by Paul Govern
Any large-scale effort to use data from COVID-19 patients to serve biomedical or public health research must first concern itself with patient privacy issues.
With expedited funding from the National Science Foundation (NSF), Vanderbilt University Medical Center’s Bradley Malin, PhD, and colleagues will develop privacy-risk assessment techniques and software tools to assist data custodians, and particularly public health officials, to make informed data sharing decisions that appropriately balance public health goals with personal privacy in the context of a pandemic.
“As COVID-19 patients interact with the health care system, the data routinely generated through these interactions presents an invaluable opportunity to advance our understanding and grapple with the pandemic,” said Malin, professor of Biomedical Informatics, Biostatistics and Computer Science.
Current patient data sharing strategies tend to assume risks and benefits that don’t change with time or location. Malin and colleagues will develop what looks to be the first data sharing framework keyed to the dynamics of pandemics. Their decision-support tools will allow re-identification risk to vary based on considerations such as whether a pandemic is rampant or comparatively manageable at a given time and place, or whether release of more detailed location information could serve changing public health imperatives.
Joining Malin in leading the project will be Murat Kantarcioglu, PhD, of the University of Texas at Dallas. To help support the project, the two researchers have received $200,000 through the NSF’s expedited Rapid Response Research (RAPID) funding mechanism.
Malin has long studied data release strategies, and the risk that de-identified patient data will be re-identified. When the public’s interest, as represented by biomedical science, appears to conflict with the preservation of patient privacy, public policy typically comes down on the side of privacy.
“In a pandemic, where these policy decisions may be more pressurized, to allow optimum tradeoffs it will be helpful to have an agile framework for understanding the privacy implications of personal-level data sharing,” Malin said.
