Analysis requests are sent from a central analysis machine to several data-holding machine storing the harmonised data to be co-analysed. The data sets are analysed simultaneously but in parallel, linked by non-disclosive summary statistics. Analysis is taken to the data – not the data to the analysis.
DataSHIELD is implemented entirely via free, open source software: at heart, a modified R statistical environment linked to an Opal database deployed behind the firewall at each data-holding organisation. Analysis is initiated in a standard R environment at the analysis machine, with communication between the analysis and data-holding machines controlled via secure web services. The same infrastructure and approach may also be used with just one data source – this is then referred to as “single site DataSHIELD” providing a freeware-based approach to creating a secure data enclave.