To support a US Core Profile, a Server: SHALL Be able to populate all profile data elements that are mandatory and/or flagged as Must Support as defined by that profile’s StructureDefinition.

To support a US Core Profile, a Server: SHOULD declare support for a US Core Profile by including its official URL in the Server’s CapabilityStatement.rest.resource.supportedProfile element.

Servers may deploy and support one or more US Core Profiles to represent clinical information

Server may deploy and support ... the following US Core interaction: “Quick Start” defined for each Profile

Server may deploy and support ... the following US Core interaction: Clinical Notes

Server may deploy and support ... the following US Core interaction: Medication List

Server may deploy and support ... the following US Core interaction: Basic Provenance

Server may deploy and support ... the following US Core interaction: Screening and Assessments

Servers implementing ... [clinical information representation] can claim conformance to the US Core Profile content structure ... by implementing all or parts of the US Core CapabilityStatement into their capabilities.

Servers implementing ... [interactions] can claim conformance to the ... RESTful interactions defined by implementing all or parts of the US Core CapabilityStatement into their capabilities

[To claim conformance to a US Core Profile] a conformant Server:

SHALL Be able to populate all profile data elements that are mandatory and/or flagged as Must Support as defined by that profile’s StructureDefinition.

[To claim conformance to a US Core Profile] a conformant Server:

SHOULD declare conformance with the US Core Server Capability Statement by including its official URL in the Server’s CapabilityStatement.instantiates element: http://hl7.org/fhir/us/core/CapabilityStatement/us-core-Server

[To claim conformance to a US Core Profile] a conformant Server:

SHALL specify the full capability details from the US Core CapabilityStatement it claims to implement.

[To claim conformance to a US Core Profile] a conformant Server:

SHALL… Declare support for the US Core Profile by including its official URL in the Server’s CapabilityStatement.rest.resource.supportedProfile element.

[To claim conformance to a US Core Profile] a conformant Server:

SHALL… Declare support for the US Core Profile’s FHIR RESTful transactions.

Required binding to a ValueSet definition means that one of the codes from the specified ValueSet SHALL be used.

For CodeableConcept, which permits multiple codings and a text element, this [Required binding to a ValueSet definition] rule applies to at least one of the codings

For a [required binding to a ValueSet definition], a CodeableConceptwhich permits multiple codings and a text element … [using] only text is not valid.

[When populating a coded element with a [required binding](http://hl7.org/fhir/R4/terminologies.html#required] to a ValueSet definition] US Core Responders SHALL provide a code exclusively from [the required binding] ValueSet

[Extensible binding] (http://hl7.org/fhir/R4/terminologies.html#extensible) means that one of the codes from the specified value set SHALL be used when an appropriate concept exists;…

[When using an [extensible Binding] (http://hl7.org/fhir/R4/terminologies.html#extensible)] If no suitable code exists in the [extensible] ValueSet, alternate code(s) MAY be used..

For CodeableConcept [with an [extensible binding] (http://hl7.org/fhir/R4/terminologies.html#extensible) … Text-only representation SHOULD NOT be used - coded values are expected whenever possible to support interoperability.

When claiming conformance to ... [to a US Core profile extensible binding rule, a] US Core Responders Shall provide: A code from … [the] valueset 'DataElement.code.code' if the concept exists in the valueset [for a DataElement.code that has an extensible binding] Or an alternative code if the concept does not exist in the valueset [for a DataElement.code that has an extensible binding] Or text in ... `[DataElement.code.text]’ if only text is available [for a DataElement.code that has an extensible binding]

Alternate codes may be provided in addition to the standard codes defined in required or extensible ValueSets. These alternate codes are called "additional codings".

[The additional codings] may be equivalent to or narrower in meaning than the standard concept code.

Not all modifier elements are Mandatory or Must Support, and there is no requirement for supporting them

Servers MAY communicate a system-wide profile in their CapabilityStatement to identify which additional elements, including modifier elements, they support

If the source system does not have data for an element with a minimum cardinality = 0 (including elements labeled Must Support), the data element SHALL be omitted from the resource.

If the data element is a Mandatory element (in other words, where the minimum cardinality is > 0), it SHALL be present even if the source system does not have data.

For [mandatory] non-coded data elements [where data is not available], use the DataAbsentReason Extension in the data type ... [with] the code unknown - The value is expected to exist but is not known.

[In situations where data is not available] for [mandatory] coded data elements … [with] example, preferred, or extensible binding strengths (CodeableConcept or Coding datatypes) If the source system has text but no coded data, only the text element is used.

[In situations where data is not available] for [mandatory] coded data elements… [with] example, preferred, or extensible binding strengths (CodeableConcept or Coding datatypes): For Coding datatypes, the text-only data is represented as a display element

[In situations where data is not available] for [mandatory] coded data elements… [with] example, preferred, or extensible binding strengths (CodeableConcept or Coding datatypes): … If there is neither text nor coded data ... [then] use the appropriate “unknown” concept code from the ValueSet if available.

[In situations where data is not available] for [mandatory] coded data elements… [with] example, preferred, or extensible binding strengths (CodeableConcept or Coding datatypes): … If there is neither text nor coded data ... [then] if the ValueSet does not have the appropriate “unknown” concept code, use unknown from the DataAbsentReason Code System.

[In situations where data is not available] for [mandatory] coded data elements… [with] required binding strength (CodeableConcept or code datatypes): use the appropriate “unknown” concept code from the ValueSet if available

[In situations where data is not available] for [mandatory] coded data elements… [with] required binding strength (CodeableConcept or code datatypes): If the ValueSet does not have the appropriate “unknown” concept code, you must use a concept from the ValueSet. Otherwise, the instance will not be conformant

[If the source system does not have data for a Mandatory element for a coded data element with required binding strength] If any of these status codes is missing [meaning it lacks an "unknown" or otherwise appropriate concept code from the ValueSet, a] 404 HTTP error code and an OperationOutcome SHALL be returned in response to a read transaction on the resource.

[If the source system does not have data for a Mandatory element for a coded data element with required binding strength, and the ValueSet does not have the appropriate "unknown" concept code, then] if returning a response to a search, the problematic resource SHALL be excluded from the search set

[If the source system does not have data for a Mandatory element for a coded data element with required binding strength, and the ValueSet does not have the appropriate "unknown" concept code, then] if returning a response to a search, … a warning OperationOutcome SHOULD be included indicating that other search results were found but could not be compliantly expressed and have been suppressed.

The FHIR RESTful Search API requires that Servers that support search SHALL support the HTTP POST-based search.

For all the supported search interactions in this guide, Servers SHALL also support the GET-based search.

When searching using the token type searchparameter (how to search by token) The Server SHALL support both [system and code values]

When searching using the reference type searchparameter (how to search by reference) The Server SHALL support both [id and Type values]

When searching using the date type searchparameter (how to search by date) The Server SHALL support values precise to the day for elements of datatype date

When searching using the date type searchparameter (how to search by date) The Server SHALL support values precise … to the second + time offset for elements of datatype dateTime

Servers are strongly encouraged to support a query for resources without requiring a status parameter.

If business requirements prohibit [querying a resource without a status parameter], they SHALL follow the guidelines here.

For searches where the Client does not supply a status parameter, an implementation’s business rules may override the FHIR RESTful search expectations and require a status parameter to be provided

For searches where the Client does not supply a status parameter, … systems [that require a status parameter to be provided] are allowed to reject such requests as follows:

SHALL return an HTTP 400 status

For searches where the Client does not supply a status parameter, … systems [that require a status parameter to be provided] are allowed to reject such requests as follows:

SHALL return an OperationOutcome specifying that status(es) must be present.

For searches where the Client does not supply a status parameter, … systems [that require a status parameter to be provided] are allowed to reject such requests as follows:

SHALL support search with status if status required

SHALL NOT restrict search results ( i.e., apply ‘hidden’ filters) when a Client includes status parameters in the query.

[For systems that require a status parameter to be provided,] if a system doesn’t support a specific status code value that is queried, it SHOULD return an HTTP 200 status with a search bundle.

[For systems that require a status parameter to be provided,] if a system doesn’t support a specific status code value that is queried [and returns a search bundle],... [t]he search bundle SHOULD contain resources matching the search criteria

[For systems that require a status parameter to be provided,] if a system doesn’t support a specific status code value that is queried [and returns a search bundle],... [t]he search bundle SHOULD contain ... an OperationOutcome warning the Client which status code value is not supported.

For searches where the Client does not supply a status parameter, … systems [that require a status parameter to be provided] are allowed to reject such requests as follows:...

SHALL document this behavior in its CapabilityStatement for the “search-type” interaction in CapabilityStatement.rest.resource.interaction.documentation.

When an element is Mandatory, the data is expected always to be present.

For querying and reading US Core Profiles, Must Support on any profile data element SHALL be interpreted as follows…:

US Core Responders SHALL be capable of populating all data elements as part of the query results specified by the US Core Server Capability Statement.

SHALL NOT include the data elements in the resource instance returned as part of the query results.

For querying and reading US Core Profiles, Must Support on any profile data element SHALL be interpreted as follows…:

In cases where information on a specific data element is missing, and the US Core Responder knows the precise reason for the absence of data (other than suppressed data), US Core Responders SHOULD send the reason for the missing information.

[When sending reason for missing information, follow] the same methdology outlined in the Missing Data section but using the appropriate reason code instead of unknown [reason code].

Implementors [US Core Responders] seeking ONC certification [in the ONC IT Health Certification program] SHALL interpret Additional USCDI Requirements as Must Support elements as documented above and below;

Implementors [US Core Responders] [not] seeking ONC certification [in the ONC IT Health Certification program] SHALL interpret Additional USCDI Requirements as … optional.

If an element is marked as Must Support and defined by a pattern [as described by ElementDefinition.pattern], then the pattern defines the elements and element values that the Server SHALL be capable of providing.

Primitive elements are single elements with a primitive value. If they are marked as Must Support, then the Server SHALL be capable of providing the element value to meet the Must Support requirement.

[W]hen claiming conformance [to a profile with a must support primitive element] … US Core responders SHALL be capable of providing the value [of the primitive element]

For any complex element marked as Must Support, the Server SHALL be capable of providing at least one of the sub-element values.

If any sub-element is marked as Must Support [for a complex element], it must also meet the Must Support requirements and satisfy the Must Support requirements for the parent element.

[I]f any sub-element is marked as Must Support or Additional USCDI [for a complex element] and the parent element is not, there is no expectation that you must support the parent.

[I]f any sub-element is marked as Must Support [for a complex element] and the parent element is not… [and] the parent element is represented in the structure, Servers SHALL support the sub-element(s) marked as Must Support.

When claiming conformance [to a must support complex element with no must support sub-elements] … US Core Responders SHALL be capable of providing a value in [the] sub-element

When claiming conformance [to a must support complex element with one or more must support sub-elements] … US Core Responders SHALL be capable of providing a value in [each must support sub-element]

Systems [US Core Responders] can support the other elements [of a complex element, not labeled as a Must Support], but this is not a requirement of US Core

The U.S. Core Data for Interoperability (USCDI) may require additional elements, [which is a requirement for certification in the ONC IT Health Certification program, but not a requirement of US Core conformance for US Core Responders]

In certain profiles, only specific resource references are labeled as Must Support.

...

• US Core Responders SHALL be capable of providing [such an element] with a valid reference to [all listed Must Support profile(s).]

Systems [US Core Responders] can support other [resource] references [other than those labeled as Must Support], but this is not a requirement of US Core

In specific profiles, only a single resource reference is present on an element labeled Must Support.

...

• US Core Responders SHALL be capable of providing [such an element] with a valid reference to [the Must Support Profile.]

[If a profile has] a Must Support element [with] a choice of datatypes for its content [and some of] the datatypes … are labeled as Must Support … When claiming conformance to [such a] profile:

• US Core Responders SHALL be capable of populating [the Must Support data type choice]

[If a profile has] a Must Support element [with] a choice of datatypes for its content [and some of] the datatypes … are labeled as Must Support[, or a profile has] an Additional USCDI element [with] a choice of datatypes for its content [and some of] the datatypes … are labeled as Additional USCDI

…

[US Core Responders] MAY support populating ... other [data type] choice elements (such as Observation.effectivePeriod), but this is not a requirement of US Core.

There are several instances in this Guide where there is a choice of supporting one or another profile element to meet the Must Support or Additional USCDI requirement. In such cases, the Server or Certifying System SHALL support at least one element.

​An individual SMART Server will publish a granular list of its capabilities, and a set of these capabilities is combined to support a specific use, a Capability Set. See SMART App Launch’s FHIR OAuth authorization Endpoints and Capabilities for more details. Servers MAY support ... [any] SMART on FHIR Capability Sets and capabilities [(see FHIR OAuth authorization Endpoints and Capabilities)]

​At least one of the following SMART on FHIR Capability Sets SHOULD be supported for US Core Servers that support User-Facing Applications ... Patient Access Standalone Apps Clinician Access for EHR Launch

​For certified systems[, those participating in the ONC IT Health Certification program], both SHALL be supported: Patient Access Standalone Apps Clinician Access for EHR Launch

Implementations [US Core Responders] supporting Backend Services ... SHALL include support for the Client-confidential-asymmetric capability.

Implementations [US Core Responders] supporting Backend Services ... SHALL include support for the ... system/scopes.

​US Core Server[s] SHALL support token introspection defined by the SMART App Launch Guide. For more details and additional consideration, see SMART App Launch’s Token Introspection.

Other systems only need to support scopes for the US Core APIs they support [instead of all US Core's required scopes]

Servers MAY support other scopes in addition to those listed below and in the Quick Start sections.

For “User-Facing Applications”, a system’s support for patient-level (patient) or user-level (user) scopes depends on its published list of SMART on FHIR capabilities (see the capability sets above). For example, if a Server lists permission-patient and permission-user in its capabilities, it SHALL support both patient-level and user-level required scopes

For “User-Facing Applications”, a system’s support for patient-level (patient) or user-level (user) scopes depends on its published list of SMART on FHIR capabilities (see the capability sets above). For example, if a Server lists permission-patient and permission-user in its capabilities, it ... SHOULD support both patient-level and user-level recommended best-practice scopes

For “Backend-Services”, System-level scopes (system) describe data that a Client system is directly authorized to access. Systems that support system-level (system) scopes SHALL support the required US Core scopes

For “Backend-Services”, System-level scopes (system) describe data that a Client system is directly authorized to access. Systems that support system-level (system) scopes SHOULD support the recommended US Core scopes

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] AllergyIntolerance [the] <patient|user|system>/AllergyIntolerance.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] CarePlan [the] <patient|user|system>/CarePlan.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] CareTeam [the] <patient|user|system>/CareTeam.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Condition [the] <patient|user|system>/Condition.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Coverage [the] <patient|user|system>/Coverage.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Device [the] <patient|user|system>/Device.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] DiagnosticReport [the] <patient|user|system>/DiagnosticReport.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] DocumentReference [the] <patient|user|system>/DocumentReference.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Encounter [the] <patient|user|system>/Encounter.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Goal [the] <patient|user|system>/Goal.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Immunization [the] <patient|user|system>/Immunization.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] MedicationDispense [the] <patient|user|system>/MedicationDispense.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] MedicationRequest [the] <patient|user|system>/MedicationRequest.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Organization [the] <patient|user|system>/Organization.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Patient [the] <patient|user|system>/Patient.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Practitioner [the] <patient|user|system>/Practitioner.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] PractitionerRole [the] <patient|user|system>/PractitionerRole.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Procedure [the] <patient|user|system>/Procedure.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Provenance [the] <patient|user|system>/Provenance.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] QuestionnaireResponse [the] <patient|user|system>/QuestionnaireResponse.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] RelatedPerson [the] <patient|user|system>/RelatedPerson.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] ServiceRequest [the] <patient|user|system>/ServiceRequest.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types SHALL be supported ... [For] Specimen [the] <patient|user|system>/Specimen.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types MAY be supported ... [For] Location [the] <patient|user|system>/Location.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following scopes that correspond directly to FHIR resource types MAY be supported ... [For] Medication [the] <patient|user|system>/Medication.rs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Condition [the] <patient|user|system>/Condition.rs?category=http://hl7.org/fhir/us/core/CodeSystem/condition-category|health-concern [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Condition [the] <patient|user|system>/Condition.rs?category=http://terminology.hl7.org/CodeSystem/condition-category|encounter-diagnosis [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Condition [the] <patient|user|system>/Condition.rs?category=http://terminology.hl7.org/CodeSystem/condition-category|problem-list-item [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs?category=http://hl7.org/fhir/us/core/CodeSystem/us-core-category|sdoh [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs?category=http://terminology.hl7.org//CodeSystem-observation-category|social-history [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs?category=http://terminology.hl7.org/CodeSystem/observation-category|laboratory [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs?category=http://terminology.hl7.org/CodeSystem/observation-category|survey [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHALL be supported ... [For] Observation [the] <patient|user|system>/Observation.rs?category=http://terminology.hl7.org/CodeSystem/observation-category|vital-signs [see SMART Clinical Scope Syntax for details on clinical data scopes]

The following granular scopes SHOULD be supported ... [For] DocumentReference [the] <patient|user|system>/DocumentReference.rs?category=http://hl7.org/fhir/us/core/CodeSystem/us-core-documentreference-category|clinical-note [see SMART Clinical Scope Syntax for details on clinical data scopes]

In addition to the capabilities defined in the Server’s CapabilityStatement, Servers SHALL document their SMART capabilities in their Well-Known Uniform Resource Identifiers (URIs) JSON file.

US Core requires ... additional metadata [to be available through the Server's Well-Known Uniform Resource Identifier (URI)]: ... [in] scopes_supported [the] array of scopes a Client may request.... The Server SHALL list all the required US Core Scopes for the US Core Profiles they support in the scopes_supported array

US Core requires ... additional metadata [to be available through the Server's Well-Known Uniform Resource Identifier (URI)]: ... [in] scopes_supported [the] array of scopes a Client may request. The Server [MAY support] additional scopes (so Clients should not consider … [the required scopes] an exhaustive list).

US Core requires ... additional metadata [to be available through the Server's Well-Known Uniform Resource Identifier (URI)]: ... [in] scopes_supported [the] array of scopes a Client may request. Servers MAY limit Clients’ scopes to those configured at registration time.

US Core requires ... additional metadata [to be available through the Server's Well-Known Uniform Resource Identifier (URI)]: ... [in] scopes_supported [the] array of scopes a Client may request. … Servers SHALL allow users to select a subset of the requested scopes at the approval time.

US Core requires ... additional metadata [to be available through the Server's Well-Known Uniform Resource Identifier (URI)]: ... [in] introspection_endpoint the URL to a Server’s introspection endpoint. ... Servers SHALL document this endpoint in the file

Therefore, a reference to a US Core resource SHOULD include a logical id (Reference.reference), not an identifier (Reference.identifier).

For all references, US Core Responders SHOULD return resources that conform to a US Core profile if a US Core profile exists for the resource type.

When responding to a query, Servers SHOULD not use inline contained resources to represent the returned data.

[I]f referencing a contained resource in a US Core Profile, the contained resource SHOULD be a US Core Profile if a US Core Profile exists for the resource type.

[M]asking data [where a specific piece of data is hidden for security and privacy reasons] SHOULD be handled based on implemented policies.

[When masking data] elements with a minimum cardinality = 0 (including elements labeled Must Support) [for security and privacy reasons], the element SHOULD be omitted from the resource.

[When masking] Mandatory elements (in other words, where the minimum cardinality is > 0) [for security and privacy reasons, use the code “unknown” following the guidance on Missing Data in the Conformance Sections.

[When using SNOMED codes in US Core Profiles I]mplementers MAY use the default system URI [of SNOMED CT], which refers to an unspecified edition/version [of SNOMED]

[To enable] terminology Servers to be able to validate US Edition-only codes [of SNOMED CT], implementers SHOULD provide the accompanying system URI to describe the edition [see example 2 on US Core general guidance]

[If a] US Core Profiles binds the Quantity.code element in the Quantity datatype to the UCUM code system, [then] systems should also use UCUM for the optional valueRange and valueRatio datatypes (which are complex datatypes with Quantity elements)

[If a] US Core Profiles binds the Quantity.code element in the Quantity datatype to the UCUM code system, [when] UCUM code [is] provided [it SHOULD be indicated in the Quantity.unit and Quantity.code elements with Quantity.system = "http://unitsofmeasure.org"]

[If a] US Core Profiles binds the Quantity.code element in the Quantity datatype to the UCUM code system, [then] if UCUM units are unavailable, represent units in the unit element

SHOULD NOT supply the unit field]

A FHIR Server SHOULD not delete records.

A FHIR Server SHOULD update the appropriate resource status to entered-in-error or inactive [when requested to delete records]

If a system supports the deletion of records, they SHOULD refer to the Deletion Safety Checks in the FHIR specification.

A FHIR Server SHOULD allow these resources [those entered in error] to be searchable by Client applications.

If the FHIR Server has updated the resource status to entered-in-error: For patient facing applications, A FHIR Server SHOULD remove the resource’s contents, leaving only an id and status. Note that this typically will not conform to the US Core or FHIR StructureDefinitions.

If the FHIR Server has updated the resource status to entered-in-error: ... For provider-facing applications, the content may be supplied with content and additional detail (such as the reason for the status change) that the patient viewing system would typically not have access to.

the data provider SHOULD do their best to translate (safely) to the requested language [when accessing records in a native or requested language]

[When Clients request a resource in a specific language] Servers SHOULD make reasonable efforts to translate what can be safely translated.

[When Clients request a resource in a specific language] Servers SHOULD populate the Resource’s language element with a code based on the underlying language of record, not the requested language.

[When Clients request a resource in a specific language] Servers SHOULD … [use] the Human Language Extension when the language of a display, etc, is known to differ from the stated (or inferred) language.

[When Clients request a resource in a specific language] [Servers SHOULD use] the Translation Extension when the Server provides additional translations by its own choice or in response to a different Accept-Language than what the resource is stored in.

[When Clients request a resource in a specific language] Servers SHOULD make it known what languages are supported in their CapabilityStatement(s) using this extension: http://hl7.org/fhir/5.0/StructureDefinition/extension-CapablilityStatement.acceptLanguage [definition]

Servers SHOULD support the _lastUpdated search parameter for US Core Profiles

Servers ... SHOULD populate Resource.meta.lastUpdated for US Core Profiles as accurately as possible.

Servers SHALL document in CapabilityStatement.rest.resource.searchParam.documentation the types of changes that can be detected using the _lastUpdated search parameter

US Core Servers [MAY support compartment based searchs, but] are not required to support patient compartment based searches.

US Core Servers [MAY resolve absolute URLs, but] are not required to resolve absolute URLs external to their environment.

Servers can [MAY] choose to return the results in a series of pages to manage the number of search results returned.

Servers SHALL support, at minimum, these … "Common Clinical Notes": Consultation Note (11488-4)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Discharge Summary (18842-5)

Servers SHALL support, at minimum, these … "Common Clinical Notes": History & Physical Note (34117-2)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Procedures Note (28570-0)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Progress Note (11506-3)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Imaging Narrative (18748-4)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Laboratory Report Narrative (11502-2)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Pathology Report Narrative (11526-1)

Servers SHALL support, at minimum, these … DiagnosticReport categories: Cardiology (LP29708-2)

Servers SHALL support, at minimum, these … DiagnosticReport categories: Pathology (LP7839-6)

Servers SHALL support, at minimum, these … DiagnosticReport categories: Radiology (LP29684-5)

systems are encouraged to support other common notes types, such as: Referral Note (57133-1)

systems are encouraged to support other common notes types, such as: Nurse Note (34746-8)

To enable consistent access to scanned DiagnosticReport clinical reports, the FHIR Server SHALL expose these overlapping scanned or narrative-only reports through both DiagnosticReport and DocumentReference by representing the same attachment URL [as] DocumentReference.content.attachment.url [and] DiagnosticReport.presentedForm.url

when DiagnosticReport.presentedForm.url references a Scan (PDF), that Attachment SHALL also be accessible through DocumentReference.content.attachment.url

Systems [Servers] may extend their capabilities [around types of clinical notes] to the complete US Core DocumentReference Type Value Set.

This guide requires [Server] systems to implement the US Core DocumentReference Profile

This guide requires [Server] systems to implement the US Core DiagnosticReport Profile for Report and Note exchange

Systems [Servers] may support other [DiagnosticReport] categories as well.

The following SHOULD be exposed via DiagnosticReport: Imaging Narrative

The following SHOULD be exposed via DiagnosticReport: Laboratory Report Narrative

The following SHOULD be exposed via DiagnosticReport: Pathology Report Narrative

The following SHOULD be exposed via DiagnosticReport: Procedure Note

Servers that support DiagnosticReport will include the clinical note narrative content in DiagnosticReport.presentedForm

FHIR Server claiming support to this guide SHOULD support the $expand operation [operation link to provide information to Clients requesting information on the note and report types the Server supports]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.category&contextDirection=outgoing for DiagnosticReport report category discovery [for read operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.category&contextDirection=incoming for DiagnosticReport report category discovery [for write operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.code&contextDirection=outgoing for DiagnosticReport report type discovery [for read operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.code&contextDirection=incoming for DiagnosticReport report type discovery [for write operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.category&contextDirection=outgoing for DocumentReference note category discovery [for read operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.category&contextDirection=incoming for DocumentReference note category discovery [for write operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.type&contextDirection=outgoing for DocumentReference note type discovery [for read operations]

[If Servers support the $expand operation for discovering note and report types, then Servers SHALL support] the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.type&contextDirection=incoming for DocumentReference note type discovery [for write operations]

[If Servers support the $expand operation then] the note and report types for a particular Server [SHOULD be] discovered by invoking the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.presentedForm.contentType&contextDirection=outgoing for DiagnosticReport report content type discovery [for read operations]

[If Servers support the $expand operation then] the note and report types for a particular Server [SHOULD be] discovered by invoking the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-diagnosticreport-note#DiagnosticReport.presentedForm.contentType&contextDirection=incoming for DiagnosticReport report content type discovery [for write operations]

[If Servers support the $expand operation then] the note and report types for a particular Server [SHOULD be] discovered by invoking the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.content.attachment.contentType&contextDirection=outgoing for DocumentReference note content type discovery [for read operations]

[If Servers support the $expand operation then] the note and report types for a particular Server [SHOULD be] discovered by invoking the #expand operation as follows: GET [base]/ValueSet/$expand?context=http://hl7.org/fhir/us/core/StructureDefinition/us-core-documentreference#DocumentReference.content.attachment.contentType&contextDirection=incoming for DocumentReference note content type discovery [for write operations]

[The MedicationAdministration and MedicationStatement] medication resources are not profiled by US Core, and systems that support US Core are permitted to use them

When using a code [to represent a medication][, the code SHALL follow the extensible binding rules to Medication Clinical Drug (RxNorm) - i.e., unless RxNorm does not cover the concept, the RxNorm code SHALL be used.

USCDI recommends the National Drug Codes (NDC) as an optional medication terminology. They can be supplied as an additional coding element [when representing a medication]

When referencing the Medication resource, the resource may be included in the returned bundle, as an external resource, or as a contained resource if it can’t stand alone. … The Server application MAY choose any combination of these methods

if an external reference to Medication is used, the Server SHALL support the _include parameter for searching this element

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: A MedicationRequest resource query SHALL be all that is required to access "all medications" or "all active medications" for a patient. (In other words, no other medication resource type needs to be fetched)

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: The [MedicationRequest resource] query result SHALL include all MedicationRequest resources with a MedicationRequest.intent = "order" representing authorized medication orders directly derived from the system's orders.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: The [MedicationRequest resource] query result SHALL include all prescribed and "self-prescribed" MedicationRequest resources with a MedicationRequest.intent = "plan" representing reported medications.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: Servers SHALL use the MedicationRequest.reported[x] element to indicate that the MedicationRequest record was captured as a secondary "reported" record rather than an original primary source-of-truth record.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: Servers [MAY] use the MedicationRequest.reported[x] element to indicate ... the source of the report.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: When recording "self-prescribed" medication, Servers SHOULD use the MedicationRequest.requester element to indicate the Patient or RelatedPerson is the prescriber.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: Servers SHOULD support the encounter search parameter.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: [If Servers support the encounter search parameter, s]earching by encounter will return all medications ordered during that encounter, including medications administered in the hospital and prescribed or discharge medications intended to be taken at home.

Requirements to access “all medications” [all historical, active, and future prescribed medications and medications entered in error and whose status is unknown.] and “all active medications” [all medications with an active status. Active medications do not include past, future, unknown status, and entered-in-error medications.] for a patient: Servers MAY support the search parameters categoryand encounter. This search will return all medications ordered during an encounter for a given MedicationRequest.category such as "inpatient".

To provide a list of a patient’s medications, it may be necessary to “de-duplicate” them. The de-duplication activity MAY be supplied by the Server

An agent.type=”assembler”, agent.type=”transmitter”, or other agents from Provenance Agent Type value set MAY also be included.

implementers [of US Core's framework of Screening and Assessments] SHOULD consider more constrained, domain-specific profiles derived from the US Core Profiles to meet the needs of their respective use cases.

Every Server that supports the USDCI Data Class “Health Status/Assessments”:

SHALL support representing clinical judgments using US Core Condition Problems and Health Concerns Profile or US Core Simple Observation Profile.

Every Server that supports the USDCI Data Class “Health Status/Assessments”: ... The US Core Simple Observation Profile's Observation.derivedFrom element SHOULD reference the Structured Screening and Assessment upon which clinical judgment observations are made

Every Server that supports the USDCI Data Class “Health Status/Assessments”: ... the US Core Condition Profile's Condition.evidence.detail element SHOULD reference the Structured Screening and Assessment which assist in diagnosing problems or health concerns.

For API developers using US Core, it’s important to understand when to use the QuestionnaireResponse versus Observation to represent structured assessments and surveys. Here are some guidelines to help choose the appropriate profile: Observations represent specific point-in-time facts that need to be searched, trended, the subject of statistical analysis, and directly referenced in support of other actions ... anything that meets one of the preceding criteria must be surfaced as an Observation.

Implementers SHOULD treat these [USCDI Health Status Assessments Data Element] value sets as having an extensible binding.

when recording SDOH data [with] US Core Profiles, Servers SHOULD use … Social Determinants of Health Conditions Value Set, Social Determinants of Health Procedures Value Set, Social Determinants of Health Goals Value Set, [and] Social Determinants of Health Service Requests Value Set

Servers SHOULD maintain a stable common identifier for a resource across [FHIR] versions.

In an upgraded R4 endpoint, any data in FHIR DSTU2 SHOULD be in FHIR R4.

The FHIR RESTful resource types supported in a DSTU2 implementation SHOULD be supported in a R4 implementation

The FHIR RESTful resource types supported in a DSTU2 implementation SHOULD be supported in a R4 implementation [with the] exception [of] MedicationStatement may be deprecated, and the data SHOULD be mapped to MedicationRequest.

The FHIR RESTful resource types supported in a DSTU2 implementation SHOULD be supported in a R4 implementation [with the] exception [of] Care teams as represented by CarePlan in DSTU2 SHOULD be replaced by and the data mapped to CareTeam in R4

Servers SHOULD make available the same information in DSTU2 and R4 where the more recent standard allows.

Servers SHOULD make available the same information in DSTU2 and R4 where the more recent standard allows … [with the] exception [of] MedicationStatement data [should be] mapped to MedicationRequest

Servers SHOULD make available the same information in DSTU2 and R4 where the more recent standard allows … [with the] exception [of] care teams, as represented by CarePlan, SHOULD be mapped to CareTeam in R4

Data SHOULD be maintained between [FHIR] versions (i.e., not be degraded).

To allow Clients to use a single authorization token when accessing resources from multiple version-specific endpoints, Servers SHOULD use the same base authorization endpoint across versions.

[The clinical status of the allergy] SHALL be present if verification status is not “entered-in-error”

[The clinical status of the allergy] SHALL NOT be present if verification Status is “entered-in-error”

Profile Specific Implementation Guidance:

No Known Allergies may be represented using the US Core-AllergyIntolerance profile

[When used No Known Allergies is documented the system Shall use an] appropriate negation code in AllergyIntolerence.code

[When used No Known Allergies is documented the system Shall use an] verification status in AllergyIntolerance.verificationStatus

If a patient has not been asked about their allergies, this would be represented as: AllergyIntolerance.code = “1631000175102” (Patient not asked (contextual qualifier) (qualifier value)) AllergyIntolerance.verificationStatus = “unconfirmed” or empty (in other words, then element omitted)

If a patient has been asked, but has indicated they have no known allergies, this would be represented as: AllergyIntolerance.code = “716186003” (No known allergy (situation)) AllergyIntolerance.verificationStatus = “confirmed”

Considerations for systems aligning with HL7 Consolidated (C-CDA) Care Plan requirements: US Core Goal SHOULD be present in CarePlan.goal

Considerations for systems aligning with [HL7 Consolidated (C-CDA)] (http://www.hl7.org/implement/standards/product_brief.cfm?product_id=492) Care Plan requirements: ... US Core Condition SHOULD be present in CarePlan.addresses

Considerations for systems aligning with [HL7 Consolidated (C-CDA)] (http://www.hl7.org/implement/standards/product_brief.cfm?product_id=492) Care Plan requirements: Assessment and Plan MAY be included as narrative in CarePlan.text

Although both US Core Practitioner Profile and US Core PractitionerRole are Must Support, ... Server system[s conforming to the US Core CareTeam profile are] ... not required to support references to both, but SHALL support at least one of them

[When conforming to the US Core CareTeam profile] Because the US Core PractitionerRole Profile supplies the provider’s location and contact information and a reference to the Practitioner, Server systems [conforming to the US Core CareTeam profile] SHOULD reference it instead of the US Core Practitioner Profile [when conforming to the US Core CareTeam profile] .

Servers [conforming to the US Core CareTeam profile] that support only US Core Practitioner Profile and do not support the US Core PractitionerRole Profile SHALL provide implementation-specific guidance on how to access a provider’s location ... information using only the Practitioner resource.

Servers [conforming to the US Core CareTeam profile] that support only US Core Practitioner Profile and do not support the US Core PractitionerRole Profile SHALL provide implementation-specific guidance on how to access a provider’s ... contact information using only the Practitioner resource.

For Problems and Health Concerns [records, systems SHOULD] use the US Core Condition Problems and Health Concerns Profile.

USCDI’s applicable vocabulary standards for Encounter Diagnosis are SNOMED CT and ICD-10-CM. The US Core Condition Codes only supports ICD-9-CM for historical purposes. When using ICD codes, only non-header ICD-10-CM codes SHOULD be used as the primary code for current encounter diagnoses.

[A US Core Condition Encounter Diagnosis] encounter SHOULD always be referenced in Condition.encounter.

For Encounter Diagnosis [records, systems SHOULD] use the US Core Condition Encounter Diagnosis Profile.

If the category is "problem-list-item", Condition.clinicalStatus SHOULD be present.

Updates to .meta.lastUpdated SHOULD reflect: New problems and health concerns

Updates to .meta.lastUpdated SHOULD reflect: Changes in the clinical status or verifications status of problems or health concerns.

A coverage.type of “81” (Self-pay) MAY be used to imply that the patient has no coverage or that an individual or organization other than an insurer is taking responsibility for payment for a portion of the health care costs.

Implementers should refer to the PHDSC Payer Type Committee User’s Guide for the Source of Payment Typology when selecting codes.

To differentiate Medicare Parts A, B, C, and D systems can use the following codes [when sending Coverage.type]: [For] Part A and B [use] 121 (Medicare Fee For Service)

To differentiate Medicare Parts A, B, C, and D systems can use the following codes: [For] Part C (Medicare Advantage Plan) [use] 111 (Medicare HMO), 112 (Medicare PPO), 113 (Medicare POS)

To differentiate Medicare Parts A, B, C, and D systems can use the following codes: [For] Part D [use] 122 (Medicare Drug Benefit)

If Insurers issue unique member IDs for dependents, then the memberId Coverage.identifier should be used [with the unique dependent ID] instead of Coverage.dependent to uniquely refer to the dependent with respect to their insurance.

Implementers are encouraged to use the FDA Global UDI Database (GUDID) and associated APIs to parse and validate the [unique device ID] UDI

Medical devices with UDI information SHALL represent the UDI code in Device.udiCarrier.carrierHRF

UDI-PI elements present [in Device.udiCarrier.carrierHRF] SHALL be represented in the corresponding US Core Device Profile elements

If UDI is not present and the manufacturer … is available, … [it] SHOULD be included to support historical reports of implantable medical devices [where] manufacturer [is sent in] Device.manufacturer

If UDI is not present and the ... model number information is available, … [it] SHOULD be included to support historical reports of implantable medical devices [where] model [is sent in] Device.model

Records of implanted devices MAY be queried against UDI data, including: UDI HRF string (udi-carrier)

Records of implanted devices MAY be queried against UDI data, including:UDI Device Identifier (udi-di)

Records of implanted devices MAY be queried against UDI data, including: Manufacturer (manufacturer)

Records of implanted devices MAY be queried against UDI data, including: Model number (model)

Implementers MAY also adopt custom SearchParameters for searching by lot numbers

Implementers MAY also adopt custom SearchParameters for searching by serial number

Implementers MAY also adopt custom SearchParameters for searching by expiration date

Implementers MAY also adopt custom SearchParameters for searching by manufacture date

Implementers MAY also adopt custom SearchParameters for searching by distinct identifier

The diagnostically relevant time (known as the “effective time” and typically the time of specimen collection) … SHALL be present if status [of the diagnostic report] is ‘partial’, ‘preliminary’, ‘final’, ‘amended’, ‘corrected’, or ‘appended’.

When the report was released … SHALL be present if status [of the diagnostic report] is ‘partial’, ‘preliminary’, ‘final’, ‘amended’, ‘corrected’, or ‘appended’.

Updates to .meta.lastUpdated SHOULD reflect New laboratory reports.

Updates to .meta.lastUpdated SHOULD reflect changes in the status of laboratory reports, including events that trigger the same status (e.g., amended → amended).

The DiagnosticReport.category binding Must Support, at a minimum, the US Core DiagnosticReport Category Codes of Cardiology, Radiology, and Pathology

Other [diagnostic report] categories may be supported [when using US Core DiagnosticReport Profile for Report and Note Exchange]

[L]inkages between specific LOINC codes and the LP-type codes may be used as guidance [for a Server's categorization of diagnostic reports]

For Diagnostic Imaging Reports systems SHOULD support using the subset of LOINC codes listed Table 4: LOINC® Document Type Codes in HL7 Standard for CDA® Release 2: Imaging Integration; Basic Imaging Reports in CDA and DICOM, Release 1

The DocumentReference.type binding Must Support, at a minimum, the 10 Common Clinical Notes

The DocumentReference.type binding may extend to the whole US Core DocumentReference Type Value Set

[DocumentReference.type may also use] other category schemes such as the LOINC-based Document Class Value Set and IHE XDSclassCode

Although both [DocumentReference.attachment.url and DocumentReference.attachment.data] are marked as Must Support, the Server system is not required to support an address and inline base64 encoded data, but SHALL support at least one of these elements.

If there are multiple DocumentReference.content element repetitions, these SHALL all represent the same document in different formats or attachment metadata

The [documentReference.content] element SHALL NOT contain different versions of the same content.

The organization responsible for the DocumentReference SHALL be present either in DocumentReference.custodian or accessible in the Provenance resource targeting the DocumentReference using Provenance.agent.who or Provenance.agent.onBehalfOf

Although ... marked as Must Support, Servers are not required to support both [an Encounter.reasonCode or a reference with Encounter.reasonReference], but they SHALL support at least one of these elements.

If Encounter.reasonReference references an Observation, it SHOULD conform to a US Core Observation [profile applicable to the observation being made]

Although … marked as Must Support, Servers are not required to support both Encounter.location.location and Encounter.serviceProvider, but they SHALL support at least one of these elements.

If the event facility/location differs from the Encounter.location, systems SHOULD reference it directly

If the event facility/location differs from the Encounter.location, … systems SHALL use the location element for all resources where the element is available.

If the event facility/location differs from the Encounter.location … systems MAY use the standard Event Location Extension for US Core DiagnosticReport Profile for Laboratory Results Reporting and US Core Observation Clinical Result Profile.

Updates to .meta.lastUpdated SHOULD reflect New encounters/visits

Updates to .meta.lastUpdated SHOULD reflect changes in the status of encounters, including events that trigger the same status (e.g., in-progress → in-progress).

Although both Goal.startDate and Goal.target.dueDate are marked as Must Support, the Server system is not required to support both, but SHALL support at least one of these elements.

[Servers shall] use the status code: not-done to represent that an immunization was not given.

The preferred code system identifier … is http://hl7.org/fhir/sid/cvx for CVX [vaccine codes]

The preferred code system identifier … is http://hl7.org/fhir/sid/ndc for NDC vaccine codes]

When using a code [to represent a medication for a medication dispense], RXNorm concepts are used.

When using a code [to represent a medication for a medication dispense], USCDI recommends the National Drug Codes (NDC) as an optional terminology. They can be supplied as an additional coding element.

When referencing a Medication resource in .medicationReference, the resource may be contained

When referencing a Medication resource in .medicationReference, the resource may be an external resource

The Server systems are not required to support both a [medication] code and a reference [when sending medicationDispense], but SHALL support at least one of these methods

If an external reference to a Medication resource is used, the Server SHALL support the _include parameter for searching this element.

When recording “self-prescribed” medication, requester SHOULD be used to indicate the Patient or RelatedPerson as the prescriber

Although [medicationRequest.reportedBoolean and MedicationRequest.reportedReference] are both marked as Must Support, the Server system is not required to support both, but SHALL support at least one of these elements

[For organizations participating in the ONC Health IT Certification program,] the referenced resources SHOULD be a US Core Profile as documented in Referencing US Core Profiles.

Source EHR identifiers SHOULD be included to support deduplication across MedicationRequest resources.

The observations MAY have additional codes that translate or map to the Observation code or category codes [such as] … local system-specific codes [and] …more specific codes

[For an Observation a] code system value SHOULD be supplied for each additional code

Because the blood pressure values are communicated in the mandatory systolic and diastolic components [when using the US Core Average Blood Pressure Profile,] the Observation.value[x] element SHALL be omitted

The context or precondition of a patient’s [care experience] preference SHOULD be supplied in the Observation.valueString or in an extension

Servers SHOULD use the base FHIR [Observation Category Codes] (http://hl7.org/fhir/R4/valueset-observation-category.html) [in Observation.category]

Systems SHOULD support Observation.effectivePeriod to accurately represent measurements over time

An Observation.component without a value, SHALL include a reason why the data is absent

Systems that never provide a component observation without a component value … [MAY choose not] to support Observation.component.dataAbsentReason

An Observation without a value, SHALL include a reason why the data is absent

[For] an Observation without a value … [Systems MAY choose not to] include a reason why the data is absent … [if] there are component observations or … reporting panel observations using Observation.hasMember

Systems that never provide an observation without a value ... [MAY choose not] to support Observation.dataAbsentReason

[When sending US Core Laboratory Results Observation Profile] updates to .meta.lastUpdated SHOULD reflect new laboratory observations

[When sending US Core Laboratory Results Observation Profile] updates to .meta.lastUpdated SHOULD reflect changes in the status of laboratory observations, including events that trigger the same status (e.g., amended → amended)

[When sending US Core Observation Occupation Profile] for … [a] current job, [Servers SHALL] omit observation.effectivePeriod.end to indicate it is ongoing.

When the industry is known, but the occupation is not, [Servers SHALL] use the value “unknown” from the DataAbsentReason Code System

when the occupation is known but the industry is not, [Servers SHALL] omit the industry Observation.component

To represent the patient’s pregnancy status, [Servers SHALL] use the US Core Observation Pregnancy Status Profile.

To represent the patient’s intent to become pregnant, [Servers SHALL] use the US Core Observation Pregnancy Intent Profile.

[For] multi-question surveys or assessments Observation.code is an overarching assessment or screening code, and the Observation.value element SHOULD be empty

A practitioner’s clinical observation or assertion about a patient’s health status, which is not a response to a screening or assessment question, SHOULD use the US Core Simple Observation Profile instead [of the US Core Observation Screening Assessment Profile]

the Server system ... SHALL support [either] Reference(US Core Observation Screening Assessment Profile) or Reference(US Core QuestionnaireResponse Profile) for Observation.derivedFrom

Observations formally part of an assessment tool or survey SHOULD use the US Core Observation Screening Assessment Profile.

An assertion or determination derived from screening and assessment tools SHOULD reference them using Observation.derivedFrom

When using `Observation.derivedFrom’ to reference an Observation, the referenced Observation SHOULD be a US Core Observation

[US Core] treatment intervention preferences expressed by a patient may be documented in narrative (text) form or the result of selecting from a list of options provided by the content creator/implementer.

[When using US Core Treatment Intervention Preference Profile] the context or precondition of a patient’s preference SHOULD be supplied in the Observation.valueString ... or an extension

[US Core Vital Signs] observations MAY have component observations … [see] FHIR core specification vital signs table for examples

Information about the growth chart tables used to determine percentiles SHOULD be supplied in Observation.note.text

[In the US Core Blood Pressure Profile] because the blood pressure values are communicated in the mandatory systolic and diastolic components[,] the Observation.value[x] element SHOULD be omitted

[In the US Core Blood Pressure Profile] because the blood pressure values are communicated in the mandatory systolic and diastolic components[,] an Observation without a systolic or diastolic result value, SHALL include a reason why the data is absent in Observation.component.dataAbsentReason

[In the US Core Blood Pressure Profile] because the blood pressure values are communicated in the mandatory systolic and diastolic components[,] All Server systems - including those that never provide a component observation without a value - SHALL support Observation.component.dataAbsentReason for the components.

Inspired oxygen therapy may be represented with component observations when measured at the same time as the pulse oximetry measurements [in the US Core Pulse Oximetry profile]

Many pulse oximetry readings are taken while the patient is breathing room air. The concept of “room air” (unmodified, ambient air) SHOULD be represented as an inhaled oxygen flow rate of 0 liters/min

A pulse oximetry reading without inspired oxygen component observations may imply that the measurement was performed while the patient was breathing room air or that the inspired oxygen reading was omitted. To remove this uncertainty, the inspired oxygen component observations SHOULD be used [when using the US Core Pulse Oximetry profile.]

Systems SHALL support National Provider Identifier (NPI) for organizations

Systems … SHOULD the National Association of Insurance Commissioners NAIC Company code (sometimes called “NAIC Number” or “cocode”) for payers.

Systems … SHOULD support Clinical Laboratory Improvement Amendments (CLIA) for laboratories

The Complex [Extension] for Race ... allow[s] for one or more codes of which MAY include additional codes from the detailed ethnicity ... value sets drawn from the Race & Ethnicity - CDC (CDCREC) code system [when using the [US Core Race Extension] (https://hl7.org/fhir/us/core/STU8/StructureDefinition-us-core-race.html)]

The Complex [Extension] for Race ... allow[s] for one or more codes of which SHALL include a text description [of category codes when using the [US Core Race Extension] (https://hl7.org/fhir/us/core/STU8/StructureDefinition-us-core-race.html)]

The Complex [Extension] for ... Ethnicity allow[s] for one or more codes of which MAY include additional codes from the detailed ... race value sets drawn from the Race & Ethnicity - CDC (CDCREC) code system [when using the [US Core Race Extension] (https://hl7.org/fhir/us/core/STU8/StructureDefinition-us-core-race.html)]

The Complex [Extension] for ... Ethnicity allow[s] for one or more codes of which SHALL include a text description [of category codes when using the [US Core Ethnicity Extension] (https://hl7.org/fhir/us/core/STU8/StructureDefinition-us-core-ethnicity.html)]

Previous name is represented by setting Patient.name.use to “old” or providing an end date in Patient.name.period or doing both

Previous address is represented by setting Patient.address.use to “old” or providing an end date in Patient.address.period or doing both.

Non-certifying systems[, systems that are not participating in the ONC Health IT certification program,] SHOULD follow the Project US@ Technical Specification for Patient Addresses Final Version 1.0 as the standard style guide for Patient.address.line and Patient.address.city for new and updated records.

[For systems participating in the ONC Health IT certification program,] this requirement [to follow the Project US@ Technical Specification for Patient Addresses Final Version 1.0 for Patient.address.line and Patient.address.city] does not apply to historical records or documents that are exposed through FHIR-based APIs. [Organizations MAY choose not to use use Project US@ Technical Specification for Patient Addresses Final Version 1.0 when sending historical records]

SHOULD NOT be used as a patient identifier in Patient.identifier.value

Servers that support only the US Core Practitioner Profile and do not support the US Core PractitionerRole Profile SHALL provide implementation-specific guidance on how to access a provider’s location and contact information using only the Practitioner resource.

Although Practitioner.address is marked as Must Support, the Server system … [MAY choose not to] support it if they support the US Core PractitionerRole Profile

[When using the US Core Practioner Profile] Practitioner.address … SHALL[be supported] if … [the Server does] not support the US Core PractitionerRole Profile

Only professional/work contact information about the practitioner SHOULD be available to the patient

[If using LOINC codes in procedure.code] only LOINC concepts that reflect actual procedures SHOULD be used

[For organizations participating in the ONC Health IT Certification program,] when using Procedure.reasonReference Servers SHALL support at least one target resource in Procedure.reasonReference

[For organizations participating in the ONC Health IT Certification program,] when using Procedure.reasonReference …The referenced resources SHOULD be a US Core Profile as documented in Referencing US Core Profiles

The US Core Provenance resource SHALL be supported for these US Core resource types: AllergyIntolerance, CarePlan, CareTeam, Condition, Coverage, Device, DiagnosticReport, DocumentReference, Encounter, Goal, Immunization, MedicationDispense, MedicationRequest, Observation, Patient, Procedure, QuestionnaireResponse, RelatedPerson, ServiceRequest

If a system receives a provider in Provenance.agent.who as free text, they must capture [the organization] who sent them the information [and upon] request ... SHALL provide this organization as the source

If a system receives a provider in Provenance.agent.who as free text, … [upon request they] MAY include the free text provider.

The Must Support ServiceRequest.category is bound, at a minimum, to the US Core ServiceRequest Category Codes, and other category codes can be used.

The ServiceRequest.code value ... SHOULD be constrained to a subset for a particular use case or domain

Since the binding [for Specimen.type and additional USCDI elements] is extensible when a code is unavailable, just text is allowed [and conformant].

Although both Specimen.identifier and Specimen.accessionIdentifier are marked as Must Support, the Server system is not required to support both, but SHALL support at least one of these elements.

Systems SHALL establish a risk analysis and management regime that conforms with HIPAA security regulatory requirements

US Federal systems SHOULD conform with the risk management and mitigation requirements defined in NIST 800 series documents.

US Federal systems … SHOULD include security category assignment following NIST 800-60 vol. 2 Appendix D.14.

The coordination of risk management and the related security and privacy controls … SHOULD be defined in the Business Associate Agreement when available.

Systems SHALL reference a single time source to establish a common time base for security auditing and clinical data records among computing systems.

The selected time service SHOULD be documented in the Business Associate Agreement when available.

Systems SHALL keep audit logs of the various transactions.

Systems SHALL use TLS version 1.2 or higher for all transmissions not taking place over a secure network connection.

US Federal systems SHOULD conform with FIPS PUB 140-2.

Systems SHALL conform to FHIR Communications Security requirements.

For Authentication and Authorization, Systems SHALL support any SMART App Launch Version 2.0.0 and later for Client <-> Server interactions.

Systems SHALL implement consent requirements per their state, local, and institutional policies.

The Business Associate Agreements SHOULD document systems’ mutual consent requirements.

Systems SHOULD provide Provenance statements using the US Core Provenance Profile resource and associated requirements.

Systems MAY implement the FHIR Digital Signatures

Systems MAY protect the confidentiality of data at rest via encryption and associated access controls.

If any sub-element is marked as Additional USCDI [for a complex element], it must also meet the Additional USCDI requirements and satisfy the Additional USCDI requirements for the parent element.

When a Reference type element is labeled as Must Support and has a single target profile referenced, the target profile SHALL be supported.

When a Reference type element is labeled as Must Support, has multiple target profiles referenced, and specific targets are labeled as Must Support, the Must Support target profile(s) SHALL be supported.

[If a] slicer's Must Support property only defines the element level Must Support...[, i.e.,] no Must Support... property is defined for the slice, then support for that slice's definition is optional.

Servers SHALL support, at minimum, these … "Common Clinical Notes": Surgical Operation Note (11504-8)

Servers SHALL support, at minimum, these … "Common Clinical Notes": Emergency Department Note (34111-5)

Servers that support the USCDI Health Status/Assessments Data Class SHALL support the US Core Observation Screening Assessment Profile

Servers that support the USCDI Health Status/Assessments Data Class ... SHOULD support the SDC Base Questionnaire and the US Core QuestionnaireResponse Profile."

For the US Core Simple Observation Profile, Servers SHALL support all the category codes listed.

For the … US Core Observation Screening Assessment Profiles, Servers SHALL support all the category codes listed.

For the US Core Condition Problems and Health Concerns Profile, Servers SHALL support the code ,"sdoh"

For the US Core Condition Problems and Health Concerns Profile, Servers SHOULD support the other category codes listed.

For the US Core ServiceRequest Profile, Servers SHOULD support all the [listed] category codes.

[The Server SHALL support the] category of "problem-list-item"

[The Server SHALL support the] category of "health-concern"

The DiagnosticRequest.basedOn element connects the DiagnosticReport to the originating order in the EHR. Systems that initiate the order SHOULD use this element when reporting the results.

The DiagnosticReport.media.link element SHOULD be used to support links to various patient-friendly content, such as jpg images of x-rays (see the DiagnosticReport Chest X-ray Report Example).

The DiagnosticReport.imagingStudy element SHOULD be used to support exchange with systems that can view DICOM (Digital Imaging and Communications in Medicine) studies, series, and SOP (Service-Object Pair) instances referenced in the -ImagingStudy resource.

If the referenced a document or file [referenced by DocumentReference.content.attachment.url] is hosted on a Server outside the FHIR Server, it should be securely accessible using the same authorization credentials that were used to access the FHIR Server. This reduces complexity for the Client and improves the user experience.

Although the [US Core Interpreter Needed] extension is marked as an Additional USCDI Requirements on both US Core Patient and US Core Encounter Profiles, the certifying Server system is not required to support the extension on both profiles, but SHALL support the extension on at least one.

There is no guarantee that vaccine lot numbers are globally unique, and they are not recommended for matching or de-duplication across systems unless used with other data elements such as a vaccine product code, manufacturer code, or date of administration. Implementers MAY communicate the Immunization.manufacturer to ensure global uniqueness to lot numbers.

Servers SHALL return all active medications following the Get All Active Medications guidance.

[For US Core Laboratory Result Observation Profile, even] when the specimen type is already implied by the LOINC code used in Observation.code (e.g., a LOINC code for Blood Glucose), the Observation.specimen element SHOULD also be populated with the referenced Specimen resource to explicitly communicate the collected specimen type.

[For US Core Laboratory Result Observation Profile, the] type of specimen [in Observation.specimen] SHOULD not conflict with the LOINC code [in Observation.code].

[For the US Core Observation Screening Assessment Profile, the] category type "survey" is required.

[For the US Core Observation Screening Assessment Profile,] when using Observation.derivedFrom to reference an Observation, the referenced Observation SHOULD be a US Core Observation.

Servers can use the US Core Interpreter Needed Extension on [the US Core Patient Profile] or the US Core Encounter Profile to communicate whether a patient needs an interpreter.

Systems SHOULD designate the patient's preferred language in the Patient.communication.preferred element.

For the USCDI Laboratory Order, Imaging Order, Clinical Test Order, and Procedure Order Data Elements, implementers SHOULD use the corresponding category codes listed in the table

An Observation.component without a value, SHALL include a reason why the data is absent.

Systems that never provide a component observation without a component value are not required to support Observation.component.dataAbsentReason.

Although 'Observation.performer' target profiles US Core Practitioner Profile and US Core Patient Profile are labeled Must Support. Servers are not required to support both, but SHALL support at least one.

[If a] slicer's Additional USCDI property only defines the element level Additional USCDI...[, i.e.,] no Additional USCDI... property is defined for the slice, then support for that slice's definition is optional.

When a Reference element is labeled as Must Support has multiple target profiles referenced, but none are labeled as Must Support

...

• US Core Responders SHALL be capable of providing [such an element] with a valid reference to at least one target profile.

The author is communicated by the elements and the author's role by the referenced target resource (for example, Patient, Practitioner/PractitionerRole, RelatedPerson, Device). Details about the author's role are contained in the target resource's contents. Many of these elements are labeled Must Support or Additional USCDI Requirements. However,[even if they are not labeled Must Support or Additional USCDI Requirements] all of these elements and target resources SHOULD be supported in the profiles if the system captures the data.

Although 'Observation.performer' target profiles US Core Practitioner Profile and US Core Patient Profile are labeled Must Support. Servers are not required to support both, but SHALL support at least one.

… this profile uses USPS Two Letter Alphabetic Codes for US states. When encoding addresses outside the US, systems SHOULD use the ISO 3166 subdivision codes.

Due to implementer feedback, some US Core Profiles reference the base FHIR PractitionerRole resource instead of the US Core PractitionerRole Profile. However, the US Core PractitionerRole Profile SHOULD be used as the default profile if referenced by another US Core profile.

When selecting role codes...implementers SHOULD choose the code that reflects the specific duties performed within that role rather than the specialty unless the individual's professional specialization characterizes the role.

A procedure including an implantable device SHOULD use Procedure.focalDevice referencing the US Core Device Profile

Other information can be tracked such as what activity has occurred in Provenance.activity, and details about where the entity came from (for example, a document source's identity and role) in Provenance.entity.

Unless exchanging legacy or text-only data, medical specialty codes SHOULD be taken from NUCC or SNOMED CT.

Unless exchanging legacy or text-only data, procedure codes SHOULD be taken from SNOMED-CT, CPT, HCPCS II, ICD-10-PCS, CDT, or LOINC*.

US Core guidance provides more flexibility for situations where implementers cannot fully comply with the FHIR guidance. This flexibility is sometimes necessary and expected for legacy and text-only data. However, for newly recorded, non-legacy data, a system SHOULD adhere to the extensible binding rules.

For data not captured by the system transmitting the information, the coded data SHOULD be automatically converted to fine-grained codes from the specified ValueSet.

For data not captured by the system transmitting the information, ....the system MAY provide the existing code or the free text, and a high-level abstract code, such as the SNOMED CT code "71388002"(Procedure), to remain conformant with the extensible binding.

To represent the patient and provider attestation metadata, the optional DocumentReference Attestor R5 Extension ... SHOULD be use

Unless exchanging legacy or text-only data, condition codes SHOULD be taken from SNOMED CT and ICD-10-CM, USCDI's applicable vocabulary standards for the Problem Data Element.

The Service Delivery Location Role Type value set is inherited from the base resource, and, although its binding strength is extensible in the base resource, US Core implementers MAY treat it as preferred.

When exchanging Social Determinants of Health (SDOH) data, implementers SHOULD use the Social Determinants of Health Goals set which is curated by the Gravity Project to represent goals for all SDOH Domains.

Implementers SHOULD conform to the binding strengths listed for each USCDI Order context...Laboratory Order|[LOINC Common Laboratory Orders Value Set]|extensible, Imaging Order|[LOINC Radiology Codes]|preferred, Clinical Test Order|[LOINC Clinical Test Codes]|preferred

the US Core ServiceRequest Profile MAY be used to represent a derived "actionable" order based on a portable Medical Order(PMO).

[When] used to represent a derived "actionable" order based on a PMO...The ServiceRequest.category SHOULD be bound to Portable Medical Order Categories

[When] used to represent a derived "actionable" order based on a PMO...ServiceRequest.code` SHOULD be aligned with the appropriate Category

The ServiceReqest.reasonReference SHOULD be used to reference the US Core ADI DocumentReference Profile that indexes the PMO.