for CRP, hemoglobin, ferritin, and vitamin D
  • AllergyIntolerance for a subset of patients
  • MedicationRequest as medication agreement
  • MedicationStatement as medication use
  • one dataset-level Provenance export metadata record

Main Clinical Codes

Diagnosis examples:

Code Display
34000006 Crohn disease
64766004 Ulcerative colitis
24526004 Inflammatory bowel disease

Medication examples:

Code Display
387501005 Mesalazine
372574004 Azathioprine
386891004 Infliximab
407317001 Adalimumab
443465002 Ustekinumab

Procedure examples:

Code Display
73761001 Colonoscopy
1351202006 Ileocolonoscopy
76164006 Biopsy of colon

The fixture labels these as illustrative SNOMED CT-coded clinical concepts. Their terminology status has not been independently validated for clinical use. The fake BSN identifier system is urn:oid:2.16.840.1.113883.2.4.6.3.

Data-Quality Issues

Low-rate deterministic data-quality issues are injected to make adapter and linkage behavior easier to test:

  • missing BSN
  • minor name typo
  • stale address
  • missing telecom
  • birth date off by one day
  • duplicate local patient row for the same person

These are practical fixture assumptions, not national statistics.