Precision Health member Kayte Spector-Bagdady, JD, MBioethics, is principal investigator on an R01 grant to improve data-sharing between hospitals and commercial entities by incorporating whether and to what extent patients find these data-sharing practices acceptable. Thorough and accurate research requires datasets that are diverse and representative of the population, yet Black and Hispanic patients are more likely to be uncomfortable with their data being shared. This project (“Hospitals Sharing Patient Data and Biospecimens with Commercial Entities: Evidence-Based Translation to Improved Practice”) aims to find areas of compromise that simultaneously respect patient autonomy and increase the demographic diversity of research databases.

Spector-Bagdady is associate director at the Center for Bioethics and Social Sciences in Medicine and an assistant professor of obstetrics and gynecology. She is also chair of the Research Ethics Committee, a member of the Michigan Medicine Human Data and Biospecimen Release Committee, and an Associate Editor of the American Journal of Bioethics. Co-investigators on the project are:

• Chad Brenner, PhD, MSE
• Luyun Chen, PhD
• Melissa Creary, PhD, MPH
• Reshma Jagsi, MD, DPhil
• Sachin Kheterpal, MD, MBA
• Jodyn Platt, PhD, MPH
• Brian Zikmund-Fisher, PhD

How did this project develop? Out of what prior research did it grow?
Kayte Spector-Bagdady: Widely sharing patient data and biospecimens can enable life-saving advances in translational science. Hospitals and commercial entities focusing on precision medicine and other genetic and artificial intelligence technologies have increasingly partnered to share these data resources. For example, Dr. Kheterpal has applied modeling techniques to the integration of Electronic Health Records (EHR), fitness wearable technologies (e.g., Fitbit), and genomic data into innovative polygenic risk scores. Collaborations between hospitals and industry can benefit these advances by funding and accelerating discovery.

To ensure equitable access to scientific advances, datasets must include patients reflecting the demographic distribution of disease. Dr. Brenner has found that while genetic factors may play an important role in the variability of head and neck squamous cell carcinomas, the lack of data derived from non-Caucasian cohorts impedes the development of personal medicine. But how do we best diversify research data while balancing the autonomy interests of individual participants against benefit across communities?

Previous research by Dr. Platt and me demonstrates that many patients are uncomfortable with hospitals sharing their data with industry. In addition, we know from Dr. Jagsi’s research that individuals who identify as Black or Hispanic are consistently more likely to report discomfort. Dr. Creary has argued that when policies aimed at achieving justice for marginalized populations fail to recognize how those populations have already “embodied the cumulative effects of marginalization,” additional distribution of resources will never achieve “justice” insofar as the concept is already “bounded by greater socio-historical constraints.”

We therefore identified the need to measure patient response to actual hospital data-sharing practices with the statistical power to differentiate preferences across race, ethnicity, and other relevant scales in order to identify promising areas for data commercialization policy compromise with an equitable impact.

What health research issues will this project address?
KS-B: This project intersects with a number of key issues related to precision health and genetic research. At its core, it evaluates the relationship between patient autonomy concerns and data-sharing practices that have emerged between hospitals and their industry partners. Our analysis also addresses issues related to justice by explicitly focusing on health disparities and the unanticipated consequences of commercialized data.

This project encompasses many areas—gathering, sharing, and stewarding data; racial bias; underrepresented populations; transparency; and ethics. How did you establish research parameters for it?
KS-B: While there are many entities that sell or share health or health proxy data, many kinds of people who generate them, and many different regulations governing such interactions, this project is going to focus specifically on (1) hospitals which share (2) patient data and specimens to (3) private industry.

What makes U-M strategically positioned to take on such a project?
KS-B: As one of the top publicly funded research institutions in the county, U-M is well-positioned to serve as a test site for new and improved approaches to data sharing with industry. It has a devoted translational science institute with a demonstrated interest in facilitating this kind of work. Precision Health at U-M is an excellent, innovative initiative for which Dr. Zikmund-Fisher is a faculty advisor. Members of our research team serve on U-M’s recently established multidisciplinary Human Data and Biospecimen Release Committee, which reviews industry and research requests to commercialize health data and biospecimens collected at Michigan Medicine. U-M is the perfect laboratory to study current practice, set better policies, and field test them all in the same institution.

“Compromise” is a key objective of the project—how will you go about discovering areas for compromise?
KS-B: Our approach is to use an exploratory framework that gathers information from multiple levels of stakeholders. Using a combination of interviews and surveys, we will explore current data-sharing practices, and the relationships hospital stakeholders have with patients and industry partners such as biotechnology companies. We will acquire multiple perspectives on the likely types and acceptability of patient impacts that might affect data-sharing practices, as well as explicit information regarding factors that would affect stakeholders’ willingness to compromise. Importantly, our mixed methods ELSI (ethical, legal, and social implications) approach will also utilize sampling strategies that are designed to ensure that we include patient populations that have been traditionally underrepresented in research of this kind.

What is the desired outcome of this project? Where will it lead?
KS-B: In addition to characterizing both hospital and patient perspectives on data sharing, the ultimate goal of the project is to facilitate the development and implementation of improved data-sharing practices. We believe that identifying areas for compromise between patients and hospitals to improve data-sharing practices with industry in ways that are respectful of individual patient autonomy and equitable in impact across diverse communities will have a major impact on the accessibility and generalizability of translational science.