Talk:Detailed Clinical Models
UML as formalism for DCMs
Comment by William Goossen on this: This use of UML is agreed in the Patient Care committee and has many experts in favor of this. However, in particular the OpenEHR representatives challenge the use of UML. At this stage UML does allow to specify all characteristics of the examples we worked with. That cannot be done with an HL7 template as they currently exist. It can also not be done in the OpenEHR or CEN/ISO 13606 archetypes. So in order to be complete for the clinical material, we believe UML is appropriate. In particular because the CEN/ISO 13660 standard and the OpenEHR reference information models are in UML and a transform without loss of meaning can be assumed at this stage.
(Moved from front page--Kevin)
Concerns about HL7 would be appropriating the DCM initiative (moved from main page)
During the ISO meeting, parties expressed a concern that HL7 would be appropriating the DCM initiative. This is definitely not the case. DCM will be in the generic, and thus specifically for use in the multiple standards space. However, some parties must do the work. Further, the generic clinical content expressed in a DCM needs a transformation in order to make it implementable. In particular the use of generic and publicly available DCM to create HL7 implementable artifacts will probably be the space to settle HL7 'ownership'. Patient Care WG will have to sort this out in the current underway update of project 320 DCM. In particular we have to consider renaming it the HL7 implementation specification for DCMs.
William Goossen, Durham, ISO, 20 October 2009.
UML as a formalism
So is there some kind of inform/paper/web/wiki page stating what things can be done in UML that can't be done with templates and archetypes? I'd like to take a look to that if possible.