|
Presenters: Roxane Townsend, Mark Frisse
We of course did not do our homework appropriately; we got off into
other discussions that we felt were very important. The first thing
we had to figure out was how does this fit into the big healthcare redesign
that has been mentioned multiple times. This is how we think about it,
the healthcare redesign is the whole bean person and healthcare IT is
the backbone. One of the things we talked about was organizational stability.
Mark Frisse: The questions that we asked
raised a bunch of issues. Given that this thing has to be built, what
are some of the things this coordinating team has to get their  head
around?
One is define interim governing entity. You have to have key constituents
and you have to evolve.
You have to define the ongoing costs. For example, we all have to do
audits; what if you were asked to audit 15,000 records?
What are the scopes of services? What is in scope now and what is not?
Each of these things will have to be goals put on your timeline.
You have to start thinking about forming a formal non-profit entity
with a three month target.
Data sharing agreements take a long time.
There are a certain amount of people that have to play; you cannot
leave here without knowing who the key people are. If you cannot get
them on board, you need to give back the money.
We do not think you need to have a user strategy right away; we put
a six month target on this element. I meant the end of my initial user
strategy at the end of six months.
You have to define the benefits, and you have to start now.
These are some tasks that we put forward. Everyone needs to trust this
thing to give it a shot; otherwise it will not be successful.
Susan Christensen: I want to reiterate
that this focuses on the immediate need to get some things done. It
does not preclude beginning the entire process.
Presenter: Mike Kaiser
Our understanding of the scope is that we are supposed to be able to
exchange clinical data across two health providers across two markets.
The partners could be Ochsner, HCSD or PATH.
The next question was, what is the relative importance of the HIE project
in your market?
Compared to internal institution projects (IT and non-IT)? LOW
Compared to other market wide non-IT projects? LOW
Compared to other statewide health information exchange projects? LOW
We have a lot going on and it is hard to make this work a priority,
even though we all did attend this event.
How much have the leading institutions within your market prepared for
exchanging health information?
We think we are pretty well prepared, we have LINKS, CLIQ, Ochsner,
Blue Cross and Medicaid’s databases. We also have the PATH network
which is a safety net alignment for data sharing.
What new information do you need to prepare your market for HIE?
We need to know more about the technical systems for each other
We still need to define entities. We still don’t know if there
is a definition of a minimum data set.
Presenters: Ob Soonthornsima, Stephanie Mills, Jeannie Hinton, Bill
Braithwaite
We think that one of the key challenges is level of participation; another
thing is the degree of difficulty. How do we do this in nine months
and if so, is it meaningful? One of the things we can think about is
when you think about the level of participation, we are sharing the
data today but it is one on one. When you look at this from the other
way around and when you look at a single point of exchange, all of a
sudden trust becomes a huge issue. What we are trying to do is work
on the biggest issue which is trust, which relates to definition, scope,
time and logistics. This graph was meant to demonstrate: what is the
big deal?
Stephanie Mills: We had some really good
discussions and came up with some general concepts we agree on. To elaborate
on Ob’s model, what we are doing now has a relatively low degree
of  difficulty
and a decent amount of participation.
We need to keep it simple. We would advocate for a phased approach,
the toe in the water versus jumping in the pool approach. This will
help align our initiatives and build the trust between us. We need to
include the key stakeholders at the table.
We agree on having shared data elements and standards.
We need more definition around LaCare HIE. Expectations, resources,
standards, and sustainability all need to be discussed.
Jeannie Hinton: We have a lot of data
elements that we can share right now. Looking at low, low, low perhaps
New Orleans is not the place to try this first. To go through some examples
of the data elements, we have prescription information. The data is
there, it is just a matter of creating the pilot and going from there.
Bill Braithwaite: How do we link data from different sources?
Identifying a patient from two different institutions is a challenge
with different patient identifiers, etc.
Second, the methods for integrating data into an EMR may all vary.
The last item we discussed that is needed was that people are reluctant
to share data if they are afraid of being sued. There needs to be some
form of litigation protection.
Presenter: Barr Bauer, Jeff Penton
 We
talked about entities and RHIOs. This is how the plumbing would be hooked
together. People that view this would view through a portal, which we
call a health enterprise viewer. Data would flow up and would be viewable
through the portal. Above this we have a thing we call central services,
or MPI. The things that we want to do here are that we want to display
and share data, and we want to be able to long arm data. One of the
success criteria is the ability to look for a specific portal. This
is the basic architecture. (Refer to Hypertile for detail)
Discussion
Why not one MPI for the region? How does this map onto the clinical
and business issues? All that does is create a lot of layers.
It is not an issue of how many people you can fit into an MPI. The
relationships between the organizations matter.
The MPI gets populated with the patient demographic data. If the organization
participates in more than one, they will have to do this twice.
We talked about this for a long time. Each city could have their own
RHIO but we went with this option instead. We are trying to be as modular
as we can, if one market goes down, hopefully the other markets will
still be up.
I think one of the core concepts we have to resolve before we leave
is where the data is going to reside.
I guess a big issue is how much of a data set do people want?
As a doctor and to the patient, what is the cost of not being able
to see all of that information? There are transaction costs that are
not measurable in the HIT department that we need to think about.
Let’s not forget shadow costs, costs that would happen if we
did what you described. We fake it pretty well, we repeat tests, and
deal with lack of information to treat our patients.
I am being a little Pollyanna-ish here, but at the end of the day this
is the patient’s information.
Let’s take that entire discussion offline.
The basic model we have for the data is what data do you want to share?
That will include any data that you share down here (refer to the Hypertile).
There needs to be an immediate working session to put together what
that continuity of care record (CCR) looks like.
If I take the intersection of all these organizations, does the scope
of this project allow for different people to contribute different things?
We understand that not everyone will have the same data. We recognize
that not everyone has that ability.
What entity will house this?
The state will operate this from Baton Rouge.
For this project, DHS is prime and the state will manage.
I hope you think hard about this, there is not a set of policies for
participating institutions. That meeting should inform the architecture,
because architecture is policy. I would love to see the community here
figure out how you want to dialogue and share; I think this is important
for your  long
term success.
Jeff Penton: As you can tell, we have a
lively group. I am going to start at the end and work backwards. What
do we view as our success criteria?
We saw it as having the ability to have some sort of clinical record.
As we stated before, a working group does need to be pulled together
to figure out what makes that CCR.
Remote documents, or long arm, send us a pointer back to a piece of
information. This is stuff that we do not want to house but that we
do want access to. We put that up as a stretch goal.
We want to demonstrate inter-operability between the two markets and
to share data from external providers.
Data back to providers is a second stretch goal.
The dates that we are looking for are as follows:
April 1: RHIO agreements, BA agreements in place with DHH. Once the
agreements are in place, six weeks later we hope to have our first data
feed on May 15.
The technical discussions can start right way; we do not need agreements.
We hope to have all the data feeds by June 15, around 10 weeks away.
We are going to provide a lot of the applications for your environment.
Once we have all the data there will be ongoing development to provide
this functionality. We hope to begin User Acceptance (UA) testing by
August 15 with a code freeze by August 30.
We want to go live by September 29.
Discussion
The scope is actually what you can provide by entity.
I think we need to be very careful in our assumptions about what the
scope encompasses.
We have two regions. Someone who has a portal will be able to reach
out and get information, but not do the long arm.
There is still work to be done on the portal.
In Tennessee, DAAs went very quick for us, maybe one to two months.
I think this is very aggressive. If you cannot get nine hospitals, perhaps
you can stagger them in. We are a year and a half into this, with live
feeds, but no one is using this. We finally have a system that is useable.
In terms of the contract, it is more than just hospitals, right?
Correct.
I think determining who is ready to play involves determining what
is that minimal data that needs to be exchanged. Take whatever we have
got from whatever entities are in play.
As we have talked about, the exchange of information means being able
to view it, not necessarily handing it off.
When we talk about exchange, it means that I have the information and
someone else has the ability to view it and use it for the patient’s
benefit.
On the first Hypertile presented for Team D, this should say Markets
not RHIOs.
We started with one MPI handling all the markets, if something happens
to your uber MPI, then all of your markets are down.
I would strongly encourage that you pull back the scope of this thing
to not try and aggregate data in a clinical data repository, but that
you create an MPI and generate a pointer to the information. You will
run into some severe problems, for example, laboratory results. Furthermore,
you have 10,000 data elements that you have to aggregate. Do something
that is possible to do, take the data you have; the insurance databases
are very good. I think the scope is way too ambitious.
We need to scope for success, that is the type of data we want to share
and the minimum data set for the MPI.
What is the data set of the CCR? (minimum data set)
What is the minimum number of entities required to demonstrate contract
requirements?
How did this evolve in Indianapolis?
We are centrally managed but we are completely federated. I think it
is very important to determine whom do I trust and who do I share things
with.
Tom Lloyd: We are going to go to work.
You have been working in four areas and now we are breaking the teams
into different groups. We let a little conversation happen here to
begin peel ing away the onion.
|
