Date: Thursday 9/24/2020

Time: Noon-1:30pm ET

Chair: Bryn Rhodes

Scribe: Howard Strasberg

CDS Hosted CQI

Agenda brought forward by CQI

Meeting Minutes

  • Joint CDS/CQI Minutes
  • CPG-on-FHIR - Bryn
    • Quality improvement ecosystem requires standards for data and knowledge
    • Example - antenatal care guideline and related quality measure
    • EBM-on-FHIR used to convert research into guidelines
    • CPG-on-FHIR used to convert guidelines into CDS
    • EBM-on-FHIR and CPG-on-FHIR are both specification IGs (as opposed to model IGs or content IGs)
    •  CPG-on-FHIR is a Gemini project - close collaboration with IHE CCG Profile
    • Scope of CPG-on-FHIR
      • knowledge artifacts
      • data elements
      • activities - e.g. order test, prescribe medication, provide counseling
      • libraries - reusable logic
      • recommendations - e.g. avoid co-prescribing opioids and benzodiazepines
    • L1-L4 knowledge representation framework
    • L3 + T3 (forms) = Questionnaire (SDC)
    • L3 + T2 (logic) = Library (CQL) + ActivityDefinition + PlanDefinition
    • L3 + T1 (data) = CodeSystem + ValueSet + StructureDefinition + Measure
    • L1→L4 uses an agile approach, not a waterfall approach
    • WHO digital accelerator kits are examples of L2
      • working with WHO to turn some of these examples into L3
    • PlanDefinition - ECA rule
      • event represented as trigger
      • condition represented as condition
      • action represented as ActivityDefinition
    • service responds with CarePlan with RequestGroup
    • knowledge artifact can be implemented as a service or locally
    • Project status
      • All comment reconciled
      • 80% applied
    • Question about EBM Plus model 
      • Alignment being evaluated
    • Matthew Burton - Covid19 digital guideline working group
      • example of CPG-on-FHIR
      • forced to do concurrent development of narrative and computable guidelines
      • created ECA rules for CDS
      • can also get patient-specific, practice-level metrics (eCQM)
      • eCase report (registries)
      • agile approach
        • guideline development group works closely with knowledge engineering team
        • localization for setting specific factors at each implementation site
        • e.g. ED severity score and disposition guidance (ACEP)
        • cross-functional, integrated team
        • "chunking" (composite artifacts)
        • iteration, rapid-cycle feedback
        • test-driven knowledge engineering
        • leverage knowledge base
      • BPM+ tools were used by the knowledge engineering team
        • translated into CPG knowledge architecture and profiles
      • work was part of Covid19 healthcare consortium
  • EBM-on-FHIR 
    • https://tinyurl.com/COKAfolder
    • no universal standard to make evidence FAIR
      • EBM-on-FHIR is a project to solve this problem
    • Covid-19 Knowledge Accelerator (COKA) - virtual organization 
    • early stages
      • standard format
      • common terminologies
      • tools for initial implementation
        • move from digital data exchange to computable digital exchange
      • adapting tools for real-world implementation
      • develop an implementation guide
        • learn from real-world steps before writing the IG
    • four resources at maturity level 1
      • Evidence, EvidenceVariable, Statistic (dataype), OrderedDistribution (datatype)
    • two resources at maturity level 0
      • Citation, EvidenceReport
    • Brian reviewed an example of an Evidence resource
    • Evidence and EvidenceReport are used to share the evidence
    • use Group resource to define the populations and the EvidenceVariable resource to define the exposures
    •  code systems and value sets
      • created 35 value sets (and thus 35 code systems)
      • the process wasn't set up for the creation of code systems
      • created a code system development protocol
      • currently working on 4 code systems
        • study design
        • statistic type
        • statistic model code system
        • risk of bias code system
      • people may join any of the expert groups
    • demonstrating that data exchange can work
      • map between MEDLINE elements, CORD-19 JSON elements and FHIR Citation elements
      • goal is to support interoperable, reusable efforts for Citation repositories
      • Systematic Review Data Repository (SRDR) sends outcome definition data as FHIR EvidenceVariable resource in JSON
      • project in discussion to develop tools for data entry for clinical trial reports
    • Floyd: question about metadata to establish trust in the data
      • not part of the workstream
  • Integrated Trial Matching for Cancer Patients and Providers Project Discussion
    • Steven Bratt
    • PSS-1659 - Getting issue details... STATUS
    • mCODE - minimal Common Oncology Data Elements
      • FHIR-based core set of data elements for cancer
    • http://hl7.org/fhir/us/mcode/
    • CodeX - HL7 FHIR Accelerator
      • Use Case 2 is integrated trial matching for cancer patients and providers
    • Use Case - patient or provider searching for an offsite clinical trial
    • use mCODE data to match clinical trials
    • working on how to structure the clinical trial eligibility criteria 
    • Phase 1 - retrospective study
    • Discussion
      • alignment with EBM-on-FHIR resources?
      • ResearchStudy (BRR WG)
      • Group resource could be useful to describe the group
      • is there a role for the Cross-Group Projects Workgroup?
      • best home might be BRR

Action Items

  • None