Skip to content
2026-07-02

Mapping to OMOP

The Cohort Discovery Service requires data to be formatted using the OMOP Common Data Model (CDM). This page explains the recommended approach for mapping source data to OMOP.

OMOP CDM

OMOP is used globally by academic and healthcare organisations to support observational health research. It is patient-centric, tabular, extendable, built for analytics, and has a relational design. For background, see the OHDSI OMOP CDM documentation.

Mapping source data to OMOP can be challenging. Separating the mapping logic from the ETL process allows non-technical subject matter experts to create mapping rules that are then applied by automated ETL tools.

graph LR
    A[1. Extract\ndata subset] --> B[2. Profile data\nWhite Rabbit]
    B --> C[3. Map to OMOP\nCarrot Mapper]
    C --> D[4. ETL to OMOP DB\nCarrot-CDM]
    D --> E[(OMOP Database\nin Secure Network Area)]

    style E fill:#3db28c,color:#fff

Figure 2 — Example architecture separating OMOP mapping from ETL

Architecture diagram showing the four-step process: extract data subset, profile with White Rabbit, map to OMOP with Carrot Mapper, and ETL to OMOP database with Carrot-CDM

Step-by-step

  1. Extract a subset of your raw identifiable data for OMOP mapping
  2. Profile the extracted data using White Rabbit to understand its structure
  3. Map the profiled data to OMOP using Carrot Mapper, generating a JSON mapping file
  4. Transform and load the extracted data using the mapping file via Carrot-CDM

Supported tools

Process Tool Description
Data Profiling White Rabbit OHDSI tool that scans source data and produces detailed information on tables, fields, and values. First step in the ETL process. Used for structural mapping.
Map to OMOP Carrot Mapper Web-tool that maps the White Rabbit scan output to generate a JSON mapping file defining the ETL guidelines for the dataset.
ETL to OMOP Carrot-CDM ETL tool that automates extraction of pseudonymised data, transformation to OMOP CDM, and loading to the query tool database.

Tools are not mandatory

These tools are not required. Other tools are also available, for example Usagi (OHDSI) for manual OMOP code mappings. HDR UK has worked with the CaRROT tools and can provide support for them specifically.

OMOP mapping support

HDR UK and partners at the Health Informatics Centre (HIC) at the University of Dundee offer OMOP mapping services using CaRROT tools. Commercial vendors also provide this service.

Get mapping support

Contact your HDR UK contact to discuss OMOP mapping support options.

OMOP CDM version support

Version Support status
OMOP CDM 5.4 Recommended for new mappings
OMOP CDM 5.3 Fully supported — no migration required
Other versions Contact HDR UK before proceeding

See the Bunny deployment requirements for the latest supported versions.

What fields are required?

The minimum OMOP fields required for Cohort Discovery are defined in the OMOP Requirements section.