|
Tom Lloyd: Let’s keep this brief,
no more than eight minutes per report. When it comes to the technical
conversation, we want to know from someone who’s worked on the
HHS-ONC contract whether what is being described will support the preferred
technical models. Also, will these preferred technical models meet the
HHS-ONC contract requirements for inter-operability? Okay, let’s
hear from the first team.
Presenter: Tony Keck
 We agreed on where the high resolution system resides. It’s a
governance solution. In the model, the data resides in the native system
and we reach into it through the long arm.
Our assumptions include resource constraints and that New Orleans may
be comprehensive enough to share locally. We also assumed working towards
a long-term solution of CCR. That will take years so the long arm approach
may be better because it’s quicker. The long arm approach recognizes
that the source organization can’t always provide the data when
needed.
Technology allows folks to participate at different levels and times.
The pros of long arming force a strategic path toward a mixed “national”
model. There’s also flexibility and keeping as many participants
engaged as possible.
The cons include a need for authentication utility. There are a lot
of coordination costs involved, as well as service outages causing
data
to become unavailable. There’s also a security issue with credentialing
and audits, as well as coordination costs.
Comment: I’d add performance to
the coordination costs.
Tony Keck: We all agreed we can participant
at different levels with different organizations, but this works for
a start.
Presenter: Eric Ford
 The CCR is causing us angst. Our preferred solution was to create a
minimum data set. There’s some uncertainty about the next module
after that, but there is some eventuality. It will be stored centrally
and there will be a universal viewer. Phase two includes pointers to
additional data.
Discussion
Why the discussion about viewers and why are they tied to the data
set?
That’s a good question.
KatrinaHealth has one viewer. The front-end user doesn’t need
to know what system the data comes from. The viewer is linked to how
it comes together.
That is really the CCR model.
It’s a hybrid.
You have to know how to use the links.
You don’t have to do that.
That means mapping the data and integrating it.
My point is that it doesn’t matter where the data is, it matters
how you see it.
Eric Ford: We assumed that we would prime
the MDS with payer data, and that it would be faster and easier to implement.
The pros and cons were identified. It would be easy at the edge, achievable,
and it will have a large core of available data. The cons include
privacy
concerns and authorization being more complex.
Discussion
Our group discussed the argument that having consolidated data was
clinically most valuable, but the con is that it may not be robust enough.
The lowest common denominator data set may eliminate that. Is that what
you’re saying?
No.
Roxane asked yesterday whether your model addressed what happens if
areas go down. Will that area be a problem?
So you’re priming the pump with care data.
That’s consistent with large data sets.
Crossing Blue Cross and Medicare has lots of data, and then you could
pull out specifics.
In Memphis, we think the assumptions around privacy, etc. really lead
to about five points and questions to answer. We have some great follow-up
questions to work on.
There’s been a lot of talk about backup. It is not more useful
to have a minimum data set. This system will require backup no matter
how you do it. Both will require serious disaster planning and backup.
I agree. That adds another level of complexity, but that’s Phase
two.
Everything everyone wants can be done with caveats. We can handle the
extremes, but there are still governance issues around auditing and
security. That could be worked around, although it increases the complexity
and training for it. How the data is presented will be discussed during
the project; otherwise, I see no show stopping issues.
Presenter: Roxane Townsend
 We talked about a lot of things. We laid out a model showing ONC and
DHH as “clouds”. The steering committee is a bean where
each publisher would publish and be represented. The committee is
between
DHH and the workgroups and subs to ensure the two mesh well. The workgroups
will be driving and ensuring the steering committee and the system
are
conforming.
We identified some next steps and reset issues. We need to revisit
the membership and vote on the steering committee. Consumer participation
was also discussed. The communication and education piece and the feasibility
and business case, as well as evaluation recommendations were other
reset buttons.
The next steps include having the steering committee establish guiding
principles and governance decisions. We also need to set a series
of
workshops and determine a need for legislation and budget that we want
to ask for to cover liability.
Discussion
How will DHH be staffed?
The steering committee will primarily be staffed by DHH.
|
