ActRelationshipType cleanup

Problem

There is a degree of variability in use of Act Relationship types and problems with the existing values which preclude safe interpretation. This is an issue for decision support, detailed clinical models (who are primarily concerned with using the right act relationships rather than consuming them), eMeasures, research, public health, secondary use, and which may even limit the ability of human readers to interpret structured statements if an explicit human readable narrative (such as provided by CDA) does not accompany.

Existing Entries w/o Definitions

• CURE
• CURE.ADJ
• DIAG
• MTGT.ADJ

Existing Entries w/ "Challenging" Definitions

• SYMP symptomatic relief has a diagnosis of "Used in the diagnosis of the indicated disease."

Existing Entries Which Seem to be Used as Synonyms in Models

• TRIG and PRCN

The definition of has trigger starts with "A pre-condition that if true should result in the source Act being executed." This suggests that if there is a difference between these concepts that TRIG is a subtype of PRCN.

Requirements

General Case

Most of the general case requirements are met. These include:

• Temporal relationships
• Causal relationships
• Composition

Further Documentation Helpful

Several of the ActRelationshipType' have interactions with specific 'ActMood' values and/or call for addition information.

• PRCN has pre-condition states "A requirement to be true before a service is performed. The target can be any service in criterion mood. For multiple pre-conditions a conjunction attribute (AND, OR, XOR) is applicable." This is likely to require some degree of examples and further clarification for use. Given the vast number of use cases for this specific ActRelationship (and consequences for false positive/negative determinations) we need much clearer language on use than what is currently in ActRelationship.conjunctionCode.