Page tree
Skip to end of metadata
Go to start of metadata

Overview

  • Developed and run by OHDSI.
  • OMOP allows for systematic analysis of disparate observational databases. The concept is to transform data contained within those databases into a common format as well as a common representation (vocabulary), and then perform systematic analysis using tools that have been written based on OMOP.
  • Latest version is 6.0.
  • Goals
    • Conduct research to evaluate the performance of analytical methods on their ability to identify true associations and avoid false findings
    • Develop tools and capabilities for transforming, characterizing, and analyzing disparate data sources
    • Establish a shared research so the research community can collaborate
    • Open source community standard.



Uses

  • Supports the conduct of research to identify and evaluate associations between interventions and outcomes caused by these interventions.
  • Optimized for:
    • Identifying patient populations with certain healthcare interventions and outcomes
    • Characterizations of these patient populations for various parameters
    • Predicting the occurrence of these outcomes in individual patients
    • Estimating the effect of these interventions have on a population

Implementation

  • Data from UC Davis, UC Irvine, UC Los Angeles, UC San Diego, UC San Francisco, and the national Veterans Health Administration have been transformed into the OMOP CDM.

OMOP - HL7 and FHIR

  • A mapping exists between OMOP data elements and FHIR - http://build.fhir.org/ig/HL7/cdmh/profiles.html#omop-to-fhir-mappings
  • Georgia Tech has a FHIR project to build a FHIR resource on OMOP
    • OMOPonFHIR supports DSTU2 and STU3
  • Differences and Limitations:
    • Mandatory fields in OMOP vs. optional items in FHIR
    • Mandatory elements in OMOP with no relevant value in FHIR
    • Different codes for units
    • Extensions are used to insert data in FHIR
    • Fact_relationship tables are used in OMOP to show relationships between primary and secondary diagnosis


Model

  • Organized in tables
    • Clinical Data Tables - person, observation, visit, condition, death, drug exposure, procedure, device, measurement, note, survey, observation, specimen, fact relationship
    • Health System Tables - location, care site, provider
    • Health Economics Tables - payer plan period, cost
  • Is designed based on domains:
    • Domains are modeling in a person-centric relational data model - all clinical Event tables are linked to the PERSON table (except health system data tables)
    • Drug, device, procedure, condition, observation, measurement, spec anatomic site, meas value, specimen, provider specialty, unit, metadata, revenue code, type concept, relationship, route, currency, payer, visit, cost, race, plan stop reason, plan, episode, sponsor, meas value operator, spec disease status, gender, ethnicity, observation type
  • Source codes are maintained

Model Diagram: (https://github.com/OHDSI/CommonDataModel/wiki/Standardized-Clinical-Data-Tables)


OMOP Oncology

  • OMOP has a workgroup focused on Oncology.
    • Level of detail necessary for cancer studies is not regularly present in each data source.
    • Identification of cancer treatment regimens are complex compared to other diseases.
  • Representation of Cancer Diagnosis
    • Diagnosis: combination of histology and topography
    • Diagnostic schema: group of cancer diagnosis with similar diagnostic features
    • Diagnostic modifiers: features that define the cancer diagnosis (stage, grade, laterality, biomarkers, etc.)
  • Representation of Cancer Treatments
    • Treatments: higher-level procedure or regimen; capture the overall modality
    • Treatment modifiers: refine description of cancer treatment (number of fractions, treatment volume, total dose, etc.)
    • The Oncology workgroup has created a standardized ETL to ingest NAACCR data into the Oncology CDM Extension


Diagram: (https://github.com/OHDSI/OncologyWG/wiki/Cancer-Models-Representation)

OMOP Oncology and mCODE

  • Primary cancer condition diagnosis
    • OMOP rolls up histology and topography into one concept
    • mCODE has attribute for histology and topography but doesn’t roll them up. There is a separate PrimaryCancerCondition code.
    • OMOP has a grade modifier; mCODE does not.
    • Both OMOP and mCODE use SNOMED codes for condition.
  • Treatment
    • OMOP has more radiation and surgical modifiers
    • Both OMOP and mCODE use RXNorm for medication
  • Cancer Disease Status
    • OMOP records disease episode (first occurrence, recurrence, remission, progression)
    • mCODE has the CancerDiseaseStatus profile, SecondaryCancerCondition, and the status attribute.

References


  • No labels