November 8th, 2011 Security Working Group Conference Call
Contents
Security Working Group Meeting
Attendees
- Kathleen Connor
- Ed Coyne
- Mike Davis Security Co-chair
- Jon Farmer
- Suzanne Gonzales-Webb CBCC Co-chair
- Jim Kretz
- Glen Marshall
- John Moehrke Security Co-chair
- Milan Petkovic
- Ken Salyards
- Richard Thoreson CBCC Co-chair
- Tony Weida
- Craig Winter
Agenda
- (05 min) Roll Call, Approve Minutes & Accept Agenda **This meeting discussion will continue into CBCC hour**
- (15 min) PASS Discussion - Security Input for new projects for PASS
- (15 min) HL7 Confidentiality Code Refactoring Project - Kathleen Connor
- 3 types of Policy Data Binding discussion added (e-mail string between Mike Davis, Kathleen Connor, John Moehrke, Richard Thoreson, Tom Bonina)
- (xx min) Revisit - International Member participation - Doodle Poll
- (15 min) HL7 Composite Security and Privacy Domain Analysis Model (DAM) - reconcilliation update
- (15 min) Caveat codes
issues about email topics- the right approach to technical standards regarding managing privacy consents in a scalable way… we are working with the Confidentiality Codes—but it seems like we’re circling; what’s feasible and what isn’t. I’d like to go over this territory again.
- (5 min) New and Other Business
DRAFT Meeting Minutes
Roll Call, Approve Minutes & Accept Agenda **This meeting discussion will continue into CBCC hour**
Data Segmentation under S&I page – you will find the use cases being proposed. They are marginally associated with US-law where data tagging has some potential use. US Code title 38, CFR42, high tech self pay, and emergency access. There are a lot of use cases—for HIMSS we would ‘glom’ them together to show the use of the standards and where they apply and speak to where we see a gap in standards.
It should be noted that Lori Fouquet is driving the use case development (for the interoperability showcase for IHE); so we will have some consistency.
There is a possibility at HIMSS that we are showing what the data segmentation has to offer—but we haven’t really talked about the master plan with our use cases that could be demonstrated across standards organizations.
When there are multiple privacy use cases in a workflow—they seem to talk negatively about the other cases—this appears to show the public that we do not have ‘’ANTHING’’ available. If we coordinated, we can show yes, this is available, this is in progress and this is in the works etc. (instead of implying that OASIS, IHE and other groups are doing separate things) Let’s work on the message we want to display—cohesiveness across the groups.
The last visit to HIMSS (HITSP) the main point of the demonstration was that standards---we could still do things today and we didn’t need to wait for new standards to be developed, standards are not concrete and we expect gaps to be present.
PASS Discussion - Security Input for new projects for PASS
- Discussions held at HL7 for the next steps for PASS (at SOA meetings) so that for the next meeting in January we can coordinate Security, Privacy and PASS
• Context was coming out of the PASS the meeting – rejuvenation a PASS overview. Is there anything new—interesting? How do we engage? Mike brought up a gap and with some overview---S&P meta-data tagging, augmenting… • Making the Architecture overview more ‘cool…’ how does the tagging get done, how are attributes tied to the objects, the tag processing for access control –where do you look in a message that is routed and controlled? These are real-world problems where this is an issue. SOA has a number of hData projects going on—Don would like to know if there are other coordination that we can do to tie in with hData. • What kinds of services are needed, to make useful for processing making access control real. It sounds like Don is aligning an update of the current SFM—with a focus on tagging.
- Kathleen presentation slide deck (need to add link)
The policy binding is the different types/optimal ways to do it, have slides that cover what we did last week. Questions we might ask about it. Mike was asking to have Don’s group look at it from an architectural view The tag processing for access control might have some tie to these use cases and how that might get implemented.
Policy exchange – one or another, do we say there are 3 ways of the policy and I’m using option A (or B or C) then there has to be an agreement of the attributes for interoperability purposes. Policy – what is it… if we are talking Policy – a computable policy in some way. The rule for implementing the policy has to be known and we are provides some . when we talk of data attributes – those are potentially tags on a policy or part of the policy itself.
- I would call this a terminology service, where they have the information objects have the objects for exchange. The terminology service knows what the objects are and the tagging occurs to that level. We don’t want to go back and recompile/rewrite all systems we have to comply with a new standard. We can do this at a service level. Create the interoperable things to exchanged and with tag the object appropriately. There are collections of data that are ultimately defined by the episode.
- There needs to be a terminology service that medical has—security cannot own everything. The service has to be answer … these medical items are what security wants to protect
- Does the group like the idea of a (virtual) tagging service? V-Tag
- There are some rules emerging in the conversation
- Is there a paper written on this subject? (for more information, rather than starting from scratch)
- We have a RIM and CDA document…do we want to create a new thing that is incompatible with what is already there. We do need to get down to facts.
- We can have an invalid discussion—proposing a talk on XML elements and XML attributes on V3 rim compliant payload. How do you derive the properties of an arbitrary node? (but we’ve always had that problem)
- At enforcement time something has to answer the query (is this a psych note?
- If it’s a policy sensitive node—what policies are entailed in that?
The policy is—if I give you this day, you cannot redisclose without some authorization ; our V-tag service and has done its thing and the data has a no Redisclosure tag—when received by the receiving organization, it’s not part of the data; and Dr. Bob and decide to share with another hospital. The policy will say no redisclosure, then the information to e sent out is denies. It seems to me that the tag is retained for future use. That’s a tag on the tag.
Proposal of 2x2 matrix virtual or physical, header or body (content).
ACTION ITEM: Jon Farmer to send out proposed matrix to WG.
Tag definition? Tag = has a property value assertion. You may derive other tags from it… There should be a paper that we can cite—rather than reinvent it.
Meeting adjourned at 12:05
HL7 Confidentiality Code Refactoring Project - Kathleen Connor
- Not Discussed moved to next week’s meeting agenda
3 types of Policy Data Binding discussion added (e-mail string between Mike Davis, Kathleen Connor, John Moehrke, Richard Thoreson, Tom Bonina)
- Not Discussed moved to next week’s meeting agenda
Revisit - International Member participation - Doodle Poll
- Not Discussed moved to next week’s meeting agenda
HL7 Composite Security and Privacy Domain Analysis Model (DAM) - reconciliation update
- Not Discussed moved to next week’s meeting agenda
Caveat codes
- Not Discussed moved to next week’s meeting agenda
