Notes from Content Packaging Workgroup
Status of Notes: DRAFT
Date: April 14, 2010
Time: 1pm-2pm
Attendees: Honora Burnett, Arien Malec, Lin Wan, David Kibbe, Saurabh Singh, Vasu Manjrekar, Vassil Peytchev, John Moehrke, Keith Boone, Paul Saxman, Karen Witting, Vince Lewis, Mark Stine, Chris Lomonico, Omar Bouhaddou, Jim McNeal & Rich Kernan
Actions for This Week
#
Date
Action
Status
Owner
Due Date
12
4/21/10
Arien will write a Requirements Document for Content Packaging Specification
Open
Arien
4/28/10
13
4/21/10
Vassil will provide a “layering discussion on the Wiki
Open
Vassil
4/28/10
Actions from Last Week
#
Date
Action
Status
Owner
Due Date
8
4/14/10
Provide post on the wiki further thoughts on an alternative, super simple content packaging formula that meets use case of presupposition – that is simple from developer and end user perspective?
Closed
Arien
4/21/10
9
4/14/10
Provide post on applicability of ZIP on Wiki
Closed
Keith
4/21/10
10
4/14/10
Provide Post about MIME on Wiki
Closed
Vassil
4/21/10
11
4/14/10
Next time discuss message signature heading
Closed
All
4/21/10
Notes
Agenda and Framing
· Review of last week’s action items
· Discussion
· Next Steps
Discussion & Framing Issues
1) Difference between header metadata and content package metadata
o Policy issues with PHI in header metadata
o Internet Message Format headers in current specification as strawcase

o Listen to Devon McGraw’s speech to HITPC Privacy & Security Policy Workgroup
§ Has standing on whether we encrypt the payload
o Discussions that have evolved have confused even David so he will restate it:
§ We need to decide the minimum required or mandatory header data
§ Common use case: XDS derived content/messages then how do we sync up these headers
§ Then metadata inside an object
§ Simplest possible HISP would never see anything other than encrypted payload
§ Policy implication about what goes in header vs content
§ Can’t simply say the person-person message takes away all the complexity surrounding this, there is still complexity
o Comment from David Kibbe
§ Separate metadata in the header from the package
§ Don’t understand with why we’d be concerning ourselves with what is the metadata in the package, but not in the header
§ Why are we concerning ourselves with this
· Discussion about: Agreement important separation between header/content metadata & using internet metadata format as the minimal header metadata
· Comment from Keith Boone
o Header information about the message should be distinct from the content
o Issues to bring up about the specifics of MIME field
§ Minimum sets is applicable if you are doing SMPT, but different for REST and SOAP
§ Internet message format isn’t MIME
§ Discribing minimal set of metadata, and this is a starter set
§ Affinity domain associated with NHIN work?
· Comment from Mark Stine
o Not requiring the date field

2) Minimalist use of metadata strawcase
o Multipart MIME as baseline
o Minimalist XDS metadata: http://nhindirect.org/message/view/Content+Packaging+Workgroup/23045631
o Simpler XDS notation or easier path to expressing simplest metadata http://nhindirect.org/message/view/Content+Packaging+Workgroup/23044739#23097977
Discussion: we’ve got multi-part MIME as minimalist, is there a step up using XDS metadata
· Comment from Keith Boone
o Usual case, find agreement between document and header
o Use cases: newborn delivery
§ Document is summary of labor/delivery that included health information about mother/newborn
§ Exchange and local identifiers may not always be the same
§ Normally most common cases say that these are tightly linked, so we should document them – close ties and could almost automate the generation of XDS metadata when it is that constrained
o XDM model as a grouping tool
o How much metadata should be reflected in the XDS header as to just trusting the metadata in the documents
o Issue in discussion from Wiki
§ Text message, and an attached healthcare document (CCD) that minimalist use doesn’t give a good way to express “this is about the same patient”
§ Don’t be reliant upon complex HER processing
o Should we layer the straw-man spec (second layer: how to use XDM, HL& 2.5 in a later layer)?
o Cleaner distinction between headers concerning the message lower level transport and the use of XDS and a specific type of includable object
o Requirements document for CP specification
o Arien will write a Requirements Document for Content Packaging Specification
· Comment from Lin Wan
o Implication for the receiving system
o Email client could ignore XDS Metadata and format up like an email – doesn’t preclude getting into the content
o ZIP Package contains multiple set of metadata inside it
· Comment from Vassil
o Still some layering confusion in what we are layering. Vassil’s understanding is:
§ Information that is necessary for the routing without information about the content
§ If we use XDS metadata there are certain things that will help us achieve – the information that is associated with certain patient can be anything
§ One way to package this metadata with associated topic is XDS or over email in Zip file
§ Provides a directory structure so that when the file is unzipped the HTML file will provide links to other content in the package
§ Vassil will provide a “layering discussion on the Wiki
· What parts are optional, recommended and required at each level?

Next Time:
3) Standardization of mutually specified by Affinity Domain XDS metadata items (C80 sufficient?)
4) Packaging XDS metadata strawcase
a. Use of multipart/related as with current strawcase?
b. Only XDM zip files?
5) Another addition from Vassil Peytchev: Content Types
6) Another addition from Paul Saxman: How does MIME/XDS fit in with the other WGs?