TREEHOOSE - an Open Source, Cloud-based Trusted Research Environment

Analysis of sensitive data for the purpose of national statistics, social policy and healthcare research is challenging as it requires significant governance and ethical oversight. Where and how the analysis is performed is controlled by the data controller as the data cannot be freely shared often resides within a monolithic computing environment. These computing environments are referred to as Data Safe Havens (DSHs), Trusted Research Environments (TREs) or Research Data Centres (RDCs). TREs may not have all the usual tools a data scientist or analyst is used to and may have limited computing resources. Supporting these environments by the data controller is resource intensive. The public cloud provides access to powerful, flexible and scalable computing resources based on the needs of the user. There are no capital costs and maintenance requirements are reduced. The Health Informatics Centre (HIC) has >10 years experience of managing a TRE and we have developed an open source – infrastructure-as-code – Trusted Research Environment (TRE) toolkit which can be deployed in the public cloud for a safe, secure and scalable computing environment suitable for working with clinical patient data or other sensitive data.