(199) Proof-of-concept for a UK federated data network using the OMOP common data model: the UK regulatory studyathon

Background: The UK is a rich source of real-world data (RWD) and offers many possibilities for supporting and improving the timeliness and robustness of regulatory decisions for medicines and medical devices. However, significant challenges remain around data availability, access, quality, and heterogeneity. Common data models (CDM) have been used to standardise real world datasets and facilitate research across them through federated analytical approaches.

Objectives: In 2023, a `regulatory studyathon’ was held to explore use of the Observational Medical Outcomes Partnership (OMOP) CDM as applied to UK healthcare datasets through consideration of two priority regulatory-focused questions.

Methods: The ‘studyathon’ was a week-long event attended by multiple UK data partners. The study questions explored prescribing of fluoroquinolones before and after regulatory actions taken to minimise the risk of rare but serious adverse events, and the epidemiology of rectal prolapse and rectopexy using implantable mesh. Common analytic code was developed prior to the ‘studyathon’ and refined during the week.

Results: Six UK databases, representing over 60m patient lives and including primary/secondary care plus paediatric/adult services, participated in the ‘studyathon’: Barts Health, Clinical Practice Research Datalink (CPRD) Aurum linked to Hospital Episode Statistics (HES), CPRD GOLD, Great Ormond Street Hospital, Health Informatics Centre, and Lancashire Teaching Hospitals. Results were produced for both studies, including a description of fluoroquinolone prescribing trends in secondary care for the first time. Differing approaches to OMOP conversion had been utilised by in-house or external collaborators, depending on existing resource and expertise. Data on mesh devices, including unique device identifiers (UDI), were recorded for administrative rather than clinical purposes, but it was possible to incorporate these into the CDM during the week. Interpretation of the results was complex given the different care settings and localities.

Conclusions: The regulatory ‘studyathon’ was successful in bringing together data partners from across the UK to answer the selected questions. The need for consistency in the OMOP conversion process was highlighted, as well as the importance of local context in interpreting the results. Opportunities for exploring device safety were uncovered, particularly given inclusion of UDI’s into the CDM. This could be enhanced by supporting clinicians in collecting device data systematically. The ‘studyathon’ demonstrated the feasibility of a UK federated data network using OMOP CDM for providing evidence to support regulatory decision-making.