Jonathan Razza is the director of solutions engineering at Liaison Technologies.
I was recently at the at Interconnected Health 2012 conference in Chicago, and amongst the various topics being discussed, the most common theme popping up is the need for more standards to allow PHR, EMR and HIE networks to really take hold. I’m in no position to disagree with this, but I do find it interesting that an industry that uses so many acronyms (many of which represent existing data model and messaging standards) still feels like they don’t have standards that are sufficient to solve their data integration challenges. Is it too much of a ‘pie in the sky’ idea to establish just a few data model and messaging standards that cover nearly all use cases, and expect wide adoption of the standard across the industry? Maybe it is.
Many of the problems the healthcare industry is trying to solve are similar to the problems that companies in supply chain struggled with over a decade ago. Granted, there are some added complications such as privacy/consent (HIPAA), the need to track a long-running history for a patient record, and inherently more complex and semantically ambiguous data sets. But then again, technology is more than a decade ahead of where it was when supply chain went down this same road, and we should be able to re-use many of these same concepts. Existing standards will continue to evolve and new standards will be created, but are new standards really the solution to today’s integration problems? Just ask any mapper how many different document standards they have mapped to/from, and how many times they have had to create maps to and from what is supposed to be a single standard and it becomes obvious that 1) a common standard is not enough to integrate systems together, and 2) integration between two different standards is really not a big deal.
Healthcare companies could be making better use of existing standards that allow for customized extensibility such as HL7, and go ahead and use this to represent their data and integrate with companies using perhaps a different version of HL7, their own extensions to HL7, or really any other proprietary or public standard out there such as NIEM. Rather than wait for the “perfect standard” to come out in hopes that it will solve all problems and cause widespread adoption, why not go ahead and use the best standard out there and start reaping the benefits of integration? Why not go ahead and start saving money, get ahead of the competition, and maybe even save some lives in the process? Standards are useful, but interoperability really does not require both endpoints of an integration to use the same standard. Integration-as-a-Service (IaaS) providers are in business in large part because reality is, companies who need to integrate with each other usually don’t agree on which standards to use.
This xkcd comic shows the problem with waiting for the “perfect” standard: