Documentation and Testing XDD Meeting 2010-09-24 AND 2010-09-30

From Direct Project
Jump to: navigation, search

Meeting: 2010-09-24


Chris Harris from NextGen
Yvan Charpentier from NextGen
Karen Witting
David Tao
Vince Lewis

Comment on Karen's flow


  1. Should note that organizations receiving less than full metadata need manual workflow
  2. Should note that conversion organizations need to map domains to XDR endpoints


  1. Need to reformat the table to make it clear which is sender and receiver


  1. Has some sequence diagrams to contribute


  1. Should have a section to link back to IHE docs and content container spec
  2. Need security considerations section feed from S&T review
  3. May need to address security & trust model conversions (e.g., Direct S&T to Exchange S&T)
  4. Note that HISP security and trust model and the XD* conversion conversion may take place in different organization

Review metadata table


  • Should be the underlying document's author, not necessarily the sender, will stay as R2

Language Code

  • No additional constraints

Mime Type

  • Specify that it should be the same MIME type as the RFC 5322 document

Patient ID


  1. Overloads the meaning of Patient ID, better to specify use of message ID
  2. Better to flag difference between patient specific and non patient specific

(Note that XD* classifications have UUIDs so name collision is no big deal)


  1. Need a third value of null (I don't know which of these it is)
  2. If you include a patient ID, the assumption is that you trust it.
  3. Would prefer R2


  1. Still concerned with auditing
  2. Fine to use message ID
  3. Need to specify that the message ID from RFC5322 needs to get carried over


  1. Just noting that if you don't understand the OID, the harm is minimal

Source Patient ID

  • Move from R to R2, no further guidance

Submission Set Author

  • Should be the message sender

Review SOAP headers

NextGen comments on gaps

Meeting: 2010-09-30


Karen, Beau, Arien, Vassil, Sri, Chris (NextGen)


  1. Review metadata attrs
  2. Review sender/receiver matrix
  3. Review SOAP headers
  4. NextGen comments on gaps
  5. Document TODOs and owners
  6. New classification for how much patient information you have


Author: Change "MUST" to "when available MUST"

Add a slot: authorTelecommmunication

UUID: Must not start urn:uuid, must be unique per doc and submission set

intendedRecipient: should be XON|XCN|XTN
(note that intendedRecipient is required for XDR we think)

Create a new section for XD* Metadata extension

Sender/Receiver matrix:

Collapse receiver column to email receiver

Yvan questions:

SOAP headers:

Vassil: Change To SOAP header to accept multiple To, document why

Yvan's questions:

Do we need reference to original message ID? Answer - handle deeper down in content space (e.g., referral document)

Content questions: yes, handles multiple docs, and handles any kind of content

Section editors

Need section editor for transport conversion:

SOAP-> SMTP, SMTP->SOAP (Arien to write first pass)

We argued ourselves out of the need for a flag to indicate not patient specific

  1. If any of submission set or document patient ID exists, you know the cross community patient ID
  2. If any of sourcePatientId or sourcePatientInfo are filled, you know the sender's take on identity
  3. If none are supplied, you must do further processing, which may include:
    1. Examining classCode/formatCode/typeCode/contentTypeCode, or mimeType to see if the document type or submission type is a known non-patient specific format (e.g., PQRI XML)
    2. Inferring based on other known information (e.g., this is a message channel that is used only for a specific content type)
    3. Shunting the message to a queue for manual inspection and handling


  1. Arien to rename document to XD* Conversion for Direct Messaging and scrub out XDD
  2. Arien to remap sender/receiver matrix
  3. Arien to take first pass at transport conversions, Vassil to review
  4. Arien to see if Vince can take on packaging conversion section
  5. Karen to write context section for Metadata Conversion section
  6. Karen to put in header #s
  7. Arien to clean up extra stuff at the end
  8. Security Considerations: Arien to hand back to S&T WG
  9. Karen to write a section on Patient ID
  10. Arien to write metadata extensions section
  11. Vassil to document multi-valued TO for header and reason why