Evidence — Statistical result of a PICO question
The elementary grain of a systematic review in FHIR: for a given PICO question, the observed statistical result, its confidence interval and its GRADE certainty.
Purpose of the resource
Evidence carries the formal statistical result of a research question framed
as a PICO (Population, Intervention, Comparator, Outcome) model. It references the
EvidenceVariable resources that define these components, declares the observed
statistic (mean, OR, RR, HR), carries the certainty per GRADE, and
indicates risk of bias.
Key fields
| Field | Type | Cardinality | Role |
|---|---|---|---|
url | uri | 0..1 | Canonical URL. |
identifier | Identifier[] | 0..* | Business identifiers. |
title | string | 0..1 | Title. |
citeAs[x] | Reference(Citation) | markdown | 0..1 | Recommended citation. |
status | code | 1..1 | draft | active | retired | unknown. |
useContext | UsageContext[] | 0..* | Clinical context. |
approvalDate | lastReviewDate | date | 0..1 | Quality cycle. |
publisher | contact | author | editor | reviewer | endorser | string / ContactDetail[] | 0..* | People & roles. |
relatedArtifact | RelatedArtifact[] | 0..* | Related artefacts. |
description | markdown | 0..1 | Description. |
assertion | markdown | 0..1 | Textual assertion. |
variableDefinition | BackboneElement[] | 1..* | Population, Exposure, Outcome (reference to EvidenceVariable). |
synthesisType | CodeableConcept | 0..1 | Statistical synthesis type. |
studyDesign | CodeableConcept[] | 0..* | RCT, cohort, case-control. |
statistic | BackboneElement[] | 0..* | Result (Mean, OR, RR, HR…). |
certainty | BackboneElement[] | 0..* | GRADE certainty. |
JSON example
{
"resourceType": "Evidence",
"id": "example-anticoag-stroke",
"url": "http://example.org/Evidence/anticoag-stroke",
"status": "active",
"useContext": [{
"code": {
"system": "http://terminology.hl7.org/CodeSystem/usage-context-type",
"code": "focus"
},
"valueCodeableConcept": {
"coding": [{
"system": "http://snomed.info/sct",
"code": "266257000",
"display": "Atrial fibrillation"
}]
}
}],
"variableDefinition": [{
"variableRole": {
"coding": [{
"system": "http://terminology.hl7.org/CodeSystem/variable-role",
"code": "outcome"
}]
},
"observed": {
"reference": "EvidenceVariable/stroke-event"
}
}],
"statistic": [{
"statisticType": {
"coding": [{
"system": "http://terminology.hl7.org/CodeSystem/statistic-type",
"code": "C44256",
"display": "Mean Difference"
}]
},
"quantity": {
"value": -0.47,
"unit": "RR"
}
}],
"certainty": [{
"rating": {
"coding": [{
"system": "http://terminology.hl7.org/CodeSystem/certainty-rating",
"code": "high",
"display": "High quality"
}]
}
}]
}Linkage with GRADE
The certainty field is aligned with the GRADE methodology:
high— strong evidence, unlikely to change.moderate— moderate evidence.low— limited evidence.very-low— very limited evidence.
Domains (Risk of Bias, Consistency, Precision…) are carried by
certainty.rating.subcomponent.
Common pitfalls
- Statistic without unit — RR/OR/HR must be explicit.
- Incomplete variableDefinition — at minimum, Population and Outcome are required for a valid PICO question.
- Certainty without rationale — prefer to decompose into GRADE sub-domains.
Related resources
- EvidenceReport — aggregated report.
- EvidenceVariable — PICO variable.
- Citation.
- Measure — analogous quality measure.
- FHIR R5 hub.