There are reservoirs of electronic data with detailed information on just about everyone—spending and buying habits, health and medical histories, run-ins with the law, not to mention online behaviors.
These vast datasets are valuable, and many big businesses and individuals want access for a variety of purposes. Hospitals and health systems are increasingly sharing or selling patient data, often without the knowledge or consent of the individual. In some cases, information is down to the genome level.
But does sharing promote medical breakthroughs that benefit all, or could it lead to exploitation and bias? What are the rights of the individual in this electronic maelstrom?
A panel of five experts from the Yale School of Public Health and the Yale School of Medicine shared their insights Thursday (February 13) about balancing the need to access electronic health data to advance medicine and save lives with the right of individuals to privacy and protection.
Jennifer Miller, PhD, assistant professor at the school of medicine, told a large gathering at the Whitney Medical Library that as a starting point the Health Insurance Portability and Accountability Act (HIPAA), created to protect patient privacy, needs to be updated. HIPAA was enacted in 1996—before Google’s founding.
She said that patients should be able to opt in or out of having their data shared. A dynamic consent model should also be created, empowering individuals to give permission for every use of their data, and to revoke their consent if necessary. Miller also said that review boards could be created to oversee data sharing and grant permission if the need is warranted and to ensure safeguards are in place.
Gregg Gonsalves, PhD, assistant professor at the Yale School of Public Health, said that public health experts and other scientists need access to databases to improve health outcomes in a host of areas, including the ongoing opioid epidemic that has claimed thousands of lives in Connecticut alone.
“We have a health crisis on our hands. Data is very important,” Gonsalves said. “We need data to make better decisions.”
But in his experience, getting data to promote science is oftentimes very difficult. Some potential resources have not returned his calls; others flatly say “no.”
Gonsalves described how he and colleagues often chase after small datasets in an effort to answer looming public health questions. Law enforcement, he noted, does not face the same barriers.
We need data to make better decisions.
Gregg Gonsalves
Theodore Holford, PhD, the Susan Dwight Bliss Professor of Biostatistics at YSPH, detailed how he has used large datasets to study the health effects associated with changes in tobacco policy and regulation. What would happen, for instance, if the legal age to purchase tobacco was moved from 18 years old to 21? Holford has also examined the number of lives that have been saved since the U.S. surgeon general’s landmark warning on tobacco use was issued in 1964.
Holford emphasized that access to health data is essential for scientific research, but it needs to be done in tandem with protecting the privacy of participants.
Joshua Wallach, PhD, assistant professor at the Yale School of Public Health, outlined how various proposed rules have been introduced since 2014 which would require the U.S. Environmental Protection Agency (EPA) to make the raw data from research available to the public if it used to support federal regulations. Although these proposed rules appear to support research transparency and reproducibility efforts, many fear that the open-science movement is being co-opted, politicized and used as a justification for discrediting research findings related to air pollution and environmental exposures, he said.
What remains unclear is how these proposed rules would impact research involving humans, especially epidemiological studies conducted in the past, Wallach said. Additional concerns have been raised about patient privacy, the inability to reproduce studies conducted during natural disasters, and the “picking and choosing” of studies used to guide regulations by the EPA.
“I am an avid supporter of research transparency, and believe that all non-human subject research data should be made openly available,” Wallach said. “However, for future research containing personally identifiable information, only de-identified data should be made available to researchers who will utilize the information for scientific purposes.”
The panel’s final member, Donna Spiegelman, ScD, the Susan Dwight Bliss Professor of Biostatistics at YSPH, said progress has been made on a host of public health problems (HIV/AIDS, obesity, tuberculosis and depression, among many others) through data sharing. Future access to health data will allow these advances to continue.
She outlined some of the factors that are drawing increased attention to data sharing, including scores of media stories on the topic (one recent New York Times article had a headline that said people should be “freaking out” about personal privacy), HIPAA and an emphasis on protecting data through institutional review boards.
Spiegelman closed by offering her own proposals on how the future of sharing health data might be structured for everyone’s benefit. All studies with human subjects should be monitored by data and safety monitoring boards and any abuses of data privacy that harms study participants should be “severely sanctioned,” she said.