Skip to end of metadata
Go to start of metadata

Skip to end of metadata


Go to start of metadata

The CQI and CDS Workgroups need to determine the benefits, challenges, work efforts, tooling and cost-related requirements to maintain a FHIR version-specific QI-Core data model for authoring and implementing eCQMs and clinical decision support artifacts Vs a version-neutral, logical data model (QUICK) that can output version-specific content as needed.

  • Diagrams showing the different paths included in descriptions of QI-Core/QI-Core logical view and QUICK logical model sections below.
  • Please list use cases for QUICK logical data model here:
    • Use case 1: 
    • Use case 2: 

Steps to Add/Enhance

Current build version of QI-Core for FHIR R4: http://build.fhir.org/ig/HL7/fhir-qi-core/index.html

Current QI-Core logical view (build off of QI-Core): http://build.fhir.org/ig/HL7/fhir-qi-core/quick/QUICK-index.html

Current QDM to QI-Core mapping guidance: http://build.fhir.org/ig/HL7/fhir-qi-core/qdm-to-qicore.html

Current status:

History of data model for quality (measures and clinical decision support):

QI-Core and QI-Core logical view

  1. QI-Core Implementation Guide: STU 4 (v4.0.0 for FHIR 4.0.1) completed and in process of publication.
    1. Developed by consensus with the FHIR community, CQI and CDS Workgroup efforts
  2. Existing eCQM work used FHIR R4 directly with guidance from the QI-Core content, but tooling had not been available to author directly with QI-Core content.
    1. There is a QI-Core logical view for each version of QI-Core (STU 3 and R4)

  3. Tooling in progress to author QI-Core content and use the constraints, extensions and profiles directly is now available under the heading, QI-Core Logical View.
    1. Gives an ‘authors level’ perspective of the data model

      1. Allows measure authors to choose the entire profile and specify only the items specific to the measure; the rest of the elements are populated automatically (e.g., status)

      2. Streamlines authoring with QI-Core using CQL

      3. Output in CQL shows QI-Core authoring view; ELM show fully expressed QI-Core details
    2. Testing planned for the May 2020 HL7 Connectathon

Example:

QUICK Logical Model

  1. QUICK prototype developed by Claude Nanjo at University of Utah tested in September 2019 and February 2020 HL7 Connectathons
    1. Developed top down and can map to various versions of the FHIR model, or other models
      1. A logical model informed by the Quality Improvement Domain Analysis Model (QiDAM) that, in turn is based on QDM (quality measure model), and vMR (clinical decision support model)
      2. Not dependent on FHIR versions
      3. Currently in early prototype and testing in HL7
      4. Possibility to extend beyond quality measure use cases beyond eCQMs (i.e., expand beyond an EHR-only phenomenon). The NCQA Digital Quality Framework has coined a term - digital quality measures (dCQM) - using multiple data sources for reporting - not only EHRs.  eCDS - electronic clinical data system reporting - deployed in HEDIS to promote more comprehensive quality measurement and clinical decision support. 

Further QUICK logical model testing pilot planned for May 2020 HL7 Connectathon


Current status

  • eCQM conversion from QDM to FHIR uses FHIR directly, informed by the QI-Core guidance.
  • QI-Core Logical View (formerly QUICK logical view)
    • Enables direct authoring using QI-Core
    • Uses established tooling, currently in place
    • May streamline authoring of eCQMs
    • Provides CQL output for human readable; ELM output with detailed QI-Core specification to run against FHIR servers, no extra step for implementers
  • QUICK Logical Model may have value to help with FHIR version issues
    • May reduce burden on implementers with various versions of FHIR implemented at the same site
    • Requires constraining to US Core to avoid implementer burden
    • Requires continued manual mapping to each new FHIR version to create a QI-Core for that version
    • Requires tooling to perform translation to each version of QI-Core for measure implementation output (not yet developed – potential cost and sustainability concern)
    • Requires determination of level of effort and cost (anticipated to be large)
  • Both QI-Core logical view and QUICK logical model will be tested in HL7 Connectathon in May 2020.

Expectations of Data Model Steward

  • work with FHIR and US Core to address findings related to existing resources, manage profiles and implementation guides
  • manage HL7 Jira tickets in a timely manner
  • support ongoing maintenance of the model
  • assure availability and maintenance of required tooling 

Describe benefits and challenges of version-specific Vs version-neutral logical data model efforts

This step is where the project team seeks input from CQI and CDS. WG members please fill in the table below

  • What is the effort to author eCQMs or CDS using a version-specific data model closely tied to FHIR versions
    • FHIR - note that all eCQMs converted to FHIR to-date have expressed CQL with FHIR resources informed by QI-Core and the QDM to QI-Core mappings
    • specifically using QI-Core - Tooling is soon available (and will be tested in the May HL7 Connectathon) to allow authoring eCQMs directly with QI-Core logical view
  • What is the effort to author eCQMs or CDS using a version-neutral logical model (QUICK)

BenefitsChallenges
FHIR

QI-Core

QUICK

Describe manual effort requirements for version-specific Vs version-neutral logical data model efforts. 


Update Frequency (Input on data model)When US Core Updates (within the same FHIR version)When FHIR version updatesOther (specify)
FHIR



QI-Core



QUICK




Principles for tooling to maintain data model (general - not specific to version-specific or version-neutral)

  1. Open source, publicly available
  2. Fully funded
  3. Assured maintenance
  4. Sustainable

Describe tooling requirements for version-specific Vs version-neutral logical data model efforts. 


Update Frequency (Input on data model)When US Core Updates (within the same FHIR version)When FHIR version updatesOther (specify)
FHIR



QI-Core



QUICK



Describe tooling availability for version-specific Vs version-neutral logical data model efforts. 


Update Frequency (Input on data model)When US Core Updates (within the same FHIR version)When FHIR version updatesOther (specify)
FHIR



QI-Core



QUICK



Pilot Activities / Commitment to Implement

Show that some number (3 or more?) of implementers have implemented or "are committed" to implement.  This needs to implementation in a production environment, i.e not just tested and plans for the future.




  • No labels