HL7 CDS Standards
- 1 Overview
- 2 Contributing ONC Initiatives
- 3 Foundational Standards -- Information Models and Terminology
- 4 Foundational Standards -- EHR Integration of Web Applications
- 5 Foundational Standards -- Expression Language
- 6 Knowledge Representation Standards
- 7 Infobutton Standards
- 8 CDS Service Standards
Standards for CDS include foundational and enabling standards that may not be specific to CDS, as well as standards that are specific to CDS and/or electronic clinical quality measurement (eCQM), the latter of which is the domain of the HL7 Clinical Quality Information (CQI) Work Group (http://wiki.hl7.org/index.php?title=Clinical_Quality_Information_Work_Group). The standards listed on this page are organized around those domains.
- Please contact the HL7 CDS Work Group list-serv with any questions.
- Many of the standards listed below are co-sponsored by other HL7 Work Groups, and several of them are primarily owned by other Work Groups.
Contributing ONC Initiatives
Many of the standards listed below were developed and validated through a variety of public-private initiatives sponsored by the U.S. Office of the National Coordinator for Health IT (ONC) (https://oncprojectracking.healthit.gov/wiki/display/TechLabSC/ONC+Tech+Lab+Standards+Coordination+Home), including the ONC and CMS-sponsored Clinical Quality Framework (CQF) initaitve and the ONC-sponsored Health eDecisions (HeD) initiative.
Foundational Standards -- Information Models and Terminology
Interoperable CDS solutions, like other interopeable health IT solutions, depend on a common understanding of underlying data and terminology.
General terminologies used by CDS solutions include:
- The International Classification of Diseases (ICD) (http://www.who.int/classifications/icd/en/)
- The Systematized Nomenclature of Medicine - Clinical Terms (SNOMED CT) (http://www.snomed.org/snomed-ct)
- Logical Observation Identifiers Names and Codes (LOINC) (https://loinc.org/)
General information model and data access APIs used by CDS solutions include:
- HL7 C-CDA standard (http://www.hl7.org/implement/standards/product_brief.cfm?product_id=408)
- HL7 FHIR standard (https://www.hl7.org/fhir/)
- HL7 FHIR US Core Implementation Guide (http://www.hl7.org/fhir/us/core/)
- The Argonaut Implementation Guide for FHIR (http://argonautwiki.hl7.org/index.php?title=Implementation_Guide)
Information models developed specifically to meet CDS needs:
- HL7 Virtual Medical Record for Clinical Decision Support (vMR-CDS) and associated standards:
Information models developed to jointly meet CDS and eCQM needs:
- QICore FHIR Implementation Guide (http://hl7.org/fhir/us/qicore/history.html)
Current focus of the HL7 CDS Work Group and its members:
- Build on the HL7 FHIR standard due to strong EHR vendor adoption and support
- Collaborate with the HL7 Clinical Information Modeling Initiative (CIMI) Work Group (http://www.hl7.org/Special/Committees/cimi/index.cfm) to generate detailed clinical models that encompass the semantics included in the vMR-CDS
- Develop comprehensive set of FHIR profiles based on these CIMI models, and building on the QICore FHIR Implementation Guide
- Accelerate EHR vendor adoption of these FHIR profiles, including by advocating for inclusion of these profiles into the US Core FHIR profiles and by developing and sharing support for these profiles (e.g., through the Healthcare Services Platform Consortium -- http://hspconsortium.org/)
Foundational Standards -- EHR Integration of Web Applications
The integration of externally developed applications (apps) into the EHR is rapidly becoming a mainstream approach to integrating additional functionality into the native user interface of EHR systems. The framework being adopted by EHR vendors for this is called SMART, short for Substitutable Medical Apps Reusable Technologies). When combined with the use of FHIR for a data model, this framework is called SMART on FHIR. The HL7 SMART specification is at http://www.hl7.org/fhir/smart-app-launch/index.html. While not exclusive to CDS solutions, SMART is an important enabler for embedding CDS solutions into EHR systems in an interoperable manner.
Foundational Standards -- Expression Language
Another type of foundational CDS standard is expression languages for clinical logic. For example, an expression language is needed for expressing in a computable manner the following: if a patient has diabetes mellitus, and if the last hemoglobin A1c test result was less than 7%, recommend re-testing if it has been over 6 months, but if the last test result was greater than or equal to 7%, then recommend re-testing if it has been over 3 months.
These following standards enable the expression of such logical statements:
GELLO, R2 Normative
The current focus of the CDS WG is to build on the Clinical Quality Language (CQL). Of note, CMS has announced that it plans to use CQL for the specification of eCQMs (https://ecqi.healthit.gov/cql).
Knowledge Representation Standards
Building on the foundational standards described above, there are standards for representing clinical knowledge, so that they can be shared and consumed by various health IT systems. These standards support rules (e.g., if patient meets criteria set A, recommend B), order sets, and documentation templates.
Arden Syntax, v2.10 Normative (supports rules)
HL7 CDS Knowledge Artifact Specification (KAS), Release 1.3 (supports rules, order sets, and documentation templates)
HL7 FHIR Clinical Reasoning Module (supports rules, order sets, and documentation templates; FHIR verison of the CDS KAS specification above)
Contextualized potential drug-drug interaction clinical decision support (a formalism to represent computable logic for potential drug-drug interactions)
Also, while not focused on the representation of fully structured and executable medical knowledge, the Guideline Elements Model (GEM) can be useful for the representation of semi-structured knowledge that can serve as an intermediary between narrative prose clinical guidelines and fully structured and executable knowledge (http://gem.med.yale.edu/default.htm).
The current direction of the CDS Work Group is to define a common conceptual model between the CDS-KAS and the FHIR Clinical Reasoning module so that they are aligned and to continue to enhance these specifications.
Infobutton standards allow the integration of informational resources into the EHR, while using relevant context information (e.g., patient age, gender, current user task in the EHR) to refine the informational resources presented. Currently, there is exploration of how Infobuttons can leverage the CDS Hooks decision support service specification described below.
Infobutton SOA Implementation Guide, R1 Normative
Knowledge Request RMIM, R1 Normative
CDS Service Standards
As a complementary approach to sharing standard knowledge resources (e.g., CDS-KAS artifacts) that can be imported and consistently interpreted by various health IT systems, CDS capabilities themselves can be shared as a securely accessible Web service. The current generation of standards in this area is called the Decision Support Service (DSS) standard, which was developed jointly by HL7 and the Object Management Group (OMG). A FHIR version of such an evaluation interface was included in the FHIR Clinical Reasoning module, which has been harmonized/unified with the CDS Hooks specification. The current focus of the CDS Work Group is on the refinement and adoption of the CDS Hooks specification due to its growing adoption by EHR vendors.
HL7 DSS Standard, Release 1 Normative
HL7 DSS Standard, Release 2 (Standard for Trial Use) (see link at bottom of page)
HL7 DSS Implementation Guide, Release 1 (Standard for Trial Use) (see link at bottom of page)
HL7 DSS Implementation Guide, Release 1.1 (Standard for Trial Use Update 1) (see link at bottom of page)
HL7 FHIR Clinical Reasoning Module (includes initial FHIR representation of DSS functionality; being unified with CDS Hooks)
Contextualized potential drug-drug interaction clinical decision support (an implementation guide that specifies how to integrate EHR systems with CDS services for potential drug-drug interactionsusing CDS Hooks)