
BERD Linking public health datasets: Why data linking may do more harm than good, and how governance can help
April 28 @ 2:00 pm - 3:00 pm
BERD is offering a brand-new seminar in our data equity series: SPH faculty member Kathleen Conte will speak on the unintended potential harms of linking public health datasets and measures we can take to allay these. Dr Conte will illustrate with examples from homelessness research. This talk will be given in virtual format on Monday, April 28, 2–3pm (Pacific).
The existence and availability of many administrative datasets provides the opportunity to piece together information about individuals, which in turn opens the possibility to understand communities in a deeper way. This linking (or matching, in the language of policy) of records across administrative datasets is thus a powerful tool and has the potential to aid as well as harm people represented in the datasets. For example: in homelessness research, it may be of interest to identify those most vulnerable in a population, characterize disparities, or demonstrate cost savings associated with the provision of housing and supportive services to people experiencing homelessness. The linking of homeless service system data to various other administrative datasets is an increasingly common approach for these goals, yet there is a lack of discussion about possible repercussions, which may include perpetuating racial inequities and making vulnerable people targets of surveillance. In this talk we introduce data linkage from a public health standpoint, highlight the harms it can cause to individuals and communities it is intended to help, and discuss how governance can be used to allay these unintended harms.