Health Data Research UK is building a national pilot Real-World Evidence Network. The aim is to demonstrate the potential of the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) to enable efficient cross-national federated research studies, using health data derived from multiple sources and importantly, from diverse healthcare settings. (For more on OMOP see our Common Data Models page). This pilot will serve as proof-of-concept for a sustainable UK Network to generate real-world evidence to improve UK-wide population health outcomes.
The Network is led by Daniel Prieto-Alhambra, Ed Burn, and Martí Catala Sabate from NDORMS at the University of Oxford, working in partnership with HDR UK.
The seven successful applicants funded to join the Network in its pilot phase are:
Barts Health NHS Trust provides healthcare for more than 2.3 million people in East London, across four hospitals and including people from many different ethnic and socioeconomic backgrounds. Their data includes over 10 years of information on hospital services like heart care, cancer treatment, children’s healthcare, and trauma services.
CPRD collects anonymised patient data from 1 in 3 GP practices across the United Kingdom representing 29% of the population. The CPRD Primary Care OMOP Common Data Model (encompassing 47,193,158 patient lives and 16,189,976 registered patients) and CPRD Primary Care and Linked Data OMOP Common Data Model (encompassing 34,146,603 patient lives and 9,907,027 registered patients) provide a longitudinal, representative UK population health dataset. CPRD curates the data and undertakes relevant quality assurance, so that their data are fit for purpose in a wider range of health-related applications (Mahadevan et al., 2022; Shiekh et al, 2023).
The DataLoch data service is a Trusted Research Environment that covers South-East Scotland, with a population of 850,000. The data includes a specialised cancer database curated by the Edinburgh Cancer Informatics team within the University of Edinburgh and a General Healthcare dataset.
GOSH is a national centre for serious and rare disease in children and thus their data is unique and able to offer insights that no other centre can provide. Their data set comprises all their Electronic Patient Data since 2019 covering 149,000 patients.
LTH is a digitally mature secondary care provider, major trauma centre and multi-specialty tertiary referral centre in Lancashire and South Cumbria ICS (LSC). LTH have access to routinely collected healthcare data for over 2.25 million patients spanning 15 years, covering most aspects of secondary care.
LTHT is one of the largest NHS trusts in the UK and has a strong reputation in leading medical research and keeps detailed electronic health records, laboratory and pathology results, imaging data, and pharmacy records. A large amount of data for patients with cancer (MEDOC – Minimal Essential Description of Cancer) includes important information about demographics, clinical features, biomarkers, treatments, and health outcomes.
The UCLH patient population includes national referrals for tertiary and quaternary services (cancer, neurology etc.) and general medical admissions from an inner-city teaching hospital that treats more than a million outpatients per year and has more than 100,000 inpatient admissions. Their data contains records for 6 million patients, including 13 million diagnoses and 50 million medication events.
Each data partner will receive milestone payments for onboarding to the network, and further payment for each successfully completed study.
After characterising the contributing data assets, the Network will conduct two exemplar studies during its pilot phase:
In addition to policy-relevant evidence generation, the proposed exemplar studies will help prioritise future developments for the network. Further updates will be provided as the studies progress.
Q. Why are you only using OMOP? There are other forms of Real-World Evidence that could be used.
A. The aim is to conduct large studies by combining data from multiple sites. This requires a uniform data structure so that data collected in different health care settings, including different tiers of healthcare, can be included. OMOP is the most widely used common data model for observational health data worldwide and is supported by widely available software, a network of EHDEN certified SMEs, and a vibrant user community convened by OHDSI.
Q. How were the data partners and studies selected?
A. An open call was issued on 11th July 2024. Data partners’ applications were scored by a review panel comprising representatives from MHRA, UKHSA, NICE, NHSE Data for R&D Programme, NDORMS, HDRUK, and a representative of the patient and the public voice from use MY data. This review panel also co-designed the use cases.
Q. How is data security ensured?
A. The Network complies with the “Five Safes” approach to data privacy developed by the Office for National Statistics. Specifically;
For any further questions or comments, please contact the Alliance: UKAlliance@hdruk.ac.uk
Related Links
