Patient-Generated Health Data

Gateway to Consumer-Driven Healthcare, or Just Another Distraction for Clinicians and Caregivers? 

I was first exposed to patient-generated health data (PGHD) quite early in my clinical practice. One patient was recording her self-administered blood pressure readings several times a day and bringing dozens of readings to her appointments. I had little guidance as to how to turn the data into something useful. Average the readings? Incorporate measures of variation? Correlate highs and lows with some estimate of medication adherence? I didn’t really have the time or tools to do any of this. I’m sure many of you can relate. We all have the experience of ingesting large numbers of blood glucose readings at various times of day from people with diabetes. Or daily weight readings in heart failure patients. I recall being fatigued even then by what could quickly amount to data overload.

But then I noticed a trend. The patients most engaged in generating and sharing their own data were the most… well, engaged in their health. Although my use of the data was imperfect and may not have always been grounded in science, there was magic in the process.

So where does that leave today’s physicians?

In the U.S., physicians are more likely than ever to have patients who are managing their health using wearable devices and mobile apps. In fact, according to a survey by Accenture, that number has doubled in the past two years, increasing from 16% in 2014 to 33% today. What’s more, 40% of these consumers have discussed or shared mobile app data with their doctor in the past year. This evolution is driving physicians to question how—or even if—they should be incorporating PGHD into their practice. It’s a good question, with multiple implications.

PGHD refers to health data that is created, captured, and shared by patients. There are many sources of this data, including health risk assessments and patient-monitoring devices as well as wearables and mobile apps. PGHD can augment clinical data that is available in electronic health records and claims data, so can potentially provide physicians with a more comprehensive picture of an individual’s health status. And as the U.S. care delivery model transitions to proactive population health management, PGHD can support early intervention, shared decision-making, and ongoing health maintenance. In fact, in the Accenture survey, 77% of consumers and 85% of doctors said that using wearables helps a patient engage in their health.

But whether patients should be encouraged to enter this data in an electronic health record (EHR) is less straightforward and more nuanced.

Electronic health records and their accompanying patient portals have become nearly ubiquitous as a result of HITECH, including its Meaningful Use provisions and associated financial incentives. But patient portal use remains limited, typically to administrative tasks such as appointment scheduling and bill paying, which raises the question of whether patients will consistently adhere to any manual data-entry process for health information. The question is further complicated by skepticism relating to patients’ ability to accurately recall and record data as well as the integrity of the underlying data sources themselves. All that means that PGHD will be more readily accepted and adopted in clinical practice when it is properly sourced, accurately parsed and structured, and automatically transmitted from interoperable technology.

Of course, there are advantages and disadvantages to the use of this information. PGHD, by definition and practice, means that patients are engaged in their health and healthcare, which is clearly positive. And PGHD complements clinical and claims data to provide an additional window into patient health. PGHD can also be gathered and transmitted actively or passively—in real-time or near real-time—which can increase engagement and help facilitate prospective interventions such as alerts and reminders.

On the other hand, PGHD brings with it some clear disadvantages. One challenge is normalizing the data that comes from multiple formats and that is often derived from a wide range of algorithms—and which can complicate standards-based practice. And the potential volume of information risks distracting, misleading, or overwhelming providers. If physicians don’t know what to do with the data, or don’t have an appropriate filter to make sure they’re getting the right data or method to interpret it, it can become more noise in an already noisy environment[CC1]. This is a concern that experts are already addressing with respect to EHR alert fatigue.

Clearly PGHD is raising new opportunities, but as it exists today, this data is often not actionable for healthcare providers. Even though some EHRs are beginning to bring in PGHD and display it as part of a patient’s record, that data typically isn’t linked to a discrete diagnosis or procedure code. That means it can’t be easily related to a care gap that is addressable by the application of clinical guidelines.

Ultimately, PGHD is a vanguard of potential new data sources that will help feed models for identifying and prioritizing risk. As these data sources become more sophisticated, they will begin to add valuable new vectors for population health management tools.

J. Siemienczuk, MD