The MIREDA Partnership is a UK-wide collaboration that harmonises electronic birth cohorts into a common data model to enable federated, reproducible life-course research across nations and Trusted Research Environments (TREs).
The partnership brings together major maternal and child health datasets from England, Wales, and Scotland, transforming them into the OMOP Common Data Model (CDM) to support consistent, cross-cohort analyses while data remain securely within their host environments.
Together, the harmonised cohorts represent over 18 million births, enabling research into rare exposures, policy impacts, and long-term health outcomes across the life course.
MIREDA is funded by the UK Medical Research Council and brings together five major electronic birth cohorts:
- Born in Wales
- Born in Scotland
- Born in Bradford
- Born in South London (eLIXIR)
- Clinical Practice Research Datalink (CPRD) birth cohorts
Each cohort remains within its own secure TRE, but harmonisation into OMOP enables:
- Federated analysis without sharing raw data
- Reproducible analytical pipelines
- Cross-nation and international comparisons
- Large-scale studies of rare exposures and outcomes
This GitHub organisation provides open technical resources developed by the MIREDA partnership, including:
- Standardised mapping specifications
- Carrot Mapper csv files
- Metadata and vocabulary mapping examples
- Cohort-specific and harmonised mappings
While most transformation to OMOP CDM was processed using the carrot-transform software, additional post-processing scripts for the purpose of creating unsupported domain tables and remedying limitations of the software will be found here.
- Post-processing and quality-assurance scripts
- Helper tools for handling OMOP limitations
- Reproducible study pipelines
- Example federated analyses
- Cohort-comparable study definitions
- Technical documentation
- Harmonisation decisions
- Data structure guidance
- OMOP pregnancy modelling approaches
- Mother–infant linkage solutions
- Vocabulary mapping challenges
- Practical solutions from real-world cohort harmonisation
The harmonised MIREDA datasets are accessed through the Health Data Research UK Gateway:
https://healthdatagateway.org/en/collection/119
Data are not held centrally. Each cohort:
- Remains within its own Trusted Research Environment
- Is governed by local data controllers
- Requires separate access approval
Federated analysis enables the same code to run across all sites without moving sensitive data.
-
Explore repositories for:
- Mapping specifications
- ETL pipelines
- Analytical examples
-
Review documentation for:
- OMOP structure in birth cohorts
- Mother–infant linkage methods
- Harmonisation decisions
-
Apply for data access through the HDR UK Gateway if you wish to run analyses.
If you use MIREDA resources, please cite:
Seaborne et al.
UK electronic birth cohort data harmonisation using a common data model
(Final publication details to be added once available.)
Funded by the UK Medical Research Council (MRC)
Partnership Grant: MR/X02055X/1
Additional support from:
- Health Data Research UK
- NIHR Birmingham Biomedical Research Centre
MIREDA Partnership
Centre for Population Health
Swansea University Medical School
Project pages:
https://gtr.ukri.org/projects?ref=MR%2FX02055X%2F1
https://centreforpopulationhealth.org/projects/mireda/
Unless otherwise stated, code in this organisation is released under an open-source licence.
See individual repositories for licence details.