This week, I read an article from February that provided an overview of the health IT infrastructure required for population health management (PHM). It had thorough examples and some nice graphical depictions of delivery systems and budding ACOs taking on the challenge of marrying encounter data from claims systems with clinical data from EHR to create “a 360 degree view of the patient.”
But two questions arose as I finished reading.
First of all, the article was from February 2013. Over a year ago. As Cora recently reported, progress remains fuzzy in this market, and it seems like little has changed. In these fast-moving times, when a year-old article reads more or less like the rest of today’s web fodder, there’s a problem.
Has our understanding of effective PHM evolved at all over the last 12 months?
Secondly, the talk of a 360 degree view is simply misleading. Claims data anchors us as patients to the transactions we take part in, while EHR data is more robust but rooted largely in what happens in those magical 12 minutes we spend with some nurses and a doctor.
Is there more that goes into “a 360 degree view” of our health?
Yes to both. To the first, patient generated health data (PGHD) is emerging in 2014 as a recognized data layer that is available to be captured and crunched. To the second, we now have dozens of tools to capture patient data in a variety of formats and settings; progress over the next year should be measured in incorporating that data into a longitudinal patient record.
The basic need for PGHD in PHM is one we all understand at this point: A population with diverse, dynamic health care needs presents many blind spots for a system to manage effectively, especially on a person-by-person basis. By creating a window into a patient’s life outside the walls of a clinical setting, as the story goes, these new data sources can assist clinicians in the full scope of care delivery, from prevention to maintenance to post-acute monitoring. Nothing has done more for the #PGHD movement than the surge in wearable computing, with tech titans Samsung and Apple entering what looks to be a sustained slugfest over consumer smartphone health applications and wearables.
Meaningful Use’s Stage Three requirements, while still far from complete, will offer the most definitive guidelines for incorporation of this data into the healthcare enterprise. Encouragingly however, industry seems to have taken the lead here, with early indications from Washington suggesting they will play supporting role to what is already happening in the private sector.
Challenges remain to be sure, but the growing wave of efforts we’ve seen early on in 2014 is particularly encouraging. HCOs of all types are starting to see the value in PGHD, and that they are not going to wait for further guidance before moving forward. There are many types of data sources that are available, each with their own potential benefits and challenges, but across each area steady progress is being made.
PGHD: Early Categories, Benefits, and Challenges
|Type of PGHD||Benefits||Challenges|
|HRA||Widely available; broad industry traction; steady incorporation into PHM engines||Limited insight; low frequency of collection; not applicable to all populations|
|Surveys/PRO||Growing source of data; largely untapped potential for insight into pt behavior; actionable info at point of care||Integration requires custom wiring; lack of standardization of both instruments and collected data|
|Device Generated Data||Passive monitoring; increasing consumer buy-in; potential for advanced chronic disease mgmt.||“Wild west” with lack of standardization and format; aggregation/level of analytics primitive; workflow issues abound|
Health insurers have long been folding in HRA data from their employer groups on top of their claims and TPA data to target specific cohorts for patient engagement interventions spanning disease counseling, education, classes, and more. Vendors like Phytel, Truven and several others have been pulling these capabilities into their portals and PHM suites over the last couple of years. These data-sets represent valuable low-hanging fruit for a notoriously backwards swath of publicly-insured care delivery settings, and hold promise for the rapid expansion of state-led Medicaid managed long term care programs.
A more frequent source of between-visit patient data is being captured in short-form surveys and questionnaires administered across web, tablet, and mobile platforms. These data-sets represent valuable insight into behaviors and preferences that have simply not been captured in conventional clinical visits. PatientPoint’s recent partnership with Xerox is one example, while several startups such as Conversa and Roundingwell have incorporated “digital check-ups” into their physician group-targeted offerings. These interventions are more sophisticated than an annual HRA, but require more two-way wiring to push specific questionnaires to specific cohorts and pull responses back to care managers to make them actionable.
Finally, the hottest topic in health IT today is wearables and device generated data. After speaking with a couple of the leading players in this field, Chilmark is encouraged that we will see broad incorporation of wearable device data into some EHRs by the end of the year. PracticeFusion announced alliances with two mHealth plays last month: PracticeFusion announced alliances with two mHealth plays last month: the AliveCor ECG will be available for physicians to use during a patient visit, while the DiaSend app will similarly allow doctors to upload patient-collected directly from a variety of devices into the EHR during a visit. The source we spoke with described this as an admittedly early effort that has some unanswered questions around data granularity, but in a refreshingly progressive approach, one that will be available to their entire customer base at once.
Samsung-backed diabetes play Glooko has likewise entered into a partnership with Joslin Diabetes Center that incorporates glucometer output data on top of a native NextGen EHR via an API, enabling at-risk providers to track patients with less lag. While they’ve indicated they’ve had serious interest from some dominant EHR systems in building out integrations, data currently pulled out of Glooko remains in flat-file, PDF format. There is not yet any industry-wide consensus about what a consolidated CDA would look like here, so we are only seeing the tip of the iceberg of how such data may impact population health analytics.
Early movers’ willingness to wade into these uncharted waters is encouraging, though they face challenges both technical – such as a lack of common standards around the dozens of device data types – as well as organizational – workflow, point of care vs. managing a patient panel. In a telling sign for the PHM market, these disparate sources of PGHD are being leveraged despite the lack of guidance and the lack of mandate. So while we anticipate data will be piping into EHRs and other enterprise data engines by year’s end, there is still substantial work to be done before these data can be sliced, diced, and made an actionable component of a PHM strategy.
We see this as an early sign that the heretofore fluffy concept of patient engagement now has teeth and is becoming a key piece of the puzzle. That being said, no one single delivery system, vendor, payor, or other stakeholder has yet pulled together these disparate concepts under one definitive PGHD strategy. Chilmark will be delving into more detail around this area amidst the maturing PHM market in an upcoming Insight Report, available to CAS Subscribers, as well as through our forthcoming market trends reports on CAPH and Patient Engagement. Stay tuned for more, and as always, please share your thoughts, insights, examples, objections, and any other comments below.