ediverse Explore the platform

Spotlight PEPPOL BIS Billing 3.0 The EU e-invoicing mandate is here — France Sept 2026, Belgium Jan 2026, Germany 2025.

EvidenceVariable — PICO component

The formal definition of a clinical research variable: population, exposure, comparator or outcome. Referenced by Evidence to structure a PICO question.

Purpose of the resource

EvidenceVariable is the atomic unit of a research question. An EvidenceVariable can represent a population (adults ≥ 18 years with non-valvular atrial fibrillation), an intervention (apixaban 5 mg twice daily), a comparator (warfarin), or an outcome (major vascular event at 12 months). Each variable carries detailed characteristic — inclusion/exclusion criteria as DataRequirement or TriggerDefinition — and its handling (continuous, dichotomous, ordinal, time-to-event).

Key fields

FieldTypeCardinalityRole
urluri0..1Canonical URL.
identifierIdentifier[]0..*Business identifiers.
versionstring0..1Version.
namestring0..1Technical name.
titlestring0..1Title.
shortTitlestring0..1Short title.
statuscode1..1draft | active | retired | unknown.
actualboolean0..1Characterises an observed (true) or hypothetical (false) population.
descriptionmarkdown0..1Description.
characteristicBackboneElement[]0..*Criteria defining the variable.
handlingcode0..1continuous | dichotomous | ordinal | polychotomous.
categoryBackboneElement[]0..*Variable categories (for nominal variables).

JSON example

json evidencevariable-example.json
{
  "resourceType": "EvidenceVariable",
  "id": "stroke-event",
  "url": "http://example.org/EvidenceVariable/stroke-event",
  "status": "active",
  "name": "StrokeEvent",
  "title": "Stroke event within 12 months",
  "description": "Any stroke event (ischemic or hemorrhagic) within 12 months following randomisation.",
  "handling": "dichotomous",
  "characteristic": [{
    "definitionCodeableConcept": {
      "coding": [{
        "system": "http://snomed.info/sct",
        "code": "230690007",
        "display": "Cerebrovascular accident"
      }]
    }
  }]
}

Common pitfalls

  • Non-discriminating characteristic — always point to DataRequirement or TriggerDefinition.
  • Missing handling — without this information, the variable cannot be consumed by a statistical tool.
  • Population vs outcome confusion — the same EvidenceVariable cannot fill both roles in a single Evidence.