I spent an hour today on an NHIN Spec Factory call learning about the NIEM. The presentation appears here (Slideshare hated it at 5Mb, so its now on Google Docs). The NIEM is as you have probably heard, what ONC believes to be the framework around which their Standards Harmonization contracts that John Halamka blogged about a few months ago. This was an effort from ONC in the socialization of the Standards framework using the NIEM that was at least partially successful and gets good (but not great) marks from me. It was certainly better than the silence we've been hearing.
I heard a lot of familiar terms: Use Cases, Gaps, Overlaps, Standards Evaluation ... sounded very much like HITSP. But also many familiar processes from IHE (including testing), and the focus on model driven development makes me think that HL7 has quite a bit of expertise to offer as well
Some differences: Pilots and Referece Implementations... The latter are something that is required by OASIS standards before they can become final. Also, the inclusion of interim deliverables and interative ("agile") development. The key to agile is not let it become "we don't know what we are doing so we are going to wing it" development. That's a real danger. You need to have documented processes.
Another big difference is all documentation in one place for an information exchange package. That's something we all wanted out of HITSP but they didn't have the funding to implement.
According to the presenter, ONC doesn't want to write brand new specs, they want to reuse existing work.
It's all got to fit together, and apparently the communications failure out of ONC is in part a result of trying to line up 10 different ducks in a row. Not all of the contracts have been awarded, so it's a bit difficult to communicate about everything at once.
Uhuh. This is project management 101 guys, and you need to learn to be agile too. Agile means being able to change your tactics when the situation changes. If you cannot manage that, how will you manage 10 different inter-related projects. I don't need to go back over what a disaster the communications bottleneck was for the AHIC, HITSP, HISPC, NHIN and CCHIT projects was for the first couple of years under Bush's ONC? Surely this ONC can do better than that
All in all, I feel better today than I did on Monday, although a few slides did give me some pause. After digging into the slide that talked about "HHS creating a Health Information Model" it was clear that as one person put it "ONC poorly worded" this slide. I understand their intent, HHS will fund a program to do it with industry and SDO input, begging, borrowing and reusing whatever they can.
There are a lot of places that NIEM hasn't been yet. Web Services that don't use WSDL (e.g., DICOM), dealing with an industry that has a large installed legacy base, et cetera. This is going to be a learning process for everyone.
OK, so better marks today than previously, and we are all learning as we go. I hope Dr. Fridsma is a quick study.