This article was originally published in Radiology Today Vol. 20 No. 4 P. 10
A computer doesn’t look at you the same way your physician does. It sees data—lab data, radiology images and reports, office visit notes, and more. It can collect your entire health history and help researchers and clinicians use those data in many ways. Many patients don’t realize, however, that they have signed off on allowing their personal health information to be shared for research purposes, indefinitely.
But the sharing of patient data is not a black-and-white issue. Those same patients could benefit from data that have been pooled together to determine the best treatment plan for their specific type of cancer—even narrowed down to factors such as their gender, age, or genetic makeup. How can health care facilities and new technology come together to ethically use patient data for the greater good?
Today, HIPAA protects patient data from inappropriate use and maintains that patients legally own their own identifiable data. Once data are anonymized for research purposes, however, hospitals and research facilities are free to use and, sometimes, sell the data. As Adam Tanner, a fellow at Harvard’s Institute for Quantitative Social Science, shares in his book, Our Bodies, Our Data: How Companies Make Billions Selling Our Medical Records, many patients don’t realize that their data are being sold.
Even if a patient’s health care data are deidentified, they can still expose identity in other ways. Data from health applications and social media postings regarding health are often freely available and unregulated. DNA testing kits such as 23andme come with their own unique risks, including differing state regulations concerning the release of information to life insurance companies.
A famous example of unregulated sharing is the case of Henrietta Lacks, a patient whose cells were taken without her knowledge in 1951. Those cells would prove to be vital for developing gene mapping, the polio vaccine, and many other medical advances. Her cells were sold by the billions, yet her family today can’t afford health insurance, and many of us only recently heard her name thanks to the movie starring Oprah Winfrey. Consider this: Lacks was a patient in the 1950s; today, cloud technology, AI, and machine learning make the capabilities of analyzing data even greater.
So, what happens when the compendium of data available are analyzed? One of the goals of the 21st Century Cures Act is to develop targeted drugs and clinical trials for hard-to-treat cancers and rare diseases. In order for this to be possible, a data network must be created—one that isn’t locked behind the siloed walls of many EHR systems. A new rule issued by the Office of the National Coordinator (ONC) for Health Information Technology states that patients should be the actors and not “acted upon” when it comes to their medical records. According to the director of the office of policy for the ONC, “If a patient requests their record, and it’s not given to them electronically and for free, that’s information blocking.” The ONC is not accepting any excuses, as health information exchanges could be liable for penalties of up to $1 million for lack of interoperability.
Interoperability is critical to research projects that could have positive impacts for years to come. Project Baseline, a collaboration between Verily, Stanford Medicine, Duke University, and others, is collecting data from more than 10,000 informed participants over four years. The data will be used to develop a baseline of normal health information and assist clinicians in disease prevention, risk factor identification, and treatment.
We all benefit from the findings of such comprehensive research studies. AI and machine learning will soon aid these efforts, allowing findings to be gained at a previously unfathomable pace. But how can patients, who are increasingly acting as health care consumers, feel informed that their data are being used appropriately?
First and foremost, facilities should consider reducing jargon. Recently, in the world of radiology, there has been an increased push to reduce the level of jargon in radiology reporting. For example, instead of saying “edema,” a patient-friendly version would simply say “swelling.” Consent forms offered to patients at a facility should adopt a similar approach of stating matters in plain language. Additionally, if a facility or practice merges with a larger institution, best practices include informing patients through e-mail or a mailed letter that their data are now shared with a new organization that has its own unique policies regarding consent.
Providers also have a golden opportunity to offer improved access and control to patients for their own data. Let’s say a patient wants to receive a copy of an imaging exam they had a few years ago. Today, this often involves a medical records release form and a CD that is either mailed or picked up on site. The process can take well over a week. Patients should be offered access to a patient portal where they can act as the owners of their medical data and images and use them as they please, including seeking second opinions.
Additionally, digital tools can make a big difference in how data are used for research and insight. A recent trend we have seen is facilities developing their own internal research clouds, rather than selling or using data elsewhere. Many large medical facilities have great amounts of data but lack a unified and organized method of storing and analyzing them. Today, the cloud offers the opportunity to form academic research PACS where data in consideration for a study can be stored, viewed, and appropriately anonymized. Researchers must first determine whether their facility allows mining of medical record data and then review projects with an institutional review board. Once the review and approval process has been completed, the cloud allows the deployment of an integrated application fabric that elevates health care efficiency and quality.
The opportunities for positive outcomes are endless, but facilities must make transparency and patient control core institutional values as they move forward.