The Direct Project (NHIN Direct) specifications are almost completely aligned with the IHE XDM – E-Mail option. They have simply added that sending a single document without the XDM wrapping is allowed. This is the kind of thing that I would expect a realm specific workgroup to acknowledge.
The Direct Project group is now struggling with Testing. They don’t have experts in writing test plans, test procedures, or test tools. The closest thing they do have is experts on unit-testing that comes along with the source code of their “reference implementation”. I think this is also a good place for a US Realm to focus a local Connectathon toward the local realm use-cases, and tie this to the local demonstration project (e.g. HIMSS Interoperability Demonstration). For this the IHE expertiese in test tools and testing events like Connectathon would be very valuable.
The Direct Project does have a component that IHE does not have today, this is the open-source reference implementation. IHE does come close with the open-source Registry from NIST; but this was more of an individual effort (one that needs to be recognized often), rather than a formal part of specification approval. I think there is an opportunity for IHE to extend a welcoming hand to this style of specification validation.
The Direct Project also has the Implementation Geographies. It is not clear how well this will workout inside the IHE context. I think it is something unique to the NHIN Direct project at this time due to the style of solution vs the ‘pure interoperability specification’ that IHE tends to focus on. I know that there might be concerns with getting too close to implementations, thus presenting a conflict-of-interest.
The Direct Project is an excellent example of what a region can and should do with the IHE specifications. They deviated only slightly, and when they did they provided the feedback to IHE for discussion and iteration. There are some useful lessons to be learned by IHE on how to excite a region and to accelerate implementation through reference implementations and implementation geographies.