Even more on CCR "vs" CCD --- we don't need to choose!
I've copied below a note I just sent to some of the good folks responsible for setting policy around which standards will be considered appropriate for meaningful exchange of clinical summaries. As I've said many times before, I believe that annointing just one standard misses the point -- the hard part is collecting the information to share in the first place. We should be encouraging anything that accelerates that task --- and leveraging all of the work that has already been done against the problem. Once information is available, transforming it between near-equivalent standards becomes a much smaller task.
If you have thoughts of your own --- in support of or argument against my thoughts --- please chime in --- it's important!
Colleagues ---
My purpose in writing is to provide some input "from the field" as you and your committees dig into the hard work of making "meaningful use" a concrete and measurable concept. In particular, I am hopeful that as you consider standards for the exchange of summary health records, you sanction and approve use of both the HL7 CCD/C32 and the ASTM CCR formats. I've also posted this note to my public blog to help encourage more comment and discussion.
Just three and a half years ago, I had very little experience in the healthcare domain --- which has proven to be both a challenge and a benefit. The challenge is of course obvious, but the benefit is perhaps more subtle. Faced with the task of exchanging summary information with the myriad of diverse players in the healthcare ecosystem, we saw the incredible variance in capabilities from system to system, and made an explicit choice with HealthVault to "take what we could get." Rather than forcing our partners to adapt to HealthVault, we asked them what they could send, and worked internally to harmonize and reconcile the information we received.
This is the same integration philosophy that the founders of our Amalga "Unified Intelligence System" took in building a system that can provide comprehensive views of patient data within an enterprise or group of enterprises. But unlike the utterly cacophonous world that Amalga works in, with HealthVault we have found that the world is converging on just two standards for summary exchange: the ASTM CCR and the HL7 CCD (in particular the more structured C32 variant).
Most importantly, it is our experience that neither of these two standards is "winning" over the other in the marketplace. Instead, for many legitimate reasons, different organizations have chosen to use one or the other in what seem to be near-equal measure. The good news is, this works just fine! It is simple for point-to-point or small-group exchanges to choose the format that works for them --- both easily represent the key information necessarily for summary exchange. Further, when connecting more diverse groups that may represent mixed use, systems and technology have emerged that make it easy to transform summaries as necessary to support heterogeneous exchange. HealthVault is just one example of such as system, where in the personal health space we accept both CCR and CCD, and provide tools for our users to reconcile information into a common record (I've included a few links at the end of this message where you can read more about HealthVault's use of the two standards).
At the end of the day, what we've learned is this: the hard work is in collecting complete and accurate summary information in the first place. Many organizations and vendors have been hard at work for the last few years building the code required to do that well. It is far more important to "meaningful use" that we leverage all of that existing work than it is to force it into any one XML standard --- especially when the market has proven that translations between these formats when required are well-understood.
I believe that the best choice to maximize real data exchange is simply to endorse both formats as acceptable for representation and exchange of clinical summaries.
I hope you find this input useful; I am of course available to clarify or expand upon my thoughts at any time if needed.
Related posts:
Comments
Anonymous
July 29, 2009
Sean, Great points! I agree that it doesn't have to be an 'either/or' situation with respect to CCR and CCD. Your input in this matter is highly valuable given the amount of hands-on work you folks have done across different systems. Both these standards provide more of the "envelope" information and the transformation between the two is relatively easier. IMO, what is really challenging is understanding (and mapping) the "semantics" of the codes or terminology used within the payload of these standards. Again I do agree that it is first important to get or exchange the information in whatever format and let our cool technologies (NLP, ontologies etc) make sense out of it!Anonymous
July 30, 2009
Chintan, thanks for the thoughts! You are certainly right that both formats leave a lot unsaid in terms of semantics. The cool thing about them both from my perspective is that they've figured out how to capture a "mid level" of semantics for everyone that has been elusive before. That is --- there's more meaning in them than, say, a random XML document --- because I at least know that a med is a med, for example --- but not so much that people can't actually produce them today in the real world, as with most other "standards" out there. It's this "scalable" structure that makes me optimistic about our chances --- we can start exchanging now in support of really valuable use cases --- and have a clear path to increment our way to higher and higher levels of structure. We actually have a project under way right now to try to help share our interpretation of a higher level of semantics within these documents, and will be releasing our first cuts within a few weeks --- really fun! Thanks again --- glad to have you and TrialX along for this ride.Anonymous
August 16, 2009
Absolutely, I have been saying the same for some time now. When it comes to document interchange, the choice of format is fairly trivial because mapping from one format to another, or amalgamating the information from 2 formats, is a straightforward activity. For example, I spent about an hour mapping most of the HL7 PHM standard used by Continua to Healthvault (<a href="http://softwarepmp.blogspot.com/2009/08/mapping-hl7-phm-to-healthvault.html">here</a>) - a developer could probably iron out the details and code this mapping formally in a day. We heard Healtvault will (or maybe is already) supporting the break down and importing of individual items within a CCD and CCR document. Any plans for HL7 PHM?Anonymous
August 19, 2009
Hey Didier --- this is cool stuff. We have not done a map ourselves to reconcile PHM documents into HealthVault, mostly just because the standard is too new for us to have gotten much demand for it yet. Our next step with reconciliation is to really fix up our CCD transforms; they are ok today but need the same level of work that we've just finished and are about to launch for CCR. After that --- we have a list of other "package" formats that will make sense to include, and we'll just knock them off one at a time. Always more to do in this business!Anonymous
October 20, 2009
Sean: I wondered what thought you might have about the use of semantic web services to locate and present CCD or CCR data from disparate provider EMRs. Could this data then be generated/presented to both the provider and the consumer? I know via ARTEMIS there is some work being done in Europe, but not sure abou the use of such technology in the US. Any thoughts?