Patient-Generated Health Data Under the Microscope
By Lisa A. Eramo, MA
For The Record
Vol. 32 No. 4 P. 22
The pandemic has heightened awareness of the value of gathering accurate information.
It’s a question on every physician’s mind: What actually happens outside of the exam room? Do patients follow treatment recommendations? Are they getting enough exercise? What are their daily glucose levels? What about their daily blood pressure readings?
All of this (and more) is patient-generated health data (PGHD), information that could potentially improve patient outcomes while also helping providers take advantage of new reimbursement models focused on episodic care.
PGHD (eg, daily pulse oximetry readings that patients send to their providers for remote monitoring) are proving to be critical during the COVID-19 pandemic, which has forced people to follow strict stay-at-home orders in many states. When clinicians monitor patients remotely—and intervene only when necessary—everyone can practice social distancing more easily.
“Patient-generated data is being seen as increasingly valuable in helping health care organizations make good use of limited resources, such as only testing individuals who report specific sets of symptoms or recommending self-monitoring of various health indicators,” says Lesley Kadlec, MA, RHIA, CHDA, director of policy and state advocacy engagement at AHIMA.
The bad news? Many providers still don’t trust the data. Therefore, they may not be willing to integrate them into the EHR and lean on them for clinical decision making.
“Patients are more engaged and empowered than ever before, but at the same time, the industry hasn’t yet figured out a way that the data can be formally integrated into clinical systems of care,” says George Demiris, PhD, FACMI, a professor in both the Penn School of Nursing department of biobehavioral health sciences and the Perelman School of Medicine department of biostatistics, epidemiology, and informatics.
Demiris, whose research has focused on the role of wearable devices and digitally augmented residential settings in facilitating passive monitoring of community-dwelling older adults to help them remain independent, says PGHD have immense potential but only if the health care industry can govern those data appropriately.
Part of the problem is that the FDA doesn’t regulate many of the consumer-facing devices and apps that capture PGHD, Demiris says. “Anyone can create an app and make it available,” he says. “To this day, lack of regulations and uncertainty about the reliability of the devices are still barriers.”
For example, Demiris says patients and providers can’t necessarily draw accurate inferences about sleep quality from an app that requires users to place their cell phone under their pillow so it can track movement. Ideally, if these apps were classified as medical or diagnostic apps vs lifestyle apps, the FDA could require evidence-based validity to help providers identify specific consumer-facing apps and devices they can use for clinical decision making, he says.
Federal Support for PGHD
At the federal level, there’s growing support for the use of PGHD. The Office of the National Coordinator for Health Information Technology (ONC), which is currently developing a policy framework for incorporating PGHD into EHRs, supports the use of PGHD to advance patient engagement, care delivery, and research. The goal? To seamlessly and securely transmit PGHD from patients to clinicians for routine care by 2024.
To make that happen, ONC must adopt interoperability standards that support the capture, use, and sharing of PGHD. Clinicians must also be able to store, retain, and analyze large volumes of PGHD with minimal concern for liability.
The 21st Century Cures Act includes a new standard for interoperability—United States Core Data for Interoperability (USCDI)—that could spur greater clinician trust in and reliance on PGHD, says Nick Hatt, a senior developer at Redox, an integration platform that securely and efficiently exchanges health care data.
USCDI includes data provenance (including author time stamp and author organization), which would allow organizations to track data origination over time. How could this eventually tie in with PGHD? Hatt provides this explanation: If a clinician reviews medical device data from five years ago, data provenance ideally could enable them to identify the specific version of the device—an important detail when determining whether there have been any recalls.
“Provenance adds a trust component to patient-generated health data in a standardized way,” Hatt says. “Using a standard as a benchmark like the USCDI to record the who, when, and where of a data point indirectly raises its value.”
It all goes back to information governance, Kadlec says, adding that incorporating PGHD into an organization’s enterprisewide information governance program can help in terms of building provider trust in the data. AHIMA defines information governance as “An organizationwide framework for managing information throughout its lifecycle and supporting the organization’s strategy, operations, regulatory, legal, risk, and environmental requirements.”
“If clinical decision making is to be made on the basis of data that are supplied by patients and subsequently documented in the EHR, clinicians must also be certain that the data has integrity—essentially meaning that it is reliable, accurate, and trustworthy,” Kadlec says. “Having the assurance that they have complete, current, and accurate information through development of an information governance program is essential for any health care organization to achieve its care delivery and operational goals.”
Information Governance and PGHD
When it comes to widespread use of PGHD for clinical decision making, experts agree that information governance is critical. To help sort out matters, consider the following questions:
How does the organization define PGHD? For example, will it include only FDA-approved devices (eg, glucometers, digital blood pressure devices, heart rate monitors)? Or will it also include general wellness devices or apps (activity and sleep trackers as well as wellness and health apps)? What about patient-reported outcomes using standardized assessments? (See sidebar.)
How will PGHD integrate into clinician workflows? The more seamless, the better, experts say. “Well-designed and properly executed health IT systems that integrate PGHD and then present the data to clinicians in a usable manner have the potential to improve decision making and enable patients to become active participants in their health care,” Kadlec says. “Health care organizations should look for ways to integrate PGHD through a combination of health IT adoption and engagement with clinicians.”
How will the organization protect PGHD? “Any information that is filed, maintained, or incorporated into the patient’s health record, including patient-generated health data, becomes a part of the patient’s health record,” Kadlec says. “These records are subject to all applicable state and federal regulations concerning privacy, security, use, maintenance, and disclosure.”
In some cases, data from medical organizations will be linked with data stored by a private entity, Demiris says. For example, some apps store clinical and nonclinical data (eg, GPS data), potentially enabling vendors to correlate physical movement with health outcomes. “There’s huge risk if there was a data breach,” Demiris says.
Negotiations between health care systems and vendors must address privacy and security. “These considerations should go beyond HIPAA compliance and data encryption and address audits and ongoing risk assessments,” Demiris says.
What is the vendor’s role? Vendors that create consumer-facing devices and apps play a critical role in information governance of PGHD because that’s where the raw data originates, Demiris says. Organizations often rely on whatever extraction methods the vendor uses and then store a summary or trends—not the raw data themselves—as notes or appendices attached to the electronic record, he adds.
Therefore, vendors must be transparent about which data are being measured and how, including algorithms used to extract insights and summaries. In addition, Demiris says every time vendors update or improve their algorithms, providers should be informed so they can better understand potential data fluctuations.
Hatt says vendors also need to be flexible and provide whatever type of data integration—such as automatic data flow from the patient to the provider vs patient-directed data flow—will best meet the organization’s goal.
“Vendors need to realize that health care organizations have huge existing integration backlogs and patient-generated data will fall to the wayside,” Hatt says. “Vendors face a delicate balancing act between reducing friction for patients and providers and reducing the burden on the IT team that needs to set up integrations. Being flexible and willing to implement in stages can be critical to success.”
How will the organization educate clinicians and consumers? Organizations need to educate patients about PGHD, including how to collect them. “People have different levels of health literacy and data literacy that organizations must address,” Demiris says, adding that patients also need to understand why PGHD are important and how their providers will use them.
Clinician education is equally as important, Demiris notes. “Clinicians are not trained on how to successfully navigate patient-generated data,” he says. “A lot of clinicians have been trained that the only valid data are the data captured during the clinical consultation. Training clinicians for this new reality should be led by the organization.”
Demiris provides the example of direct-to-consumer (DTC) test kits for food sensitivities. “To me, this is patient-generated data because an individual consumer decides to purchase a test and obtains data that may have some clinical relevance,” he says.
However, what if lab results don’t correlate with results from the DTC kit? For example, what if the DTC kit indicates a gluten sensitivity but physician-ordered labs do not? Physicians need to know about health-related decisions patients make based on these data, such as eliminating gluten from their diets. DTC genetic testing is another example. Patients may undergo surgical procedures or take certain medications because of the results they receive, Demiris says.
Because PGHD are going to continue to proliferate, Hatt says organizations must find ways to capitalize. “The patient becomes more engaged, and the provider has more context. Ultimately, both of these things lead to better care and better outcomes,” he says.
— Lisa A. Eramo, MA, is a freelance writer and editor in Cranston, Rhode Island, who specializes in HIM, medical coding, and health care regulatory topics.
PGHD: LEVERAGING PATIENT-REPORTED OUTCOMES IN CLINICAL CARE SETTINGS
When many of us think about patient-generated health data (PGHD), our thoughts tend to turn to Fitbit, Apple Watches, and a variety of other wearables. However, it’s important not to ignore a vast array of other data—including those from FDA-approved devices and biometric sensors as well as electronic patient-reported outcomes (ePROs).
ePROs are derived from standardized surveys implemented across a health care system to track patient outcomes over time. Some ePROs can apply to any patient population, such as assessing quality of life or physical function. Condition-specific ePROs apply to a specific patient population—for example, one with a defined clinical condition such as assessing symptoms for multiple sclerosis or low back pain.
“The value of ePRO data can be huge,” says Cynthia LeRouge, PhD, an associate professor in the department of information systems and business analytics at Florida International University. “A trial may last a year, but what we don’t know is what happens after that year from a more long-term perspective.”
LeRouge served as a coprincipal investigator on a multidisciplinary research team charged with developing national guidelines for governing, implementing, and reporting ePROs in clinical practice. The team’s work culminated in the publication of an ePRO toolkit that helps organizations integrate and report ePRO data for clinical care delivery.
LeRouge says ePROs provide the following benefits:
• better standardized assessments of patient symptom experiences;
• remote monitoring and tracking of patient health status, potentially minimizing the burden for in-person clinical evaluation; and
• patient-reported health status in the EHR.
Still, as with any type of PGHD, there must be an information governance structure in place to successfully deploy ePRO data for clinical decision making, LeRouge says. For example, providers must agree on what specific validated and standardized assessments they’ll use and in what circumstances. For example, for providers who want to assess depression, it means deciding when and how to use the PHQ-2 vs PHQ-4 vs PHQ-9.
Codesign with end users is also important, LeRouge says. “To see the value of patient-generated health data, including ePRO data, we cannot follow the ‘build it and they will come’ view, but rather work to understand pain points people face and create solutions to address them,” she explains.
Other questions to consider: How will the organization present missing data in ePRO reports? How will privacy and security be maintained when collecting ePROs? How should providers handle data outliers? For example, will the patient be required to validate the data or fill out the assessment a second time? Might certain patients be overlooked due to language, access, and cultural barriers? If so, how can this be addressed? Do the ePRO data actually advance patient care and outcomes?
“In my opinion, systems will need to create a process for evaluating whether and how the data improve the health of patients and adapt accordingly,” LeRouge says.
While ePRO use has historically been sought and implemented at an individual practice or clinic level, more health systems are implementing the system throughout their health enterprise. “This is a sign of increasing recognition of the value of incorporating the patient’s own voice about their experiences in a systematic way into the patient’s official health record,” LeRouge says.