[Openid-specs-heart] Patient Consent

Tue Jul 12 21:25:03 UTC 2016

It might be helpful for us to stipulate that the _confidentiality codes,
when available as metadata, can and SHOULD be offered to the grantor as
part of the resource registration process (aka UMA Phase 1). The
Authorization Server could then define its (internal / local) policies in
terms of the _confidentiality codes and factor _confidentiality along with
a specific Client's assertion on how it deals with _confidentiality codes
(deletes, ignores, passes on the metadata) in UMA Phase 2.

If we make this stipulation, then our profile work can branch based on
whether the _confidentiality codes are:
(a) available (in which case the RS might also inherit the responsibility
to allow the grantor to set or modify the codes associated with particular
resources as mentioned in the FHIR spec), or
(b) allow the grantor to define its (internal / local) policies based on
the RSs published resource set descriptions (basically all of patient-level
FHIR). This option also gives the grantor the option of hiding what it
considers sensitive from the resource server behind the shield of the
authorization server's Phase 2 scope authorization logic and may be useful
even if _confidentiality codes are available at the RS.

Simply put, _confidentiality codes make our quest for a pleasant and
effective user experience MUCH easier but they don't completely solve the
problem. We can stipulate this and move on to the other dimensions of what
patients want from HEART.

Adrian

On Tue, Jul 12, 2016 at 4:54 PM, John Moehrke <johnmoehrke at gmail.com> wrote:

> that is the whole set of securityLabels, inclusive of the _confidentiality
> codes... I am recommending we start small and deterministic with just the
> _confidentiality value-set.
>  http://hl7-fhir.github.io/v3/ConfidentialityClassification/vs.html
>
> John
>
> John Moehrke
> Principal Engineering Architect: Standards - Interoperability, Privacy,
> and Security
> CyberPrivacy – Enabling authorized communications while respecting Privacy
> M +1 920-564-2067
> JohnMoehrke at gmail.com
> https://www.linkedin.com/in/johnmoehrke
> https://healthcaresecprivacy.blogspot.com
> "Quis custodiet ipsos custodes?" ("Who watches the watchers?")
>
> On Tue, Jul 12, 2016 at 3:46 PM, Debbie Bucci <debbucci at gmail.com> wrote:
>
>>
>> To the standards purist .. my apologies in advance - not very good at
>> abstract thinking.  I need to *see* an implementation to *get* how things
>> work.    I get the content of the profiles will be a few levels removed and
>> the need to not tightly constrain anything.  Ok that said ...
>>
>> Kenneth - thanks so much for sending to the list.    After yesterday's
>> conversation, I do agree with John's suggestion that   the consent process
>> is separate from the token transaction (RPT right?).  It may happen at the
>> same time but very separate functions.   Was beginning to wonder how
>> something like C2S or other PDP/PEP implementations would integrate  -
>> coexist  with and UMA AS.
>>
>> Wanted to revisit Adrian's discussion around release of information.
>> For an initial visit -that some how seems important.
>>
>>
>> Is this the correct reference for FHIR  confidentiality codes?
>> https://www.hl7.org/fhir/v3/Confidentiality/index.html  ?
>>
>>
>>
>>
>>
>> On Tue, Jul 12, 2016 at 4:00 PM, Salyards, Kenneth (SAMHSA/OPPI) <
>> Kenneth.Salyards at samhsa.hhs.gov> wrote:
>>
>>> Hi John, sorry if I quoted you wrongly. The bottom line is if you are
>>> tagging something or redacting something you still have to have something
>>> that can facilitate that process; i.e., a segmentation engine or a tagging
>>> engine. Without this capability, tagging or segmenting cannot work. Ken
>>>
>>>
>>>
>>> *From:* John Moehrke [mailto:johnmoehrke at gmail.com]
>>> *Sent:* Tuesday, July 12, 2016 3:55 PM
>>> *To:* Salyards, Kenneth (SAMHSA/OPPI)
>>> *Cc:* Debbie Bucci; openid-specs-heart at lists.openid.net
>>> *Subject:* Re: [Openid-specs-heart] Patient Consent
>>>
>>>
>>>
>>> Ken,
>>>
>>>
>>>
>>> I didn't say that... I said that the FHIR 'resources'  structure is not
>>> appropriate as a segmentation boundary. That is, in normal REST models one
>>> puts sensitive data into different kind of Resources. Where in FHIR
>>> resources contain a wide mixture of sensitive data.
>>>
>>>
>>>
>>> The mechanism that is built into FHIR to enable privacy/security
>>> segmentation is the securityLabel. One can absolutely use this meta tag
>>> system to segment data. This is indeed what I am recommending through the
>>> suggestion that we use the _confidentiality valueset.
>>>
>>>
>>>
>>> John
>>>
>>>
>>> John Moehrke
>>> Principal Engineering Architect: Standards - Interoperability, Privacy,
>>> and Security
>>> CyberPrivacy – Enabling authorized communications while respecting
>>> Privacy
>>> M +1 920-564-2067
>>> JohnMoehrke at gmail.com
>>> https://www.linkedin.com/in/johnmoehrke
>>> https://healthcaresecprivacy.blogspot.com
>>> "Quis custodiet ipsos custodes?" ("Who watches the watchers?")
>>>
>>>
>>>
>>> On Tue, Jul 12, 2016 at 2:12 PM, Salyards, Kenneth (SAMHSA/OPPI) <
>>> Kenneth.Salyards at samhsa.hhs.gov> wrote:
>>>
>>> Hello Debbie, here is a high-level view of the Consent 2 Share (C2S)
>>> architecture. For the RS to be able to do what you are proposing, it would
>>> need to have a segmentation engine to enforce fined grain consent
>>> directives. Enforcing these consent directives is based on (for C2S at
>>> least) value sets based on diagnosis, medications, tests/labs and
>>> procedures associated with the sensitive conditions. We have defined these
>>> types of value sets which are used in C2S for data segmentation. I can send
>>> them to you if you would like to see the content. Also, segmentation can
>>> only be accomplished currently on structured clinical content. We are
>>> exploring ways in which to use the value set content to redact textual
>>> information related to sensitive conditions, however currently we redact
>>> all text in a segmented document.
>>>
>>>
>>>
>>> With the current state-of-art in EHR technology, the content of
>>> exchanged records is highly variable from simple text documents (all CDA
>>> really requires) to documents with text and actual clinical entries. In the
>>> future FHIR may improve this pitiful condition.
>>>
>>>
>>>
>>> As John has said before, there is no inherent capability within FHIR
>>> that can define a resource set that provides data segmentation at any level.
>>>
>>>
>>>
>>> Hope this helps, Ken.
>>>
>>>
>>>
>>> *From:* Openid-specs-heart [mailto:
>>> openid-specs-heart-bounces at lists.openid.net] *On Behalf Of *Debbie Bucci
>>> *Sent:* Monday, July 11, 2016 3:31 PM
>>> *To:* John Moehrke
>>> *Cc:* openid-specs-heart at lists.openid.net
>>> *Subject:* Re: [Openid-specs-heart] Patient Consent
>>>
>>>
>>>
>>> So ... the RS  *should* have an idea of what medications aligns with
>>> each diagnosis.   Wouldn't a patient request to not reveal /release HIV for
>>> some purpose of use info be enough info to provide to the RS to use (but
>>> note the RS may not comply due to various reasons  - but should record for
>>> audit purposes)
>>>
>>>
>>>
>>
>>
>
> _______________________________________________
> Openid-specs-heart mailing list
> Openid-specs-heart at lists.openid.net
> http://lists.openid.net/mailman/listinfo/openid-specs-heart
>
>

-- 

Adrian Gropper MD

PROTECT YOUR FUTURE - RESTORE Health Privacy!
HELP us fight for the right to control personal health data.
DONATE: http://patientprivacyrights.org/donate-2/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openid.net/pipermail/openid-specs-heart/attachments/20160712/4051190e/attachment-0001.html>