Date: 5/27/2021

Quarter: Q5

CQI Hosted CDS

Chair: Paul Denning

Scribe: Patty Craig

Chairs present: Yan Heras

Agenda

Meet with BR&R:

Consider CQI-CDS comment on CMS NPRM: 

Meeting Minutes

  • BR&R project:  Pharmaceutical Quality Projects
    • Project Scope Statement
      • The PSS needs to be updated to represent that CDS will co-sponsor, CQI is an interested party
    • Community for this project is large pharma
      • Did a successful proof-of-concept with the industry for Quality Specification domain FHIR Profile (2019)
    • This project is focusing on the early portion of the drug project lifecycle
    • There are currently no standards for how companies send data to the FDA
    • Future submission of chemistry data within FHIR (Module 3)
    • Currently in Phase 1
      • Completed mapping 100% of Phase 1 PQ/CMC requirements to FHIR resources
      • Designed and developed 6 foundational FHIR Profiles
        • These are not the profiles that will be used for companies to send data.  That work will come later.
      • Using FHIR R4 + R5 binding because one element didn't make it into R4 
    • Next Steps
      • Updates to Phase 1 domains to cover additional use cases
      • Phase 2 - Structuring Drug Product Manufacturing domain
        • BR&R plans on reaching out to CDS during this phase.
      • Designing Profiles
    • Bryn raised a question about FHIR-27821 (PlanDefinition.action.definition[x]
      • .Smita Hastak (BR&R co-chair) will reach out to Peter Bomberg (the Reporter) so Bryn's question can be addressed
  • BR&R project:  CDISC Mapping FHIR IG
    • Joint project with HL7 FHIR and CDISC (Clinical Data Interchange Standards Consortium).
      • Purpose:  Leverage EHRs built on FHIR R4/R5 and extrapolate, either directly or in an intermediary mechanism, into a sponsors clinical trial system (e.g., EDC, MDR, etc.). Helping to more efficiently access and integrate EHR-based sources of data for clinical research.
      • Focus is on the pre-clinical workspace.
    • CDISC is a standards development organization
    • CDASH
      • Clinical Data Acquisition Standards Harmonization
      • A standard way to collect data consistently across studies so data collection and formats and structures provide clear traceability
      • Metadata support data capture for SDTM datasets
      • Purpose
        • Provide consistent questions for each data domain
        • Questions worded the same so they always mean the same thing
        • Answers use standard controlled terminology
        • Database variable names are harmonized
    • What is the SDTM (Study Data Tabulation Model)
      • Standard Metadata and associated rules
      • A standard way to organize, format and describe study data
      • Domains for FHIR to CDISC
        • Adverse Event (AE)
        • Concomitant Medications (CM)
        • Demographics (DM)
        • Laboratory (LB)
        • Medical History (MH)
        • Procedures (PR)
        • Vital Signs (VS)

      • FHIR Resources

        • AdverseEvent

        • Condition

        • MedicationStatement

        • MedicationRequest

        • MedicationDispense

        • MedicationAdministration

        • Immunization

        • ResearchStudy

        • ResearchSubject

        • Patient

        • Specimen

        • Diagnostic Report

        • ServiceRequest

        • Observation

        • Practitioner

        • Encounter

        • Organization

        • Allergylntolerance

        • Procedure

        • BodyStructure

      • Variable Roles
        • Identifier - Identifies the study, the subject involved in the study, the domain, and the record sequence number
        • Topic - Contains the focus of the observation
          • Domains are logical groups of data based on the Topic
        • Qualifier - Provides more information about the Topic
          • Group, Result, Synonym, Record, Variable
        • Timing - Describes the timing of the observation
          • Visits, Date/Times, Relative Times
    • CDISC/ BR&R's Next Steps
      • Balloted in January 2020
        • Have 1 tracker item left to determine a resolution (FHIR-30153 - review CDISC mappings with respect to QI-Core (http://hl7.org/fhir/uv/cdisc-mapping/2021JAN/overview.html))
          • Tracker request:  Suggest the team review the [Ql-Core R4| http://hl7.orq/fhir/us/qicore/profiles.html1 since the issues identified in CDISC mapping are similar to those in mapping existing electronic clinical quality measure model (Quality Data Model - QDM) concepts to FHIR including adding constraints specific to the measurement and clinical decision support workflows. Note that much of the information needed is available as Observations, VitalSigns, Procedures, MedicationRequest, MedicationAdministration, ServiceRequests, AdverseEvents, Allergylntolerances etc. with adding extensions to allow reference as .partOf ResearchStudy or ResearchSubject. CDS and Measurement had similar issues with determining procedures that are non-invasive and as basic as "evaluated for". In most cases we had similar assignments to observations (l.e., findings or conditions identified based on those evaluations). Recommend discussion with CQI and CDS to evaluate similarities in mapping for CDISC, CDS and eCQMs to assure consistency in the practice environment and to reduce burden by creating a single path and to assure the metadata requirements are aligned.
          • BR&R / CDISC: They did the review and didn't find any changes that would be needed to align with QI-Core.
          • This scope of this project was about the clinical data.
          • There is another Vulcan project that is looking at scheduled activities.


  • CMS Proposed Rules that name FHIR
    • CMS-1752IPPS Annual Proposed and Final Rules, and Relevant Correction Notices: Fiscal Year 2022
    • CMS Newsroom: Fiscal Year (FY) 2022 Medicare Hospital Inpatient Prospective Payment System (IPPS) and Long Term Care Hospital (LTCH) Rates Proposed Rule (CMS-1752-P)
    • HL7's Policy Committee will be sending out an email shortly with information on when they need the review to be returned.
    • Among other items, CMS is asking for comment on:
      • A standardized definition of Digital Quality Measures (dQMs):
        • Digital Quality Measures (dQMs) are quality measures that use one or more sources of health information that are captured and can be transmitted electronically via interoperable systems. A dQM includes a software that processes digital data to produce a measure score or measure scores. Data sources for dQMs may include administrative systems, electronically submitted clinical assessment data, case management systems, EHRs, instruments (for example, medical devices and wearable devices), patient portals or applications (for example, for collection of patient-generated health data), health information exchanges (HIEs) or registries, and other
          sources.
        • NCQA worked with CMS on the first sentence of the above definition.  They are unaware as to whom CMS talk to derive the rest of their definition.
          • NCQA's definition:
            • Digital quality measures (dQMs) are quality measures expressed in a digital format using highly standardized language and data definitions that enable sharing of the fully specified measure electronically between systems.
            • dQMs: Use a standards-based interoperability format including machine interpretable measure logic (e.g., CQL) and data model (e.g., FHIR) Incorporate data concepts/terms (e.g., value sets) required to fully execute the measure
            • Key Characteristics of dQMs
              • dQMs may utilize a broad array of data from multiple electronic sources including, but not limited to, EHRs, registries, case management systems, HIEs, wearable devices and administrative claims.
              • Electronic clinical quality measures (eCQMs) use data derived from electronic medical records and are a subset of dQMs. A measure can be digital even if the electronic data it uses are generated through manual processes (e.g., G-code, case management system data entry).
              • Key Potential Benefits of dQMs The digital format, standardized language and quality assurance processes used to author dQMs mitigate the potential for faulty interpretation of paper-based specifications and errors associated with manually coding narrative measure descriptions. dQMs reference standard data collected in the normal course of care and perform many of the measure calculation functions that previously required additional processes. dQMs use of standardized data can improve accuracy and allow for more rigorous data validation to occur at different levels of the data collection process.
              • dQMs can be designed to generate clinically relevant patient-specific quality insights based on available clinical data at the point-of-care. This is not the case with current measures which generally provide information about what is best for an “average” patient and often are not implemented to generate timely, actionable information.
        • Discussion
          • It is an unnecessarily restrictive definition
      • Implementing dQMs that are "self-contained tools." That is, CMS is interested in promoting software solutions for dQMs that could, among other things:
        • Support the calculation of single or multiple quality measures;
        • Obtain data via automated queries from a broad range of digital sources (initially EHRs, but potentially also from claims data, patient-reported outcomes, and patient generated health data;
        • Generate measure score reports;
        • Be compatible with any data source;
        • Exist separately from data source systems;
        • Be tested and updated independently of data source systems;
        • Operate in accordance with health information protection laws and regulations;
        • Be deployable by hospitals, health IT vendors, health plans and/or CMS;
        • Be usable by non-technical end users; and
        • Have the ability to adopt to emerging advanced analytic approaches like natural language processing.
      • Closing the gap in health equity
        • Including, future potential stratification of quality measure results by race and ethnicity

Action Items

  • None