Inert vs. descriptive properties of Acts

Introduction

There is a mention of inert vs. descriptive properties of Acts in the RIM narrative for both moodCode and negationInd. During the disambiguation of negationInd it was found that we don't actually have a good overview of which properties are to be considered inert and which descriptive, besides the ones mentioned as examples. This Hot Topic addresses such an overview.

Background

HL7v3 Acts are defined by a set of properties that are a mix of

• properties about the action itself (real-world event)
• properties about the statement/record of the action

For example, an Act in RQO mood is a statement (an order) about an intended real-world event.
But even an Act in EVN mood is a statement (a record) of a real-world event that happened or is happening.

HL7 (and any complex information system) is about information, communicated between agents (human people, machines, etc.). The only way to communicate about actions is by making statements (in some language) about them. As such, there is no action in HL7 v3 without a statement about it. Mood defines the nature of the statement.

While it is important to distinguish between the object-level and the meta-level properties of actions as represented in HL7, they are still inextricably linked. One MIGHT think of this as two classes in an information model: i.e., a mood-independent ActPropositionalContent (with all the object-level, descriptive properties) and an ActIllocutionaryForce (with all the properties establishing the purpose of the propositional content). But it is important that BOTH aspects are inherent in ANY Act and they are INSEPARABLY linked as two sides of the same coin. Therefore, HL7 merges this into one class, since one is unlikely to keep an identified record of an ActPropositionalContent or ActIllocutionaryForce to reuse in multiple Acts.

Definition

Acts have two kinds of act properties, inert and descriptive properties. Inert properties are not affected by the mood, descriptive properties follow the mood of the object. For example, there is an identifier attribute Act.id, which gives a unique identification to an act object. Being a unique identifier for the object is in no way dependent on the mood of the act object. Therefore, the "interpretation" of the Act.id attribute is inert with respect to the act object's mood.

By contrast, most of the attributes are descriptive of what the Act expresses. Descriptive properties of the Act give answer to questions of the whom, where, with what, how and when of the action. The whom, by whom, where, and with what, are answered by Participations, while the how and when are answered by descriptive attributes and ActRelationships. The interpretation of a descriptive attribute is aligned to the interpretation of the entire act object, and controlled by the mood.

Another way of looking at it:

• is the property about the real-world action? then it is descriptive, it is object-language
• is the property about the record of the action? then it is inert, it is meta-language

Relevance

The nature of an Act property as inert or descriptive might not seem relevant to its use in a V3 message model. However, it is very relevant to the exact semantic of the property. For example, since the 'author' ActRelationship is inert, the author of an order (Act in RQO mood) conveys the author of the order itself (who wrote it). On the other hand, the 'performer' is descriptive, so the performer on an order conveys the (intended) performer of the event that is ordered to happen.

This becomes most apparent when related to the use of the negationInd attribute of an Act. This attributes negates all the descriptive properties of the Act, but has no effect on its inert properties. In effect, if you would view the Act as a statement, and would combine all its descriptive properties as an AND statement, setting negationInd to true would be like putting logical NOT in front of that statement. For further eleboration, see the RIM narrative for Act.negationInd.

List of Act attributes

Inert

• moodCode (added by Tom)
• id
• title
• text
• statusCode ???
• availabilityTime
• confidentialityCode
• uncertaintyCode (after correction by Gunther)
• languageCode
• levelCode (remove altogether?)

On statusCode, Gunther wrote:

I think statusCode is about the action, not about the record of the action. An active event is something that's going on now. Not an active record. OTOH, it appears an active order is about the record. This is a conundrum. There are SOME status codes which are always about the record, never about the real world, e.g. "nullified", but others may be viewed differently.

Descriptive

• classCode (after correction by Gunther)
• code
• negationInd
• derivationExpr (added by Tom)
• effectiveTime
• activityTime
• priorityCode
• repeatNumber
• interruptibleInd
• independentInd
• reasonCode

NOTE: the attributes of specializations of Act, notably Observation, should still be categorized!

List of ActRelationship types

Descriptive

• Conditional (after correction by Gunther)
• Component
• Pertains (added by Gunther)
• Outcome
• Sequel (after correction by Gunther)

Comment by Tom:

I would argue that Pertains is inert, in the sense that for an order this is pertinent to the order, not necessarily to the ordered event.

List of Participation types

Inert

• InformationGenerator
• Custodian
• InformationRecipient
• ResponsibleParty
• Verifier

Descriptive

• Ancillary (consider moving CALLBCK to InformationRecipient?)
• DirectTarget
• IndirectTarget
• Location
• Performer