Google Health PHR and CCR Hacking

Got to playing around with the Google Health API and got myself up to speed on the CCR architecture, im digging the “subsection” implementation of the document and like the RESTFUL ways of the force.

One quick point blares out about PHR’s that I particularly dont like:

  1. The ability to put in your own result.  This may seem like the point, but flag it somewhere as “user inputted data.”  If you want to track your own health I see the point, but for a repository of “certified” and “bought and paid for” results leave them the way of storage commit.  Seems to me that if you mix the two data elements, ones inputted directly by a user, and data elements certified coming from a hospital or care provider, the whole value of the record depreciates in value.  If I were a diggy diggy doc, Id HAVE to take the data presented to me in the CCR with a grain of salt for legal survial and order all the tests again.  Physicians work with the concept of “study” not discrete results.

After being blessed with the development sprint for Google Health, I gazed over the way of HealthVault and made some observations.

  1. HealthVault has a visions for use cases that are WELL thought out.  I think a lot of folks kind of scratch their head and say, yeah, well what is THAT going to do for me?
  2. HealthVault does not have a dll you can slap in a GUI and use, its all web based.  There is a reason for this, but its a difference worth noting.
  3. HealthVault device integration looks like it was built in the land of awesome.

With that being said, I have some homework to do with HealthVault, a follow up post is definitely in order.

With Google Health, I am pleased to announce that I was able to pull results from an ESB, then format it to CCR.  I was then able to apply an XSLT transform supplied by CCR Acceleration Resources Project over at sourceforge, and also jam it into my Google Health PHR.

I rock.