So, Google Health’s slow collapse — akin to a tire with a slow but obvious leak — has finally come to an end. This week, Google officially ended the project, one of the pioneering efforts in the Personal Health Records space. While GH will stumble along through January 1, 2012, the jig is finally up.
Why did a high-visibility project backed by one of the world’s premier Internet companies fail so miserably? Well, according to former Google employee Adam Bosworth, who first launched GH, the effort failed because “it’s not social,” TechCrunch reports. Another pundit, more convincingly, argued that unless PHRs are tied to reimbursement somehow, they’ll be “irrelevant” for most providers.
So, why should we care about the failure of a project that, I’d argue, was pretty much pie in the sky from day one? And more importantly, is the failure of GH relevant to people who care about the future of EMRs?
Well, for one thing, Google Health does offer some pretty interesting insights into what doesn’t work in the world of patient-centered clinical data. As I see it, they include the following:
* Clinical data projects that aren’t interoperable are eventually going to wither away.
I think it’s telling that Google is, at the last possible moment, rolling out the ability for patients to transfer health data to other services supporting the Direct Project protocol. Also notable is that Google is offering patients the option of downloading data that meets the Continuity of Care Record format. (That’s ASTM E2369 – 05e1 to any standards geeks out there.)
Does that imply that EMRs that don’t share data are going to be outmoded or a waste of time? Certainly not, as EMRs can potentially solve many in-house problems that providers face, and serve a far more expansive purpose. That being said, the failure of siloed PHRs should be a warning.
* Without a live, fluid source of data, PHRs don’t matter.
In this cynic’s mind, the idea that patients would suddenly begin to post data to PHRs on their own was, to put it simply, pretty dumb. Why would they? Consumers seldom think about their health data unless they’re at a doctor’s office, if at all, and they don’t exactly know what do do with the data once they’ve compiled it.
Since day one of the PHR craze, I’ve been wondering who thought they’d change patient behavior en masse by dangling a technology carrot. What were they thinking? I’m not just slamming Google, I’m targeting pretty much any PHR that isn’t linked to an EMR or other clinical data source directly. (I’m talking about you, HealthVault, and probably Dossia too.)
* PHRs must be run by a trusted intermediary, and marketed vigorously to patients, before patients will take heed.
I think it’s no coincidence that while Google’s PHR, and possibly Microsoft’s, haven’t won over many consumers, patients are beginning to pay a bit more attention to PHRs provided by providers and health insurers. (OK, I don’t have hard data on this, just a strong gut feeling — can any of you provide stats that support or contradict this assumption?)
A case in point: While they’ve arguably spent way, way too much to get there, Kaiser Permanente has built what may be the largest PHR user-base in the world — 3 million users as of spring 2010 — linked to KP’s big Epic installation. True, Kaiser had to spend millions in advertising and other forms of outreach to get patients on board, but what’s telling is that patients seem to have stayed once they arrived.
So, I’m just wondering when the managers behind HealthVault will throw in the towel. Hey, MS, just turn out the light when you leave, OK?
[…] EMR and PHR […]
[…] EMR and PHR […]
[…] What Should We Make of Google Health’s Failure? tweetcount_url='http://rdn-consulting.com/blog/2011/06/24/google-health-r-i-p/';tweetcount_title='Google Health: R.I.P.';tweetcount_cnt=5;tweetcount_src='RT @bobnadler';tweetcount_via=false;tweetcount_links=true;tweetcount_size='small'; […]
My current insurer, Aetna, is promoting their own PHR web service, which gets some data form their records, some entered by the patient. (I just got a reminder email about it today)
Does this seem like a viable model? Should I be worried about volunteering too much information to my insurer?
Hi Paul,
Insurers are trying to figure out that model. There are ways they can lower their costs by doing something like that. Lower costs for them should translate to lower costs for us….or at least no increased costs.
I’m not sure how worried you should be, but there could be some concern. Probably depends on the license agreement that you sign/click off. That should tell you a lot about what’s possible.