How Many Cooks Does it Take to Define a PHR Functional Model?

While doing some research on healthcare IT (HIT) standards for our upcoming report on the PHR market I stumbled upon 3 different organizations who all seem to be doing pretty much the same thing, defining a functional model for PHRs. Functional Models are basically just that, a model that defines what functions should be contained within a given software package or even a hosted Internet service. Logical functions for a PHR would include such items as medications list, allergies, current provider, maybe insurer, next of kin, home address and some basic biometric data – you get the idea.

First we have the payer-driven initiative launched by AHIP and Blue Cross Blue Shield to create a set of guidelines for health record transfer among insurers. This is a great effort on their part as among the multiple challenges health insurers face in promoting adoption of PHRs has been the lack of portability.  The average retention life of an individual with one insurer is on the order of 3-5 years, thus this portability issue is a big deal.  In June 2007, they released the technical specifications (warning PDF, all 156pgs!), that among other issues, states what types of data that will be contained within a PHR and need to be accounted for in such a transfer between insurers.  Those data types directly relate to specific functions within a PHR and must be mapped via an XML schema (insurers plan to use a combination of ICD-9 and CCD standards) for successful transfer.

I also knew of the efforts by the standards group HL7, who have developed a functional model for PHRs titled, PHR-S FM. Found a presentation, not from their website (trust me, their website is nearly impossible to navigate and found this via a search), but you’ll find it hl7-overview.ppt. Presentation gives a thin overview of the intentions for PHR-S FM (you’ll find the actual draft here, it’s ninth down in the table). Again, nice effort and looks like many of the stakeholders in the PHR space have commented upon the first draft, back in November 2007 (so it is inclusive). The revised version is currently being balloted upon, which will end April 28, 2008. Thus, we should see the final version hit the streets by summer.

Then today, I find that Project Health Design (PHD), a Robert Wood Johnson Foundation funded project looking at PHRs has created their own functional model for PHRs, which was released for comment in December 2007.

OK, we now have three, verifiable PHR functional models and that doesn’t account for what designs Google, Microsoft or Dossia, may have of their own. Hopefully they’ll adopt one of the above rather than create their own.

There is also the Markle Foundation, who has supported a number of efforts along these lines over the years. Markle, through sponsorship of various committees and work groups, has focused more on policy than prescriptive functional models. Yet, within those policies there is a significant amount of functions outlined. So we really cannot ignore them either.

And let us not forget our friends in Washington DC. While there has been no direct confirmation that CCHIT will take on the task of defining a functional model and/or some sort of certification process for PHRs (they currently have their hands full with EHR certifications), that has not stopped people from talking about it. My view is that it is only a matter of time before CCHIT jumps in as well.

While I applaud the efforts of the above organizations, their hearts and minds are in the right places, I do wish that they would start talking to one another and where possible avoid duplication and bring some clarity and consistency.   Instead, I see a significant amount of overlap, which is ultimately of little service to anyone.  Thankfully, it appears AHIP is thinking the same thing for upon release of their technical guidelines, they did state that they will work with HL7 to reconcile each organizations respective efforts. Hopefully, others listed here will do likewise.