Commentary: With great (statistical) power comes great responsibility: a comment on the ethics of using administrative data to investigate marginalised populations

As health data infrastructure improves, we have the opportunity to link increasing volumes of data in order to investigate important health problems. This is perhaps most pertinent when looking at the experiences and outcomes of our most disadvantaged groups, who are often invisible in data obtained through primary research. Whilst these data offer enormous opportunity, there are also ethical implications in their use, which are less frequently discussed than in relation to their qualitative counterparts. As a diverse group of clinicians
and academics working across public health, we share our experience and understanding of how we can improve our reflexivity in health data science and ensure that research in this area is ethically conducted in co-production with the people whose data we are using. We discuss the potential opportunities, challenges and impacts of using administrative data to investigate marginalised populations.