Ask Carla Heggie (BA’77) about balancing privacy with the need to access personal information for public health purposes, and you’ll quickly learn why that’s the wrong question. “It’s not a balance,” she said in a recent interview. Privacy and access, she explained, are not a zero-sum game. “You can give total access to information, in keeping with the legislation, while upholding the privacy rights of individuals. So it’s a win-win situation.”
A privacy expert who has worked in academia, government and the private sector, Heggie teaches in Dalhousie’s School of Information Management and Faculty of Computer Science. She is one of three panelists at the next Open Dialogue Live event centred around “Data and its impact on health.” The open-to-all online panel takes place April 12 at 6:30 p.m.
Big data, big responsibilities
The COVID-19 pandemic has demonstrated the importance of access to health and genomic data, in order to understand the evolution of SARS-CoV-2 (the virus that causes COVID-19) and its effects on different populations. But big data brings with it big responsibilities.
The key issue with accessing or releasing health information is whether it can identify individuals, says Heggie (shown right). “We have to look at harm. Administering privacy is just a risk-mitigation strategy, and it’s an ethical field. So you’ve got to look at what the harm is in releasing, and what the harm is in not releasing,” she says. “If we’re talking about a small community ... an individual can probably be identified with a few pieces of information.”
Samuel Stewart (PhD’14) teaches in Dalhousie’s Department of Community Health and Epidemiology, and is the director of Health Data Nova Scotia — the provincial organization that holds de-identified health information that can be used for research purposes. The group has a committee that meets once a month to assess data access requests.
With machine learning, information in large datasets can be quickly processed, potentially drawing out helpful correlations. But Dr. Stewart (shown left) says privacy legislation in Nova Scotia is “very outcomes-focused,” making it hard to do that kind of research. “You have to know
what you're trying to find with the data. You can have very specific elements, and then you use those elements and do the research to get the results,” he says. “But we don't always know where we're going with machine learning. We have a novel idea and we've got a method for knowledge discovery.”
Taking that approach and using analytics to generate hypotheses is “really difficult to do” with health data, Dr. Stewart says, “because you need to justify your use of the data in order to get access to it. And this is not a problem that's unique to Nova Scotia.”
Challenges linking genome data to useful epidemiological metadata
Finlay Maguire, who has posts in both the Faculties of Computer Science and Medicine, says his work during the pandemic has largely focused on genomic data. Sequencing of the SARS-CoV-2 virus and its many mutations — along with making that data readily available — has allowed researchers to quickly identify variants of concern and prepare for them. But Dr. Maguire (shown right) says that during the pandemic, that open approach “has collided somewhat with classical public health epidemiology kind of approach. There have definitely been some challenges in linking genome data to useful epidemiological metadata.”
As the COVID-19 pandemic continues, and as we prepare for the inevitability of future public health crises, the intersection of public health and data privacy will remain critically important. Learning the lessons this pandemic provided will protect us in the future.
Register:
The Open Dialogue Live panel will be moderated by Tara Sampalli, senior scientific director with the Nova Scotia Health Authority. This discussion is part of a public series connected to Here We Code — a campaign to position Nova Scotia as one of Canada’s top digital ecosystems, celebrating success stories and showcasing how digital is impacting the world around us.
The event is free, but registration is required. You can register here.
The panelists:
· Samuel Stewart (PhD’14) is an Assistant Professor in the Department of Community Health & Epidemiology at Dalhousie University, and the Director of HDNS, the provincial repository for de-identified health administrative data for the purposes of research. His research areas include secondary data analysis, natural language processing, machine learning, decision support systems, and meta-analysis.
· Finlay Maguire is an Assistant Professor in the Faculty of Computer Science and Department of Community Health & Epidemiology. His research involves using genomic data to better understand infectious diseases and the collaborative application of data science methods to social and health-related problems. Dr Maguire has co-developed widely used SARS-CoV-2 analysis, metadata, and quality control specifications and tooling.
· Carla Heggie (BA’77) is a certified Information Access & Privacy Professional who has been involved with the development, writing, delivery, and teaching of information management legislation and policy for over 34 years. An adjunct professor/instructor in the faculties of Computer Science and Management, she teaches five different courses in privacy. Her company, Teal Creative Thinking, offers consultation in the areas of information access, privacy, policy, process, and patient advocacy. As someone with a chronic illness, she has more than enough experience with most levels of health administration and the use of personal health information.