University of Oxford

OpenSAFELY is a new kind of data analysis platform for NHS records. It was created at pace during the pandemic, successfully granting researchers secure and efficient access to the whole population’s NHS GP records for the first time in history.

OpenSAFELY was used to drive globally impactful research on COVID, swiftly identifying which patients are most at risk and why, the causes and consequences of COVID infection and the different risks of different strains. The tools monitored the roll-out of vaccines and treatments, spotting which groups were being left behind, then monitored the changing effectiveness and safety of these interventions in near real-time.

NHS GP data is a globally impactful resource, but it has never previously been available at whole population scale, due to the inherent privacy risks. Medical records are a goldmine for research, but they also contain the most confidential secrets of every citizen. OpenSAFELY invented new methods to overcome these privacy challenges. Researchers get “dummy data” to develop their analysis, then submit their analyses for automated remote execution against real patient records, without ever needing to directly interact with that sensitive personal information. The data never leaves the GP practice’s control.

By innovating and implementing these ideas in a working service, the team were able to earn the full, formal trust of the British Medical Association, the Royal College of GPs, Citizens Juries and privacy campaigners.

OpenSAFELY is now used by analysts at dozens of organisations, on over 200 projects, and has been formally adopted by the NHS as the national GP data platform. The tools are now being installed in other countries and are also being used to help make pupil data safely accessible. By putting privacy first and innovating with an efficient working service, the team have earned the trust of researchers, clinicians, policymakers and, most importantly, patients.