What is federated analytics?

A Trusted Research Environment (TRE) is a highly secure computer system where health data is stored. Approved researchers can access this data remotely by using secure logins and passwords. TREs are designed to be safe, allowing only authorised individuals to access the data. Data cannot be added or removed without proper permissions, ensuring transparency and accountability. Multiple sources of data can be combined in a TRE to create a comprehensive dataset for research.

Researchers can apply to access and analyse the data using computer programs they develop. While TREs are secure, organisations must still follow legal requirements when sending data to a TRE.

Federated analytics is a system where the data does not move, and instead the computer code the researchers write is sent to the data. Sometimes the dataset is made from lots of pots of data in different settings. For example, a dataset studying asthma might include data from several hospitals around the country. Instead of the hospitals sending all that data to one TRE (as described above), in federated analytics each hospital would build their own TRE and keep their data within that TRE.

In this situation, the researcher writes their computer code and this is then sent to each TRE, analysing only the data kept in that one TRE. The results from all the different TREs are then returned to the researchers and combined.