Publishing a value set for a DEF or Criterion Act

From HL7Wiki
Revision as of 14:20, 17 May 2011 by Grahamegrieve (talk | contribs) (→‎Discussion)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Issue

We have no way to publish a value set as part of a DEF, or criterion class.


Discussion

  • Issue that raised this question: (Rene) On modeling the definitions of patient-questionnaire like observations whose instances should have a value from a limited set of values: This is probably modeled as an OBServation in DEF mood, with code an indication of the type of question (type of observation), and value bound CNE to a value set with the limited set of values.


Questions on how to define value sets with models in definition mood were discussed and included:

  • It is desired to send a “preferred” value set for an coded value attribute in a message in definitional mode. The current issue is that the observation.code attribute is a CD data type and this data type does not support the representation of a value set.
  • It is not desirable to model intensional definitions of value sets in the models. Question is whether we want to define value sets or just refer to them? Suggestion that we first describe how to do this by reference and then try to address how to do it by definition. Some difficulties were encountered in this process that shows the limitations of the models to support definitional instances based on value sets. It is proposed that this can be solved either by the template structure or the use of reference range structures that may reference the value set for an observation value. There is still a problem with intensional value sets
  • This process works for observation.value, but does not work for any other coded values in any instance.

The discussion of other possibilities led to a number of approaches, all of which had certain problems that add substantial complexity to the potential solution.

  • Suggestion that a new data type be created to support this concept (there was mixed review of this on MnM).
  • If all value sets were represented in a “code system of values sets”, then you could use the CD as a way to represent this.