As I sit here at the start of Day 2 at Health 2.0 Europe, I can't help having a sense of deja vu.  Not just the deja vu that comes from being at a previous Health 2.0 conference in San Francisco, but the deja vu that comes from the feeling that the field of Health 2.0, not the conference, is just passing the old threshold of electronic data interchange (EDI).  As Matthew Holt set the stage for this portion of the Health 2.0 Tools session, he commented that "it is becoming a data utility layer rather than a personal health record."

Unfortunately, watching Roni Zeigler show off Google Health, it is clear that there is not even a real data utility layer for health.  This is not to say that Google Health is not breaking important new ground, but the idea that I have to import information from the Cleveland Clinic or any other source is what gives me that feeling of deja vu.  For all the prowess of Google, this is just a Web 2.0 version of the old EDI approach to moving information.  Once I click that import button, I now have data in two places, Google Health and the Cleveland Clinic.  Once data is replicated, new challenges develop around which data source is authoritative, and the data runs the risk of being out of sync.

So as not to simply target Google's approach, it is equally clear that Microsoft has the same perspective.  Their announcement with Siemens is an excellent accomplishment, however, as I just tweeted from @ANewCLOUD and @CLOUDHealth, "Can the PHR be a 'platform' like the Internet when Microsoft & Google announce partners/customers? I don't license the Internet."

At an intuitive level, a true date utility layer for the Internet can not just exist within one domain.  If it did, it would be like having different flavors of HTML for finance, education, government or health. As we all know, HTML is a presentation layer that like the printing press can print whatever "words" we feed into it.  We don't have different printing presses for different topics, and so, we can't have a separate data utility layer for different topics, either.

The ultimate implications of a true data utility layer in health (or any domain) means that any player in the value chain: doctor, payor, patient or any one else, does not need to know if your PHR is in Google Health or Microsoft HealthVault.  Like HTML, I can view a web page in Internet Explorer, Safari, Firefox or Opera, and I do not need to know how the website I am visiting was created.  Until we reach this same point with a data utility layer for the Internet, we will remain trapped in the world of EDI and keep have a strong sense of deja vu.