The latest issue of the New England Journal of Medicine (NEJM) is a good one. First, is but another paper, this one authored by many including our new head of ONC, Professor David Blumenthal, regarding the abysmal adoption of EMRs At least it appears that Blumenthal is fully aware of the challenges he faces in implementing the HITECH Act to drive greater EHR adoption. (Note: There is a separate, shorter “Perspective” article authored solely by Blumenthal, which will review later.)
Another is the “Perspective” article by Ken Mandl and Issac Kohane, both of Children’s Hospital Informatics Program (CHIP) in Boston. These two were leaders of the core team that developed the Open Source platform, IndivoHealth which became the foundation for employer health cloud Dossia, and arguably, their core philosophy for Indivo is reflected in the core tenets of both Google Health and HealthVault: Personal health data belongs to the consumer and the consumer should decide how it is shared. This is a very radical concept that still has most providers, payers and other data holders shaking in their boots.
In their article, Mandl and Kohane argue that for the HITECH Act to be successful, execution by HHS needs to focus on supporting those actions that will create secure data liquidity and ready substitution of applications. To accomplish such, not surprisingly platforms are their core expertise, these authors wish to see HHS support the development of a platform that supports “communication and computation” of healthcare data.
While noble goals, which we support, are outlined in their paper, these goals are quite obviously derived from those who work in an academic setting and not in business. Unfortunately, their arguments do not account for the current business structure of the HIT industry or for that matter the motivations, or lack thereof for data liquidity among healthcare industry stakeholders. This is one of the core reasons RHIOs have never really gone anywhere.
Probably the most egregious data liquidity example is right here in Boston where the Partner’s IDN does not share digital data/patient records with their competitors literally across the street at Beth Israel Deaconess Medical Center (BIDMC). What is so surprising/egregious here is that BIDMC’s CIO, John Halamka, is the head of HITSP and Partner’s CIO John Glaser, formerly of AHIC is now headed to ONC to oversee execution of HITECH Act. If these two can’t achieve data liquidity between their two institutions, within such close proximity, can others? Unlikely, or not until we see a minor nuclear explosion in the healthcare sector where consumers become directly engaged in managing their health and demand that their records flow into a third-party repository where they can directly control the flow of their health data.
Which leads to the final paper in NEJM and the title of this post…
Doctors Paul Tang of Palo Alto Medical Foundation and Thomas Lee, head of Partners argue in their “Perspective” article that tethered, physician-hosted PHRs will be the true successors in the market as these solutions will provide the greatest value to end consumers.
Whoa, what are these guys smoking?
Yes, at this early juncture in the rapidly evolving PHR market, physician sponsored PHRs are seeing, in isolated incidents, very strong adoption. Kaiser-Permanente has done a great job driving adoption by a continuous improvement process of core PHR functionality that consumers desire. As great as KP’s tethered PHR is, and the uniqueness of KP in the healthcare market (fully, vertically integrated) it does not address the full continuum of care that a consumer may be engaged in. No tethered PHR does. Thus, the need for independent and untethered repositories such as Google Health, HealthVault and maybe even Dossia, that allow for multiple data types and sources to be aggregated on behalf and by consent of the consumer is necessary. This is where the market is heading and not the siloed PHRs that have been the norm, whether sponsored by payers, employers or in Tang’s and Lee’s case, providers. Their argument is just so… turn of the century.
KP, Aetna, United Health Group and others do see the writing on the wall and to their credit, are now working with Microsoft’s HealthVault group to allow a consumer to export their respective PHR records to HealthVault. Cleveland Clinic and BIDMC also support the third-party repository model with Google Health. These are the leaders, the trend-setters and obviously not Palo Alto Medical Foundation or Partners, which really is a shame.
Probably the most galling aspects of their short article though were the following statements:
In the absence of widely adopted data standards for interoperability, however, even these large corporate entities are finding it challenging to import and combine the information in ways that preserve its meaning. Furthermore, the large repository companies have yet to exchange data with one another.
From a consumer’s perspective, the lack of federal privacy protection for confidential health information stored by entities that are not covered by the Health Insurance Portability and Accountability Act (HIPAA), such as commercial PHRs and repositories, remains problematic.
Regarding the first quote, yes, the lack of widely adopted data standards to support interoperability affects everyone. Heck, even physicians can’t even share data today amongst themselves across practices and outside of a defined IDN, so why do the authors take a whack at Google and Microsoft. People in glass houses should not throw stones.
On the second one, the well-worn “Privacy boogeyman” is put out there. Honestly folks, Microsoft and Google have far more to loose if inappropriate use occurs compared to virtually any provider organization. Thus, you are much better off having your data with these third party entities than in a hospital or local physician practice’s EMR where data security is highly questionable.
While Chilmark Research does share the authors’ belief that there is tremendous value in consumer-physician sharing of health data to promote care, we take strong exception to their argument that a physician sponsored PHR is the way to go. Simply put, physician sponsored PHRs are completely siloed, rarely allow for consumer control and almost never support data liquidity. For these reasons, they are but a flagstone in the path to consumer engagement in health and certainly not the end point.
Mandl and Kohane are interviewed by MIT’s Technology Review on their latest article in NEJM that is referenced above.
I would read (and link to you more) if you didn’t have a truncated RSS feed.
[…] will provide the greatest value to end consumers. Whoa, what are these guys smoking?” Article John Moore, Chilmark Research, 26 March […]
Drs Tang and Lee do not appear to have done their homework. Characterizing Google Health and HealthVault as “stand-alone,” and a tethered patient portal as and “integrated PHR,” is exactly backward.
Most patients need both for now:
1. A *patient portal* (what they call the “integrated PHR”) gives patients short-term access to their health care provider–the test results, secure e-mail, online bill payment, etc.). This information is managed by the healthcare organization.
2. A *PHR* like HealthVault gives patients long-term access to their longitudinal health records. This information is managed by the patient.
Increasingly, PHRs will be able to offer secure access to the same features currently available only on patient portals and I expect healthcare organizations will eventually be able to hand off the complex IT roles that patient portals require. Within 10 years, patients will be able to view their current healthcare-related information, using an application of their choice, from any Internet connected computer and patient portals will go the way of the typewriter.
Kathleen, I agree with your overall premise. By definition, tethered PHR’s provide access to data from one institution (a clinic, hospital, HMO, etc) or system (e.g. Epic). Integrated PHR’s, again by definition, accept and display information from multiple sources.
I have to disagree with your labeling, though. HealthVault, Google, and Dossia really aren’t PHR’s at all. Even these companies don’t really consider themselves to be PHR’s. PHR’s are a set of API’s, and consumer directed tools and interfaces, like you see with iHealth, ActiveHealth, HealthTrio, and the myrad of other smaller PHR vendors that are out there (By the way, that isn’t an endorsement of any of these particular PHR systems, just a list of examples). A simplified definition of a PHR’s primary intention is, “to make it easy for patients to interact productively with their healthcare information”.
Tools like Dossia, HealthVault and GoogleHealth are “repository systems”, or “Health Record Banks”. There’s a VERY limited segment of consumers out there that will interact with their healthcare data directly through HealthVault with any kind of frequency that would be beneficial. It’s just not a user friendly interface, and it lacks the value services that the majority of consumers need to engage with. It was designed to be that way. Microsoft isn’t interested in making HealthVault pretty to consumers. They’re interested in partnering with other people who can provide the PHR component. The main value of HRB’s or repositories is to provide connectivity (and secondarily storage capacity for smaller institutions that can’t cost effectively maintain their own servers, etc).
Patient Portals are a much broader group of services, provided by health organizations, in which patients have access to “below-the-log-in” data, whether this is licensed health education content, billing inormation, or other health related services. PHR’s are often part of the suite of services offered from within a patient portal, and some PHR vendors are even starting to create pre-fabricated patient portal services that healthcare institutions can buy as a sort of “plug-and-play” solution, they aren’t the same thing as a PHR.
Excellent comments to which I agree with with only ony small exception. Unlike HealthVault and Dossia, Google Health does have more PHR type functionality and believe longer term they will continue to fill out that portion of the platform looking to partners to provide a higher level of abstraction, most likely along the lines of wellness (keeping healthy people healthy).
One piece that is almost always overlooked is that most – if not all – data for patients still resides in paper (as shown by the very low adoption of the EMR). Hence, driving paper charts into a shareable electronic format is a prerequisite. Even better would be a means to make that electronic paper (as pdfs or images) reasonably searchable.
Is there a potential cottage industry here that would have scanning of paper charts into appropriate formats that then would be under the control of the patient (possibly downloaded to HealthVault, etc)?
This might allow liquidity of the most hard to share piece – the paper chart.
Unfortunately, the RSS feed you get through your reader is the one provided by the host of this blog service, WordPress. Not much I can do today to correct this, but hope you keep coming back and link to the site as warranted.
Agree with everything you say though differ on modalities and time-frame. As to time-frame, consumers will be able to do much of what you outline within the next 3-5 yrs. In 10yrs time, doubt there will be many silo’d provider or payer PHRs as the need for portability and a complete longitudinal record wins out.
As for modality, increasingly, consumers will use their mobile phones to interact with a cloud-based Personal Health Platform. They’ll use their computer for heavy lifting tasks, but day-to-day interaction will be via their mobile phone.
As for your opening line, what I still haven’t been able to figure out is why Tang, who is also on the Advisory Brd of Google Health, would take such a position – odd.
There are already plenty of companies that offer such services. Some, like MyMedicalRecords, who position themselves as a PHR and are on GHealth provide a basic, secure fax service where a consumer, as part of the service, receive a unique fax number that they can provide to their doctor to have records faxed in and subsequently managed by the consumer. Unfortunately, all you get is a dumb scanned fax, no OCR, no data for analytics.
There are also services like Unival that will enter data from faxed in records to specific data fields in a consumer’s HealthVault or GHealth acct. Costs more, but you get data, not just a scanned fax image.
Kathleen and John,
Looking at other industries shows there will always be a place for both tethered PHRs (like KP.org) and un-tethered PHRs (like HealthVault).
Tethered PHRs let healthcare organizations take advantage of market-differentiating and cost-saving e-services like messaging, electronic lab results, electronic questionnaires and HRAs, e-visits, appointment scheduling, ect. Un-tethered PHRs like HealthVault cannot replicate these e-services because the e-services are so tied up in each specific organization’s EMR configuration and workflows. Are you ever going to be able to initiate a financial transfer or send a secure note to you bank’s customer service via a “personal finance record” like Quicken, MS Money, or Mint.com? Unlikely. The parallel is bank customer portals are very much like healthcare organization patient portals (“tethered PHRs”) and Quicken and MS Money are very much like HealthVault (“un-tethered PHRs”). Even if it were technically possible, banks and healthcare organizations would not let go of their e-services because they want patients to come to their website to be exposed to new products and services and to build their brand.
Like you stated, Kathleen, the value of un-tethered PHRs is basically in promising to securely store the patient’s health data forever (which healthcare organizations hosting tethered PHRs have no interest in doing). Of secondary value are the 3rd party applications that can feed off the patient’s data (John, my thoughts on how valuable these apps will be as payors start financing more sophisticated devices are below). I think healthcare organizations will eventually be pushed to linking with un-tethered PHRs by consumer demand or legislation but that’s not where we are now.
So even when un-tethered PHRs are commonplace for average Americans (10 – 15 years?), I expect people to still use tethered PHR e-services to interact with their healthcare organization.
Value of tethered and un-tethered apps –
3rd party applications linked to un-tethered PHRs are of little value today, in my opinion, because there is limited data in un-tethered PHRs. As technology enabling simpler uploading of biometric data like Bluetooth gets cheaper, more health plans will put these devices on their formularies, more patients will use them, and the value of apps that organize pedometer and glucometer data, for example, will increase. However, patients will likely upload their health data to where they are told to upload it. HMOs will want the devices they finance to dump data directly into their EMRs without any PHR meddling (they will not want patients to be able to revoke the connection to the device by altering their PHR privacy settings). Meanwhile, big insurers like Aetna will probably want their patients to upload data to some centrally managed repository which may or may not feed into healthcare organization EMRs or un-tethered PHRs. The only market I see uploading directly to un-tethered PHRs are the health-conscious consumers with increasingly cheaper biometric devices. Hence, if you’re a diabetic at an HMO, you’re going to find a lot more value in the diabetes “app” in your un-tethered PHR which incorporate messages from your care team, target ranges, upcoming appointment data, med regime updates, ect. than in a glucose tracking app you found on HealthVault. So the most valuable “apps” may end up being those built into tethered PHRs instead of 3rd party ones feeding off of un-tethered PHR data.
[Regarding most organizations not wanting to work with un-tethered PHRs, yes, Cleveland Clinic is running a trial with HealthVault. However, I believe that’s mostly because Cleveland wanted Microsoft’s Connection Center, which is currently ahead in cheap device connectivity and only works with HealthVault. As I mentioned, it doesn’t make sense to route device data through PRHs which pins the success of an organization’s Disease Management program on patients not fiddling with the privacy controls. Think of all the false alarms of patients suddenly not uploading health data – alert fatigue is dangerous.]
Correction in the second to last paragraph:
“Hence, if you’re a diabetic at an HMO, you’re going to find a lot more value in the diabetes “app” in your…
“tethered” instead of “un-tethered”
…PHR which incorporate messages from your care team, target ranges, upcoming appointment data, med regime updates, ect. than in a glucose tracking app you found on HealthVault.”
John, to provide a full feed you just need to tell WordPress that you want that. Once you got to you WP dashboard:
– Go to Settings (in the right hand upper corner)
– Select Reading (the second row of menu tabs across the page)
– In Reading Settings select full text, instead of summary where it says ‘For each article in a feed, show” under (it’s the fourth section of options there)
– Save changes and you are done.
Full feed is indeed de rigeur for us, information gluttons, and feed readers are the way to catch up with blog reading, hence the first commenters rather abrupt demand for a full text feed… On the bright side, someone wants to read your posts, so as usual, the user rules! 🙂
Hope this works, and if not, give me a shout (feel free to use the email provided for this comment).
Finally, a great article. I came across it following a winding link path from this post: http://blogs.technet.com/neupertonhealth/archive/2009/04/03/tear-down-the-walls-and-liberate-the-data.aspx
I work on an open source project called The Mine! project, which is meant to provide web users with ability to reclaim their data online, manage them according to their needs and share it on their own terms, without disrupting the connectivity and social aspects of online existence. http://www.themineproject.org
I am interested in personal health records but from the position of someone who is used to manage all my other data using online tools. I have been following MS Health Vault and Google Health Platform and especially Indivo (which I believe is only accessible to those who are part of the Dossia project).
I disagree though understand the point you are making about tethered apps having more ‘useful data’ than an untethered. I just don’t necessarily accept it. I prefer to go the way of Sugarstats http://www.vrmlabs.net/2009/03/sugarstats/, which turn the individual/patient into the pivotal point, not the institutions he needs to interact with. This is already happening on the web with ‘media’, i.e. information, publishing, distribution etc. and there is no reason it shouldn’t happen in other areas of our lives. Will take time to happen, of course, but then, most good things do. 🙂
Adriana, You’re great – thanks so much for taking the time and provide instructions on the RSS feed – not sure I would have ever found that on my own.
Jim and all others using RSS readers, happy to report that you should now be receiving the full post feed. Any problems, just let us know and thanks for your patience.
John, really glad to have been of help!
Incidentally, do you know you’ve not approved my comment, which is still awaiting moderation? I understand you must got diverted by the RSS solution… 😛
[…] of the PHR regarding access or portability – it is “locked” to the hospital. Such silo’d PHR models do not support care continuity nor consumer control. If you go to a competing hospital or a […]
Great Read! Will be back in the future to check out mroe of your posts.
[…] Although I didn’t fess up at the time, I was knocked a bit off kilter by John’s question because I thought my opinion on this was lifted straight from John’s own excellent post earlier in the year. After stumbling through the call, I did a quick search and sure enough—WJS. Here’s a link to John’s post titled “Siloed Tethered PHRs are a dead end.” http://chilmarkresearch.com/2009/03/26/siloed-tethered-phrs-are-a-dead-end/ […]
[…] First, there is the “it’s-my-data” wall put up by hospitals, insurance plans, pharmacy benefit managers, and others. They believe there is some competitive advantage by keeping the data inside their walls. And there are lots of excuses supporting their position–patients don’t want it, they can’t understand it, it might do them harm blah, blah, blah. This wall is starting to crack. Many institutions recognize that the consumer has a right to a copy of their data and are making the appropriate connections to personally controlled health data repositories like HealthVault or Google Health. Many others are writing about it too—John Moore asserted in a recent post, “Personal health data belongs to the consumer and the consumer should decide how it is shared. This i… […]