This page is part of the Clinical Guidelines (v2.0.0: STU2) based on FHIR (HL7® FHIR® Standard) R4. This is the current published version. For a full list of available versions, see the Directory of published versions
FHIR profiles are a key aspect of supporting interoperable computable content, providing a formal mechanism for defining the data elements used by computable content, as well as conformance expectations for the computable content itself. Readers of this implementation guide should be familiar with the Profiling FHIR topic in the base specification.
This page provides overview documentation for the following categories of profiles and documentation:
For indexes of profiles and extensions defined in this implementation guide, see the following:
For definitional artifacts, this implementation guide builds on the capability profiles defined in the Canonical Resource Management Infrastructure (CRMI) Implementation Guide, defining profiles for computable knowledge artifacts used in the representation of clinical guidelines. For background on the CRMI capability profiles, see the Profiles discussion in the CRMI IG.
Capability | Computable |
---|---|
Activity | CPGComputableActivity |
GraphDefinition | CPGComputableGraphDefinition |
ImplementationGuide | CPGComputableGuideline |
Metric | CPGComputableMetric |
PlanDefinition | CPGComputablePlanDefinition |
Questionnaire | CPGComputableQuestionnaire |
StructureDefinition | CPGCaseFeatureDefinition |
In addition, content claiming conformance to this implementation guide SHALL use the Shareable profile and SHOULD use the Publishable profile for the following artifact types:
Level | Definition | Plan/Instance |
---|---|---|
Group | CPGGroupDefinition | CPGGroup |
Location | CPGLocationDefinition | CPGLocation |
Organization | - | CPGOrganization |
Patient | - | CPGPatient |
Practitioner | - | CPGPractitioner |
PractitionerRole | CPGPractitionerRoleDefinition | CPGPractitionerRole |
CareTeam | CPGCareTeamDefinition | CPGCareTeam |
Pathway | CPGPathwayDefinition | CPGCase CPGCarePlan |
Strategy | CPGStrategyDefinition | CPGStrategy |
Recommendation | CPGRecommendationDefinition | CPGAdministerMedication CPGCommunicationRequest CPGDispenseMedicationTask CPGDocumentMedicationTask CPGEnrollmentTask CPGGenerateReportTask CPGImmunizationRequest CPGMedicationRequest CPGProposeDiagnosisTask CPGQuestionnaireTask CPGRecordDetectedIssueTask CPGRecordInferenceTask CPGReportFlagTask CPGServiceRequest |
Goal | defined in PlanDefinition.goal | CPGGoal |
Metric | CPGComputableMetric | CPGMetricReport |
Assessment | CPGClinicalImpression | |
Case Summary | CPGCaseSummaryDefinition | CPGCaseSummary |
Case Plan Summary | CPGCasePlanSummaryDefinition | CPGCasePlanSummary |
Case Plan Progressing Note | CPGCasePlanProgressingNote |
Examples of the use of these profiles are available in the Examples page. In particular, see the Congestive Heart Failure Pathway example.
To represent the activities in a computable guideline, this implementation guide follows the workflow patterns established by the base FHIR specification, definition, request, and event. For each type of activity, these profiles establish at least:
Note that the intent of these profiles is not to establish the content of the activity, only to ensure that requests and events related to guideline-based care can be traced back to the recommendation, strategy, pathway, and ultimately guideline they are representing.
See the Activity Examples for a complete example of each of the above activity types.
See the Activity Flow topic for a complete discussion of how activities move through the definition, request, and event phases.
In addition to the profiles defined by this implementation guide, computable content generally deals with two broad categories of profiles:
Interoperability profiles establish standards of data exchange between systems, and are typically defined in and distributed as part of model implementation guides. To be useful, these profiles will generally be established across a broad range of systems, all operating in a particular environment, or in support of a particular set of use cases. Examples of these types of profiles are:
Computability profiles describe the data expectations for computable content, and are typically defined in and distributed as part of content implementation guides. For example, given the following condition:
define "Active Ambulatory Opioid Rx":
[MedicationRequest: "Ambulatory Abuse Potential Opioids"] Rx
where Rx.status = 'active'
and Rx.category contains "Outpatient"
In this example, the condition is referencing the status
and category
elements of the MedicationRequest
resource. In order for this logic to be evaluated effectively, the system providing the MedicationRequest
instance must understand these two elements and provide data for them if it is available. Computability profiles are used to communicate this information through FHIR profiles.
Note that for logic expressed in CQL, this information can be inferred by static analysis of the CQL expressions and is exposed statically in the dataRequirement
element of the Library
resource. In addition, the $data-requirements
operation can be used to extract this information from a CQL Library dynamically.
Content conforming to this IG SHOULD select a set of base interoperability profiles appropriate for the intended target. For international usage, implementation guides conforming to this IG SHOULD use International Patient Summary.
In general, implementation of any given computable content is based on the intersection of the interoperability and computability profiles. As such, content authors must take care not to define computability profiles that conflict with interoperability profiles in their intended target scope.
Clinical practice guidelines need to define the complete set of data elements required for the logic to function. This is the purpose of the computability profiles. Note that although selecting a base set of interoperability profiles (such as International Patient Summary) can provide a basis for sharing common data elements, experience has shown that even for relatively smaller CPGs, the set of computability profiles is never completely representable in any standard terminology or core data model. As such, mapping is required to support the use of a CPG in a given setting to demonstrate:
In particular, IPS as a summary model is an incomplete target for the representation of data. It should be used as a basis for sharing data elements, but must be augmented by the specific data elements and mappings required to support computable representation of the logic of a clinical guideline.
As well, how data is exchanged is an importantly different consideration than what data is being exchanged, and the purpose of computability profiles in an IG is to specify what data is to be exchanged, not how. In particular, although document sharing could be a solution for CPGs with very few and/or low volume data requirements, care should be taken to ensure that implementations are free to make appropriate decisions for their environments about how data exchange is occurring.