Blueprint for harmonising unstandardised disease registries to allow federated data analysis: prepare for the future

Abstract

Real-world evidence from multinational disease registries is becoming increasingly important not only for confirming the results of randomized controlled trials, but also for identifying phenotypes, monitoring disease progression, predicting response to new drugs, and early detection of rare side effects. With new open access technologies, it has become feasible to harmonize patient data from different disease registries and use it for data analysis without compromising privacy rules. In this article, we provide a blueprint for how a clinical research collaboration can successfully use real-world data from existing disease registries to perform federated analyses. We describe how the European Severe Asthma Clinical Research Collaboration SHARP fulfilled the harmonization process from non-standardized clinical registry data to the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) and built a strong network of collaborators from multiple disciplines and countries. The blueprint covers organizational, financial, conceptual, technical, analytical and research aspects and discusses both the challenges and the lessons learned. All in all, setting up a federated data network is a complex process that requires thorough preparation, but above all, it is a worthwhile investment for all clinical research collaborations, especially in view of the emerging applications of artificial intelligence and federated learning.

Exploratory pharmacovigilance with machine learning in big patient data: A focused scoping review

“…37 Data capture will inevitably differ from one medical/cultural/social context to the other, but casting such idiosyncratic data into some common data model will facilitate exploiting several (or even many) datasets for exploratory pharmacovigilance. 38,39 Interestingly, only a single study used a standardized format for the data, and none of the studies used data from multiple healthcare systems (e.g., different countries). Collating data from multiple sources is non-trivial and comes with many issues and challenges, including legal and ethical ones, when patient data flow between countries, inevitably increasing the risk of leaks or even misuse.…”

“…Two somewhat parallel processes are gaining much traction in these years: the learning healthcare system 36 and the rise of real‐world evidence, both in turn relying on fit‐for‐purpose data 37 . Data capture will inevitably differ from one medical/cultural/social context to the other, but casting such idiosyncratic data into some common data model will facilitate exploiting several (or even many) datasets for exploratory pharmacovigilance 38,39 . Interestingly, only a single study used a standardized format for the data, and none of the studies used data from multiple healthcare systems (e.g., different countries).…”

Evaluation of real-world mepolizumab use in severe asthma across Europe: the SHARP experience with privacy-preserving federated analysis

“…The process of, and experiences with, the registry mapping, development and implementation of the FAP has been described separately [ 24 ].…”

Personal data in biomedical research

“…Protecting the privacy of the participant can be achieved by good collaboration between researchers, clinicians, information technology specialists, legal bodies and developers of good data management systems with good interoperability support. 6 The stored data can be shared within the institute where the patient is currently taking treatment and can be shared with family or across the organisations. The fundamental issues with data sharing are efficient communication, informed consent, establishing patient confidence and minimising the risk to the patient.…”