Trusting big health data

It’s here – big data. Not only that, big health data. Most of us are aware (or should be made aware) that data about us are frequently collected, from websites perused to geolocation from the GPS on our smartphones or wearable devices. These data are occasionally shared with third parties, and sometimes the data are sold for profit. But when it comes to health-related data, people may be hesitant to trust the initiatives collecting and sharing their information.

There are several efforts focused on collecting and sharing health data from multiple sources with the intent of creating accessible data repositories to advance biomedical research and improving medical care. One such effort is the All of Us initiative.

A multi-sector effort led by the National Institutes of Health, All of Us aims to enroll one million individuals nationwide and collect diverse data types, including genetic, lifestyle and environmental factors, over time. Individuals interested in participating will opt-in, meaning that the initiative is tasked with finding individuals willing to agree to having their health data collected and shared.

All of Us and similar efforts face participant recruitment and retention challenges that are neither unique nor new. In the United States, historical incidents of research misconduct – notably the U.S. Public Health Service Syphilis Study at Tuskegee and the unauthorized use of samples of Havasupai tribe members – have fomented mistrust of researchers among some groups and led individuals to refuse to participate in biomedical research.

Potentially exacerbating this mistrust is a sense of loss of privacy and absence of control over information describing us and our habits. Given the extent of current “everyday” data collection and sharing for marketing and other purposes, this lack of trust is not unreasonable.

Health information sharing makes many people uneasy, particularly because of the potential harms such as insurance discrimination or stigmatization. Data breaches like the recent Equifax hack may add to these concerns and affect people’s willingness to share their health data.

But it is critical to encourage members of all groups to participate in big data initiatives focused on health in order for all to benefit from the resulting discoveries. My colleagues and I recently published an article detailing eight guiding principles for successful data sharing; building trust is one of them.

The attributes that make data sharing worthy of trust varies a bit for every person. For me, trustworthy data-sharing efforts produce knowledge and products, such as interventions or treatments, which are then equitably distributed.

For some, what makes data-sharing efforts trustworthy may be opportunities to form partnerships with researchers and participate in governance structures with voting power over who can access the data. Others may find data-sharing efforts trustworthy if they are offered monetary compensation in exchange for sharing their health data.

Developing policies that address sources of mistrust and are responsive to the range of values that underlie views about trustworthiness will be challenging, but this is a challenge that should not be ignored. People willing to share their data are needed to sustain data-sharing efforts and ensure that we all benefit from improvements in the treatment and prevention of diseases.

What makes a data-sharing effort trustworthy to you?  I look forward to reading your comments!

–By Angela Villanueva, M.P.H., research associate in the Center for Medical Ethics and Health Policy at Baylor College of Medicine