Many patients prefer to hide sensitive clinical info from care teams: study

When given control of their own health records in electronic format, patients are ready and willing to share their info with their own doctor or nurse. But they stubbornly resist letting other, equally critical members of the care team in on the conversation.

This came to light when, participating in the first real-world trial of patient-provider interaction via EHR, just under half the patients withheld some info. Much of the “guarded” data involved sexually transmitted diseases, substance abuse or mental health.

The results were published in a supplement to the January edition of the Journal of General Internal Medicine.

In the six-month study, 105 primary-care patients were asked to state their preferences on which clinicians could access sensitive information in their EMR, along with which clinicians could see which particular bits of info. In other words, they were given the chance to hide some or all of their data from some or all of their providers—and 49 percent just said no to having every clinician see every line.

The study was designed and conducted by a partnership of the Indianapolis-based Regenstrief Institute, the Indiana University School of Medicine and Eskenazi Health (formerly Wishard Health Services, also in Indy). 

Some providers participating in the study were fine with patients exerting firm control, as long as patients knew their care might be compromised by the secrecy. Other providers strongly objected to being cut off from full access and, thus, risking being held responsible for bad or unsafe care, according to a release issued by Indiana University.   

IU explained that providers were able to override patients’ preferences, and view any hidden data if they felt the patient’s health care required it, by hitting an electronic “break the glass” button. When providers hit this button, the program recorded the time, the patient whose electronic chart was being viewed and the data displayed.

The journal supplement contains five papers, including a point-counterpoint discussion.

The need for collaboration between physicians and patients is critical, Kelly Caine, PhD, a professor of human-centered computing at Clemson University, writes in the discussion section. “Whether this means respecting patients’ wishes to withhold some aspects of their records using granular control mechanisms or talking with patients about the benefits of sharing even sensitive, embarrassing information, providers and patients need to work together from a base of trust to achieve good health care.”

“[A]s both a practicing general internist and health services researcher, I cannot support patients having the ability to hide information in their EHRs from their health care providers, especially their primary care physicians,” counters William M. Tierney, MD, associate dean for clinical research at Indiana University. “I understand that my attitude could have adverse effects on communication with my patients and their willingness to provide sensitive information to me and other health care providers. Moreover, I admit that there could be an adverse impact on the doctor-patient relationship and communications if I told my patients, especially those who have a strong desire to control access to their EHRs, that I intended to ‘break the glass’ and ignore their preferences for each and every visit. Nevertheless, I have an overriding duty to provide the best and safest patient care possible in the information-intensive business that is health care. Hamstringing my ability to provide such care is unacceptable to me.”

Given a chance to rebut, Caine concluded: “Research is needed to determine what the ‘right clinical information’ to display to providers should be. This is especially important in a future where health information is ubiquitous, extra-clinical, and potentially overwhelming. Understanding patients’ preferences for what constitutes the ‘right clinical information’ should be part of this process.”

HHS’s Office of the National Coordinator for Health Information Technology (ONC) supported the research with a $1.6 million grant.