On Jun 15, 2006, at 4:32 PM, CamSchlehuber wrote:

There have been two very different approaches to using the VHIM.  One is

used in the FHIE (Federal Health Information Architecture) and the BHIE

(Bi-directional Health Information Architecture ... to move data in both

directions between VA and DOD).  That approach keeps the data structures

very light-weight and can support virtually any version of the model.  The

information structure for requesting and passing data is computed at the

time of the request and response.  It proved quite useful when the timetable

for delivering FHIE was so short that the code had to be done before the

model was finished! 


It's also vaguely reminiscent of the HL7v3 approach to self-describing messages, at least if I understand your description (!) correctly.

The second approach is to use the VHIM to generate SQL

DDL for a given model version with every relationship hard-wired into the

tables.  (In my opinion, an extremely brittle approach


I'm inclined to agree.

... It's how VistA

was built in the early stages. 


Yes.

Some pioneering work by Bob Andrews and by

Doug Martin showed a much more agile approach!  Unfortunately, in VA this

work hasn't seemed to catch hold.)


It would be interesting to hear more about this. I've seen some approaches (often very interesting) to integrating the object and database approaches, but nothing that I've found entirely satisfactory. Then again, I've had the nagging sense that the whole object oriented approach isn't sufficiently general. Consider that object oriented modeling gives us various types of relationships and hierarchies ("is-a", "has-a", "uses", etc.) but provides no convincing framework that encompasses them all. UML tries to do this, to an extent, with the stereotype concept, but I don't know that it gets very far (or, at least, it hasn't to date).

Gregory Woodhouse

"Judge a man by his questions not 
his answers."   --Voltaire



_______________________________________________
Hardhats-members mailing list
Hardhats-members@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/hardhats-members

Reply via email to