Toolbox Strategic Focus Area 3
Develop a federated data network to allow group level queries on harmonized datasets
The ultimate goal of this strategic focus area is to develop a federated data network to allow group level queries on harmonized datasets.
The MSDA envisions to develop and implement a “federated data network”, which aims to achieve insights without the need to “ship” data to a centralized data source. It allows local querying of different cohorts or registries. The data does not leave the local storage unit, and the results of those queries are shared (=so not the actual patient-level data).
A federated data network has several advantages:
- data remains under the control of the data owner
- locally required legal and ethical approvals apply
- no patient-level data leaves the owner’s site, only aggregated counts, thereby ensuring patient privacy
- GDPR – ‘Privacy by Design’
- the analysis is “brought to the data” rather than creating a central data repository
- use of common data model allows for efficient search / analysis across multiple data sets
- requires close collaboration with data owners, which builds trust
We will continue the development of the federated structure to improve end-user experience and to reduce data custodian workload. First of all, we want to deliver a prototype of the improved MSDA infrastructure using the use case “automated cataloguing of COVID-19 variables”.