Guest Article: Is Patient Generated Health Data (PGHD) trustworthy enough to use in health record banks?

May 18, 2014

The push towards shifting the patient’s role from a passive recipient of care to an active member of the care-team looks set to gain further legislative backing. Earlier this year, the Health IT Standards Committee, along with The Joint Commission and ONC, laid out recommendations for integrating patient generated health data (PGHD) into Stage 3 Meaningful Use requirements. To see what this might mean to health IT and med tech vendors, I reached out to Zach Watson of TechnologyAdvice, who covers EHR related news, along with business intelligence, and other topics. Zach mentioned that health records banks might be an interesting future direction so I asked him to share his thoughts on PGHD, how trustworthy it might be, and what the future for PGHD and health banks might be. Here’s what he said:

Greater integration of PGHD into clinical practice and research opens the door to innovative treatment standards, specifically for population health management and risk-bearing delivery models, such as accountable care organizations. Broadly, PGHD helps reduce costs by allowing providers to view a patient’s previous procedures and tests, and avoid adverse reactions through access to up to date medication information.

On a more granular level, PGHD could give providers, researchers, and device manufacturers access to real-time biometric data that infers day-to-day changes in patient health and activity. This type of data could be essential in developing care plans and technologies for patients with chronic conditions who need daily treatment to address their symptoms.

PGHD already exists in clinical practice today in the form of patient reported outcomes, but Stage 3 requirements envision a greater expansion and integration of these data, with the patient as the main comptroller.

However, questions remain around data reliability, data standards, exchange architecture, and of course privacy.

When searching for a solution to these issues, providers, legislators, patients and other stakeholders must examine the architecture of existing EHR and health information exchanges (HIE). The most common iteration of an HIE, also referred to as a decentralized or federated model, requires that a physician’s EHR showcase interoperability with every other EHR from every other provider from every previous point of care for each individual patient. The physician’s EHR must query all these previous EHRs to receive the required health information, while verifying patient consent and correcting inaccuracies in real-time.

Such decentralized architecture has led to constant citation of low interoperability rates, and scalability issues in the context of PGHD. Patients now have to contribute data to numerous EHRs, with each possibly using a different portal for access. In this federated system, EHRs must not only perform complex functions required by their end-users, but also interface with disparate systems to form an interconnected information transfer network.

A separate framework for interoperability and patient participation in HIEs was proposed back in 2007. Dubbed “The Independent Health Record Trust Act,” the legislation developed an infrastructure for health record banks (HRBs). Essentially, these are untethered portals that act as repositories for patient information, designated by region. When physicians need a complete view of a patient’s record, their EHR queries the HRB for all the patients records and returns updated records to the HRB after the episode of care is completed.

In the HRB model, patients, and providers for that matter, don’t need to worry about portal redundancy, or about receiving inaccurate or incompatible records. One of the worries about utilizing patient generated health information lies in the human error associated with manually entered data. However, positioning the patient as the final quality assurance mechanism should ensure greater accuracy as the patient ultimately risks the greatest harm from health information mistakes.

Further, if physicians notice inaccuracies in the information, they can immediately make the changes if given permission by the patient, or send the patient a note about the inaccuracies. Patients can then deal with information within the HRB platform, instead of the incorrect information being transmitted back to multiple EHRs.

As an example, Geisinger piloted an ONC-funded project where patients reviewed their medication lists through patient portals and provided feedback on current and expired medications prior to an upcoming office visit. This data was reviewed by Geisinger pharmacists who followed up with the patients. Pharmacists changed the medication info as needed, notified the patient’s provider, and marked the source of the change in the EHR system.

How accurate were the patient’s changes? Pharmacists made the patient recommended changes 80 percent of the time. The providers who participated also reported that the program saved them time.

In terms of compatibility, developing one standard for HRBs will make interoperability cheaper and quicker to realize. Since the HRB would be under direct and complete control of the patient, providers could use the direct protocol to exchange information between platforms. The HRB model also offers system architects the opportunity to settle on a smaller range of standardized forms, such as CCDA.

In relation to privacy, patients can determine the level of data access they want providers to have. While this is far from a perfect solution – it’s feasible that patients may restrict information they find unflattering – it does offer a compromise for including patients, and the data they create, in health information exchanges.

Unfortunately, the aforementioned legislations for HRBs (HR 2991) died on the vine after making its way to the House Committee on Ways and Means in 2007. However, the bill’s failure didn’t signal the death of health record banks – eHealthTrust opened a health record bank in 2010 designed to scale with Arizona’s nascent HIE. Additionally, the Health Record Bank Alliance has outlined many of the ideas posited here in greater detail, along with compelling evidence of financial incentive for the HRB model.

Influential stakeholders need to incorporate PGHD to make their population health management attempts yield positive gains, both financially and clinically. The patient-facing infrastructure already largely exists, with an increasing amount of patients generating relevant data through wearable medical devices.

In lieu of a regional HRB, providers can still leverage PGHD using the following practices:

  • Begin by identifying existing information gaps in your EHR system. What data could you and your patients benefit from integrating into the decision making process?
  • Consider the mediums patients can use to record and transmit data: over 60 percent of patients want to communicate with providers electronically, while a further 30 percent are eager to utilize their smartphones and tablets. This is huge resource that potentially going untapped.
  • Implement structured and semi-structured forms that capture the data you need from patients, and can be easily accessed on a mobile device through your patient portal.
  • Consider personal health record applications such as Microsoft HealthVault. These PHR tools function much like health record banks, and provide many of the same benefits. HealthVault is even free to use.

PGHD has a great deal of potential, but still needs standards built around it. How will your organization utilize PGHD? Share your insights in the comments.

  • http://blog.bernardfarrell.com/ Bernard Farrell

    For patients with diabetes there’s a class of PGHD that is trustworthy. The data collected by various devices they use: blood glucose meters; insulin pumps; and continuous glucose monitors.

    Unfortunately the device manufacturers don’t make it easy to export the data. Hopefully the growth in HIE and EHR will force them to do so.

Previous post:

Next post: