Page tree

Versions Compared


  • This line was added.
  • This line was removed.
  • Formatting was changed.


> add examples, address both directions explicitly

Object graph resolution
In a CDA document, entries are always embedded in sections, and the sections tend to summarize their contents in section.text. This is, in fact the only data that is typically consumed by the receiver. As a result, the entry.text property is often not populated, as it is seen as redundant with the section.text element. Any human reader seeing the section.text will have no use for entry.text.

If FHIR, resources may be presented in document sections, using these same assumptions about context. But they may also be removed from that context and viewed elsewhere. In these cases, we face the question of how to provide resource.text.

1. Put whatever is in entry.text in resource.text. 
      If entry.text is empty, resource.text will be as well.
2. Copy section.text into resource.text. 
     There could be a great deal of redundant data. 
     Readers won't know which part is relevant.
3. Collect all text references from the structured data and put those into resource.text.
      It's unclear how to structure the data to be meaningful. 
     How do we know all the relevant data is tagged?
4. Generate the resource.text the same way section.text is generated
     Should work for entries tagged DRIV (VHIE CCDs are) but doesn't work for entries tagged COMP
     DRIV may not be consistently implemented