The server must implement $extract to:

1. Receive a completed QuestionnaireResponse.
2. Parse extraction extensions from the Questionnaire.
3. Convert the response into structured FHIR resources.
4. Return a Bundle or a single extracted resource.

Technical expectations:

• Support ObservationExtract as the common extraction path.
• Support DefinitionExtract, TemplateExtract, or StructureMapExtract when required by the Questionnaire design.

Extraction priority:

1. Simple Observation extraction, such as scores or measurements.
2. Complex resource extraction, such as diagnosis or medication data.
3. Calculated fields, such as BMI.
4. Transaction Bundle generation.

Example response format: { "resourceType": "Bundle", "type": "transaction", "entry": [ { "resource": { "resourceType": "Observation", "code": { "coding": [{ "code": "PHQ9", "system": "http://loinc.org" }] }, "value": { "integer": 12 }, "derivedFrom": [{ "reference": "QuestionnaireResponse/qr123" }] } } ] }

This is the final SDC workflow step for importing structured data automatically.

The server must implement $populate to:

1. Receive a pre-population request from a client.
2. Query patient data according to questionnaire.item.definition.
3. Pre-fill answers according to FHIRPath expressions.
4. Return the pre-populated QuestionnaireResponse.

Technical expectations:

• Support a FHIRPath engine.
• Support data lookup for resources such as Patient and Observation.
• Parse extensions such as processingPriority, itemWeight, and calculatedExpression.

Implementation priority:

1. Basic pre-population from definition.
2. FHIRPath evaluation.
3. answerExpression support.
4. Complex expressions and conditions.

Example response format: { "resourceType": "QuestionnaireResponse", "questionnaire": "Questionnaire/q1", "status": "in-progress", "item": [ { "linkId": "patient-name", "answer": [{ "valueString": "[automatically populated patient name]" }] } ] }

The server may support $render to:

• Convert Questionnaire into a renderable format, such as HTML, PDF, or JSON.
• Apply themes or custom presentation rules.
• Support multilingual rendering.

Most clients can render questionnaires themselves, so this operation is optional.