There are many issues associated with connecting physician offices running EMRs into a hospital or reference lab. In prior postings we’ve covered:
- The use of standard vocabularies or terminologies such as LOINC.
- The challenges of using HL7 Orders and HL7 Results in a standard way — typically via profiling such as ELINCS profile (also described here).
- Communications infrastructure — using a VPN with a real-time, always-on connection or using an asynchronous method such as web services.
Why do I mention this topic? Because it is “readers write” day over at HIS-Talk and there is some excellent discussion about many of these topics.
I think the labs agree [more standard integration] needs to happen, but just don’t want to invest in it. It is very painful to get a lab interface up and running. Each lab has multiple regions that act differently, have their own compendiums, etc. Because there is no standard test code, all the codes are proprietary. Testing is required for each and every one.
One of the barriers right now is a normal one for our industry: the existence of entrenched systems which would be very costly to change. Since there are many regions with just one or two dominant lab players who control their local markets, there isn’t a great deal of momentum to make the changes happen very fast. However, the ELINCS standard definitely has traction with major players such as the Markle Foundation, CMS, HL-7, etc. and it is also the standard for results for CCHIT certification which is obviously a major force.
By their very design, the use of a standard will require the implementer to jump though at least a few hoops (some of which may be on fire). Also, the device-to-EMR interface you complete today will probably not work for the same device and EMR in a year from now.
Nobody dislikes standards. Interoperability is usually good for business. There are two primary reasons why a company might not embrace communications standards:
- The compromise may be too costly, either from a performance or resources point of view, so a company will just do it their own way.
- You build a propriety system in order to explicitly lock out other players. This is a tactic used by large companies that provide end-to-end systems.
I’ve been to many conferences (TEPR, HIMSS, World Health Congress, etc.), and nobody seems to be able to tackle the thorny problem of semantic interoperability. Everyone can speak HL7, but that’s only half the problem. There are so many different entities that need to agree on what each of those data elements MUST ACTUALLY MEAN that I’m not sure we’ll ever see a solution.
Latest posts by Dave Shaver (see all)
- HL7 ADT Q&A with Dave Shaver - July 2, 2014
- Health Standards Community Membership Archetypes: Who uses HL7? - August 6, 2013
- Note from the Field: Meditech 6.0 HL7 Integration - September 6, 2011