This project will create guidance for bidirectional flow of queries and responses written in OMOP with Atlas expressions to FHIR with CQL using Digital Quality Measure (dQM) as an example. The goal is to establish common mechanisms for developing and sharing population cohorts across research, registry, and quality improvement (measurement and clinical decision support), and related stakeholder groups. The research community develops, validates, and assures reliability and feasibility of cohort definitions, yet retrieving data to match that definition requires significant, time consuming, curation. For the example, a hypothetical dQM cohort measure will be used as the content (such measure consistent with a COVID19 research cohort). Enabling translation of the dQM cohort measure from OMOP and Atlas to FHIR-CQL will allow queries for existing data based on standard FHIR and related profile definitions such that curation efforts should become less cumbersome. Similarly, the quality and registry communities have challenges assuring reliability and validity of cohort definitions developed independently and such definitions may require duplicate effort for implementers for cohorts defined separately by researchers, registries, and quality improvement stakeholders. The goal is to enable harmonization, providing a common source and reliable, valid, cohort definition for all three communities to query for required data in a common mechanism that will simultaneously reduce clinical care implementer burden for extracting useful data. The Power Point figure shows the overall strategy and goals [Power Point Diagrams| https://confluence.hl7.org/rest/hotovo/amazon-s3/1.0/buckets/11/content/download?path=OMOP_dQM_Project_Proposal_23February2022.pptx&targetId=86974054 ]: 1. Allow expression of cohort dQM consistent with a research and registry dataset/phenotype definition by any stakeholder which may include (a) value sets existing outside of the OMOP data model, and (b) criteria to express which instances represent the cohort desired 2. Enable harmonization / mapping of value sets (terminology efforts) 3. Enable translation of desired information to OMOP data model through common data model harmonization (CDMH) 4. Create expressions in Atlas 5. Map the Atlas expressions to FHIR-CQL 6. Extract data from FHIR server with patient data to create the data set 7. Analyze the dataset in an analytic engine addressing the FHIR-CQL retrieves directly or such retrieves mapped back to OMOP Goal: 1) Federated / aggregated results 2) Potentially sharable aggregate results - e.g., counts of numerator, denominator, or counts of those meeting cohort definition (e.g., cohort 1: All patients with COVID diagnosis, cohort 2: All patients meeting matched control criteria without COVID diagnosis) Rationale: 1) In research: cohorts are well defined and require significant curation to capture and analyze data; researchers also validate their datasets and address reliability and feasibility issues with respect to the data requested and retrieved. The ability to convert phenotype definitions and retrieve data directly from information captured as part of the care delivery process will significantly reduce the effort and time required to curate the data. 2) Collaborating with research phenotype definitions provides validated and reliable data to evaluate clinical outcomes with respect to effectiveness, timeliness, efficiency, and safety. 3) Identification of process data requirements to evaluate quality care delivery and clinical decision support may inform potential updates to the OMOP data model. Proposed activities: 1) Develop a framework for managing clinical data definition and extraction from clinical and financial data captured as part of the care delivery process 2) Identify examples of existing data sets to evaluate mechanisms to:     a) translate defined terminology to OMOP code sets; examples:          i) CDC COVID co-morbid conditions in VSAC convert to NC3/OMOP CDC co-morbid conditions) - Johns Hopkins BIDS          ii) HCUP/NIS Conditions/Procedures (ICD10/ICDPCS codes in CCSR groups) to HCUP value sets transformed to OMOP          iii) Possible - Additional dQM preventive care related denominator cohort with value sets in VSAC transformed/harmonized to OMOP data model and code sets - in later phases: N3C /NCATS Consortium: Johns Hopkins, Columbia, Tufts, etc. 3) Define phenotype expressions in Atlas using OMOP data model and code sets and SQL expressions to meet criteria to FHIR-CQL queries using dQM example     a) Query OMOP data store AND FHIR data store to determine concordance with results: Potential collaborators:            i) Phenotype Workstream of N3C Consortium for: N3C COVID Phenotype           ii) Potential participant for Quality Measure phenotype (TBD)     b) Create OMOP expressions and convert to FHIR-CQL queries: expression requirements          i) Atlas query development (TBD)         ii) Covert to FHIR/CQL (Possible participation by Paul Nagy)        iii) Possibly interrogate FHIR and OMOP data stores to compare results: Andrew Williams (Tufts), and Syntegra - may have such examples by April 2022 4) Identify data query and response requirements:             i) Tufts/Syntegra data store (OMOP) (still to be validated against N3C phenotype)           ii) NCATS, CDMH2 (FDA) FHIR data set query with CQL (SDTM endpoint, requires permission, issue: cohort definition requires multiple queries / filters not compiled on FHIR server, FDA Use Case)          In Later Phases:              iii) N3C, NCATS Consortium: Johns Hopkins University, Columbia University, Tufts, etc.              iv) Colorado University, GAG4, Clinical Genomics Workgroup   Plans (tentative, dependent on availability of participants and data requirements, but likely timing provided): 1) HL7 FHIR Connectathon May 2022: Mapping terminology (VSAC, external) to OMOP code sets; Creation of Atlas queries and convert to FHIR-CQL 2) Connectathon during NCQA/HL7 Digital Quality Summit - Use FHIR-CQL query to retrieve data from a FHIR data store 3) HL7 FHIR Connectathon September 2022: compare concordance of OMOP/Atlas query to OMOP data store with FHIR-CQL query to FHIR data store 4) Publish informative document for ballot January 2023 using examples identified in the Connectathons 5) Reconcile informative document; may require additional HL7 Connectathons to further evaluate new requirements suggested by ballot comment 5) Further iteration dependent on issues identified.