HL7 seems to be going by the wayside as more vendors favor some flavor of web services - whether it's straight SOAP or ESB - new technologies are supplanting HL7. The venerable HL-7 standard started out almost 20 years ago. Today, the standard is up to a snazzy XML-based version 3.0, but few vendors have implemented it. The vast majority of the market is still running on versions 2.3 or 2.4.

This blog post refers to some MSDN PR about creating a software factory for HL7.

The least one can say
is that this initiative hasn't been a brilliant success. Microsoft was
trying to build the full object model of the HL7 standard (something close to this, I guess), so that applications can be built against it. The HL7 model in itself is overly complex in many ways (and full of Americanisms,
that makes it even more complex when you try to use it on this side of
the pond). Of course, you could say that about almost anything designed
by committees, couldn't you? Maybe using HL7 as a kind of proof of
concept for software factories wasn't the best idea around.

An HL7 interface is still something most products must have, but its use is being relegated to the most casual and arms length kinds of systems integration.