Brad Malin, PhD (@bradmalin) is Vice Chair for Research in the Dept. of Biomedical Informatics and Director of the Health Information Privacy Laboratory at Vanderbilt University. Dr. Malin describes his work as “bringing you better health through data, analytics, and policy.” He adds, “I am not, however, a conventional data scientist. I am driven by a concern that our society lacks the infrastructure to make the most of the data we generate. As such, I complemented my education with training in public policy and management to investigate how biology, computer science, and societal affairs can be blended to maximize the potential.”
Key details
- “Learning to Love De-identification in Biomedical Research”
- Brad Malin, PhD, Vanderbilt University (Guest Speaker)
- Monday, May 11, Noon – 1 pm (Add to Calendar)
- Palmer Commons, Floor 5, Forum Hall
- Lunch will be provided (near Floor 5 entrance to Forum Hall)
- RSVP requested.
The talk is sponsored by the Department of Learning Health Sciences, University of Michigan Medical School.
Abstract
Over the past several decades, numerous approaches have been developed to remove and obscure patient identifying information in the context of biomedical research. Generally, this approach to privacy protection, which is often called “de-identification” has been codified in regulations and laws, including the Common Rule and the Privacy Rule of the Health Insurance Portability and Accountability Act of 1996. There is a now a great opportunity to erect learning health systems on top of de-identified medical record systems; however, there is trepidation because the past decade has also witnessed a number of investigations into how to “re-identify” such information to the patients from whom it was derived. These demonstration attacks have called the strength of such privacy protections into question. The goal of this talk is to review why re-identification happens both from a computational and policy perspective, the extent to which such violations can be averted using risk analysis strategies, and how we can leverage de-identified patient data en masse to support large scale association studies. In this talk, Brad Malin, PhD, from Vanderbilt University will draw upon his experience in building one of the world’s largest de-identified electronic medical record systems and the experiences of the NIH-sponsored Electronic Medical Records and Genomics (eMERGE) Consortium.