Concept Domains are named categories of concepts maintained by HL7. They are referenced by static models in circumstances where binding to a specific set of codes is not possible. Binding to codes might not be possible for a variety of reasons:
- The category of concepts is too broad to assign a set of codes to
- It is not possible to gain consensus on the set of codes
- The set of codes may need to vary based on different contexts
MIF: mif-vocabulary-model.xsd/ConceptDomain
| Requirement
|
Concept domains must have a name
|
| Rationale
|
By definition, concept domains are "named categories", so a name is pretty essential . . . The name is what's used to reference the concept domains from models that need to constrain the set of allowed concepts.
|
| MIF
|
mif-vocabulary-model.xsd/ConceptDomain/@name
|
| Requirement
|
Concept domains must have a formal definition
|
| Rationale
|
The purpose of a concept domain is to provide a label to a category of concepts. A definition is essential to identify the category of concepts the concept domain represents.
|
| MIF
|
mif-vocabulary-model.xsd/ConceptDomain/annotations/documentation/definition
|
| Requirement
|
Concept domains must indicate whether they are bindable or not
|
| Rationale
|
Some concept domains are too generic for it to be appropriate to establish bindings for them of any sort. They are useful in high-level abstract models that are not intended to be implemented directly. For example "ActCode" which covers the concepts that describes types of acts for all possible acts (encounters, observations, procedures, billing, consent, etc.) cannot reasonably be bound to a value set. Instead it is a placeholder for refinement. Specializations of the ActCode domain, such as ActProcedureCode might reasonably be bound.
An indicator provides information about how a concept domain is expected to be used. It also acts a mechanism to prevent bindings from being created where they should not.
|
| MIF
|
mif-vocabulary-model.xsd/ConceptDomain/@isBindable
|
| Requirement
|
Concept domains must be constrainable
|
| Rationale
|
HL7 methodology is dependent on the idea of constraint (see Constraint Derivation). A broad concept domain such as "All act types" in the RIM needs to be constrainable to narrower concept domains such as "All procedure codes" or even "All dental procedure codes" when creating models based on the RIM.
|
| Methodology
|
A concept domain can specialize other concept domains.
|
| MIF
|
- mif-vocabulary-model.xsd/ConceptDomain/@specializesDomain
- mif-vocabulary-model.xsd/ConceptDomain/@specializedByDomain
|
| Requirement
|
A concept domain needs to identify sample codes to give an idea of what types of codes might be bound to it.
|
| Rationale
|
Eventually, the concept domain needs to be bound. In addition, sample instances need to be constructed to document models and it helps to know some example concepts even if there aren't any specific codes bound. To ensure that we can validate that examples are present, it helps to have example concepts expressed explicitly rather than embedded as part of the definition.
|
| Methodology
|
Concept domains are required to have one of the following:
- A binding to a value set in a universal binding realm
- A binding to a value set in a representative binding realm
- A binding to a value set in an example binding realm
- Print names for at least three example codes
|
| MIF
|
mif-vocabulary-model.xsd/ConceptDomain/@exampleConcept
|
| Requirement
|
???? (Waiting on Woody)
|
| Rationale
|
????
|
| Methodology
|
Concept domains have properties (name-value pairs)
|
| MIF
|
mif-vocabulary-model.xsd/ConceptDomain/@property
|
