[This Transcript is Unedited]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
June 17, 2010
Sheraton Crystal City Hotel
1800 Jefferson Davis Highway
Crystal City, Virginia
Proceedings By:
CASET Associates, Ltd.
Fairfax, Virginia 22030
(703) 266 8402
TABLE OF CONTENTS
- Call to Order, Review Agenda: Harry Reynolds, Jr.
- Community Data Initiative: Linda Bilheimer
- Subcommittee Report Outs
- Plans for September 2010 and Future Meetings
P R O C E E D I N G S (9:10 a.m.)
Agenda Item: Call to Order, Review Agenda
MR. REYNOLDS: Welcome to the second day of our hearing of the National
Committee on Vital and Health Statistics. NCVHS is the main health policy
advisory committee to the Secretary on health information policy. My name is
Harry Reynolds. I work with IBM, and I will chair the committee for at least
the next few hours.
As is always usual when you are a lame duck, the World Cup has taken over
today, so I appreciate those of you that tore yourselves away. In a minute I am
going to open it up for introductions. You can if you want to say where you
work and the country that you are rooting for if that is a necessity. But we
will go around the room and get introductions. Members of the committee, if you
would please note whether or not you have any conflicts of interest based on
this morning’s meeting. I do not.
(Whereupon, introductions were performed.)
As we mentioned yesterday, we will just be meeting this morning. We will be
hearing about the Community Data Initiative and the subcommittee report-outs
based on the discussions yesterday. Then even though at 11:00 it shows that I
will be discussing the September 2010 and future meetings, I believe that is
when the Celtics fan will step up to take over the discussion there, so that I
don’t leave this in any inappropriate way as we move forward. Then we will be
moving over to the NAS Keck Center for our celebration.
It is nice to have Ron in the room. He was the chair in 1986. I had a
wonderful opportunity to spend time with him in Charlottesville. If you get a
chance, like that spirit we talked about yesterday, you will feel the spirit
from 1986 just like we feel it today. It made a huge difference in everything
that is going on. So it is an honor and a privilege for those of us that know
him to have him join us this morning and later this afternoon.
Our first discussion is about the Community Data Initiative. Jim Scanlon
mentioned a little bit about that yesterday. I would like to now turn it over
to Dr. Bilheimer.
Agenda Item: Community Data Initiative
DR. BILHEIMER: Thank you very much, and good morning, everyone. I would
like to apologize in advance. We had a little bit of difficulty getting this
presentation uploaded. It is going to send periodic rude messages across the
screen that I will delete very quickly, but that is some of the complications
of mixing two technologies, so I apologize in advance.
I want to talk to you today about two things, the Community Health Data
Initiative and the Health Indicators Warehouse which NCHS is developing, that
are very closely linked, but are actually separate and distinct entities. Both
very exciting, and I hope to share some of that excitement with you about what
is going on in HHS.
I am going to start by talking a little bit about the Community Health Data
Initiative. It is, for those of you who have read the HHS open government plan,
you know that this is one of the five flagship initiatives of the HHS open
government plan. It is designed to promote the dissemination of national, state
and community level data that can be used to develop applications to promote
health at the community and state and national levels and promote action and
promote feedback to HHS about health issues at the community level.
The person who has really headed all of this is Todd Park, whom many of you
know, who is the recently appointed HHS Chief Technology Officer. He envisioned
when he came on about a year ago the development of a network of people who
supply data like us at NCHS, and then people who developed applications that
would use those data to create value for end users.
As part of that, he wanted to see how we could provide a wide range of
health indicator data in a very user friendly form to a wide variety of users,
and then how we could recruit people to come and use these data in ways that we
had never ever imagined before. And believe me, I had never ever imagined some
of the ways in which people are now using our data. Then use these applications
to create awareness in communities about the health status of their population,
about policies that they might want to consider or change or interventions that
they might want to develop.
This is a very complicated slide which is a design of a conceptual
framework for the Community Health Data Initiative. Over on the left you can
see this target with three rings in it. Right in the center are end users.
Those end users may be patients, they may be consumers, they may be health care
providers, but they may be a whole lot of other people, too. They may be real
estate developers, they may be corporations, they may be researchers. The end
user community in this vision is very broad and very eclectic.
Then in the next ring outside that ring of users in the middle are the
potential applications. Some of them seem relatively straightforward, best
places to live, ratings and rankings, interactive health maps. Some of them may
not be things that one is necessarily quite familiar with. People got a little
nervous when they saw viral gains out there, and then things like enhanced Web
searches and challenges to communities to promote new ideas and new
developments.
Then the outer ring are the innovators, the people who do all of this. They
range from the traditional data analysis firms to a wide range of other
potential innovators, Web search providers, nonprofit organizations, technical
companies and so on.
So that is the framework for the initiative. But then off to the side there
is the issue of what are the data that you are going to use. The data are
coming from a variety of different sources. The main contribution that HHS has
made is this first one, the Health Indicators Warehouse.
The Health Indicators Warehouse is something that we at NCHS envisioned
about a year and a half ago. It had nothing to do with the Community Health
Data Initiative at all. We now provide health indicator data for a wide range
of users. Some of them you are very familiar with, Healthy People for example,
Healthy People 2010 and now Healthy People 2020. We are also though developing
indicators for the Community Health Data Initiative, for the county health
rankings that used to be known as Match, and also for the state of the USA.
What we found was that we were using an enormous amount of resources,
programming almost identical indicators with slight differences, maybe in age
breaks, maybe a slightly definitional change, for a wide variety of
organizations.
What was happening was, first of all, we were using a lot of resources,
essentially doing the same thing over and over again with a slight twist.
Second, there were many different indicators or metrics for the same essential
concept beginning to develop out there. We thought this was both inefficient
and also not good for users who wonder why in this set of data it says that the
childhood obesity rate is X and then in this set of indicators it says the
childhood obesity rate is Y. It is because of slightly different age breaks or
definitional differences or whatever.
What we thought about was the development of an indicator warehouse that
all of these organizations plus the public and a whole range of other users
could use that would be user friendly, have a very nice front end with a lot of
mapping capabilities, trend capabilities, so that anybody could go in, draw
down indicators and also link them to evidence based interventions.
We had started on this when Todd Park arrived in Washington last August. He
said we are looking for an indicator warehouse and we said, actually we are
developing an indicator warehouse. The rest I guess is history. We are now
developing the data source that is going to be the HHS hub for the Community
Health Data Initiative.
But it is not the only data source. There are many others and they are
growing. There are other federal data sets, and I can tell you about some of
those. We are getting state and local governments interested and they may be
providing data. There are private sector data suppliers and a host of potential
others. I can show you one very interesting example later this morning.
We thought about this and Todd Park thought about how we should be going
forward. He was doing this in conjunction with the White House. They organized
a meeting in March of this year in conjunction with the IOM. It was a
relatively small group of people, about 40 people at the IOM, and it brought
together Web 2.0 apps developers, business leaders, community leaders, data
people like us from the government. We were basically charged with thinking
through what we could do with community level data and what might be done to
develop some applications within 2010.
It was really a remarkable day. At the end of the day there were about 19
proposals for applications using community level health data. What was even
more surprising, a host of organizations volunteering to develop them in 2010,
if we could get them the data, which was a challenge.
At this meeting, HHS essentially made a commitment of what it would provide
in the way of data for 2010. First off, and what has got many people very
excited, is a new set of Medicare community level indicator data that has never
been released before. CMS is currently working on those indicators, and they
will be in the Health Indicators Warehouse.
The second was the Health Indicators Warehouse itself. We made a commitment
that we would have a version 1.0 of the Warehouse up and running by December of
2010. Do I have many sleepless nights? Yes.
Then third, because we had people who were ready and willing to go to
develop apps and to demonstrate how data could be used, we were charged with
seeing if we could put up an interim website with data on it, not with all the
user friendly content and the bells and whistles, but a basic website that
people could go to that would have data that they could experiment with. Those
data were for Medicare Compare, Healthy People 2010 because we don’t have
Healthy People 2020 available yet, though we will be by December, the county
health rankings data, the community health indicators status data, the SUSA
indicators, and also the AHRQ state profiles and the USDA Food Environment
Atlas.
I am very proud to say that NCHS through an incredible team effort got that
worksite up in a week. It is now available and I can send you the links to it
with links to all of those data sets. But for the most part they are all Excel
spreadsheets. Like I say, they have no bells and whistles on them at all, but
the data are out there.
Frankly, I was not anticipating that we would get an enormous outpouring as
a result of putting it up there, though I was very pleased that we had. We
should think again. Four months later in the beginning of June, HHS organized a
launch meeting for the Community Health Data Initiative, again organized by the
IOM and led by Secretary Sibelius. It was the hottest ticket in town. We took
over the whole of the building on Constitution, which had been closed; the
National Academies building has been closed for a year for renovation, but they
opened it specifically for this event. We had apps developers from all over
coming to demonstrate the apps that they had developed, for the most part from
the data that we had put up on the interim website, but in some cases from
other data sources which were equally interesting and intriguing.
The purpose of this meeting, we had a huge group of people there, and
others watching on a webcast. The webcast is still available for those of you
who want to see it. The idea was for us to get feedback, to demonstrate what
could be done to get apps developers and community organizations interested in
taking this forward and to generate ideas for the future.
I am just going to show you a few of the applications that were
demonstrated at this meeting. This is a screen shot of the Analyze the U.S.
website of Palentir Technologies. Palentir Technologies has primarily been
involved in visualization activities for the Defense Department and the
intelligence community. They were extremely excited about the idea of having
community health data and they gave a presentation that really wowed the
audience, to put it mildly.
What they did, they took a very specific set of data from the website which
was childhood poverty rates. They took county childhood poverty rates in Texas
and they identified pockets of childhood poverty in Texas, and then they
overlaid on that — they collect any piece of publicly available data from
government. That may be Script, it may be Tweets, it may be anything. They
organize it and overlay it on the data set.
What they overlaid on this first off was all the community organizations in
those high child poverty areas in Texas that serve children in need. You could
see the disconnect between where the organizations were and where the really
poor children were. Then they overlaid on top of that federal grants, and they
showed where money to support health and welfare of children was flowing in
those counties or if any was flowing in those counties at all, relative to
where the poor children actually were. It was an amazing presentation and got
the audience’s appetite quite whetted.
The next is Microsoft Bing.
MR. REYNOLDS: Doctor, are you willing to take questions?
DR. BILHEIMER: Sure.
MS. FRANCIS: I just wanted to ask you with respect to the way those data
sets were used. They were bringing together data sets that had not been
previously combined, whether there were questions raised or any kinds of issues
that we would then need to think about with respect to privacy, particularly on
the risk of reidentification of what might have been data sets that did not
reveal any sorts of individual identities?
DR. BILHEIMER: That is an excellent question. The data on the interim
webpage and that will be in the Warehouse are not micro data. They are
indicator data. They are aggregated indices for obesity rates, mortality rates,
poverty rates. We follow our disclosure requirements in terms of sample sizes,
so we do not publish data for small counties, or we combine data over
meaningful use years. So we are very conscious of the disclosure risk issues.
But no, there are no micro data involved, and we are very, very careful
about the cell sizes before we put anything out.
Please stop me at any time.
This is a screen shot of Microsoft Bing’s website. They have taken all of
the community health data, and they are using it in a variety of different
applications.
One, they are looking at quality of care in hospitals for using the
Medicare Compare data. This shows you a map of the Seattle area. What they did
here was to look at smoking rates in different parts of the Seattle area, and
then took the case of someone who was looking for an apartment who wanted to
work in Seattle and wanted to be in a smoke-free environment. They overlaid on
top of that where all the apartment complexes were and I think they even had
rents in there. So if you wanted this type of a place to live and you wanted to
be in a place where very few people smoked, this might be the part of the
region that you would want to live. Intriguing, it was very interesting to
watch.
The next is a screen shot of Google fusion tables. Google has taken all of
the community health data and put it into their fusion tables. So for any of
you who have used fusion, you can now with the device bring all sorts of
different data sets together and map them and chart them in different ways. The
community health data are now in there and can be used in that way.
They also are developing an application using the Compare data, the
Medicare Compare data, to identify high quality hospitals and to make
comparisons of hospitals in different markets. I think it is really interesting
to watch, because Google and Microsoft seem to be having a competition here
about who can use community health data most effectively. There have actually
been some blogs on the Web writing about who is better, Bing or Google, at this
point. So we are having fun watching that to see who is going to come out ahead
on that one.
This is a very different application. It is a web shot of Sonoma County’s
Network of Care for Healthy Communities. Sonoma County is the case study for a
project that is being developed jointly by the National Association of Counties
and the Healthy Communities Institute.
What this dashboard does is, it has both an individual component, so people
can look at their own individual health measures, but it also primarily has a
community dashboard. You can look at the indicators. As you can see down here,
it has little dashboard indicators showing you how you compare with other
California counties on this indicator, private insurance, and there is another
one with obesity, and a whole host of indicators.
It also provides you with links —
MR. REYNOLDS: Doctor, we have a question, please.
DR. BILHEIMER: Sorry.
DR. OVERHAGE: If I can put my pessimist’s hat on for a moment, this is a
neat visualization. How rapidly do data sets turn over? You have a display here
with a needle. Those are very cool fun visualizations when you have something
this dynamic. These data sets are relatively static, and perhaps a year or two
old, if you will.
A lot of the things we try and do, if you are looking in the rear view
mirror and you only get to glimpse it once every ten miles, you can’t drive
very well.
DR. BILHEIMER: And the point is well taken. It is a point that we have made
repeatedly. With the data that we have, we are constrained on what we can put
out.
I think how people are thinking about this is that these data are not going
to be able to be used for detailed day to day, week to week, month to month
tracking of the effects of interventions, policy changes and so on. They are
more a call to action type of indicator, which is how the county health
rankings have been used. Then, which is where we get into some of the other
data supplier issues, how does the community then say, take the information and
say, we have a real problem here with obesity, we would like to introduce these
interventions, how are we now going to track at the community level how
effective we have been. Then two years down the road or three years down the
road, there may be another signpost from here.
That is one of the issues that we have tried to get across to people. We
will do our best to get data out, but there are only so many things you can use
it for, and criteria well timed tracking of real time interventions is not
where action is for these. Obviously there are other data sources that enable
you to do that. I think what we are seeing is the interest in using both.
DR. OVERHAGE: The optimist side of the comment is that this kind of thing
exposes the data and its limitations in a way that may help people drive better
sources. So putting a light on the data is always a helpful thing in terms of
improving it.
DR. BILHEIMER: Yes, I think that is right. At the meeting on June 2, we got
enormous feedback on data that people would like to see. Hearing that in a
group meeting with a lot of people who are concerned about data policy going
forward is really helpful.
At the same time, we have tried to make it clear all along that we are not
dealing with micro data. Others are dealing with micro data, and that is fine,
and people have come to us and said, how can we get the micro data that
underlie your indicators. The answer is, you have to go to the agency and
follow their procedures for getting micro data.
DR. FERRER: Can I follow up? So right now your publicly available data is
actually constrained based on what Dr. Overhage just spoke to, correct?
DR. BILHEIMER: Yes.
DR. FERRER: So if you went one step further and tapped the academic medical
centers which are the suppliers of the data that Dr. Overhage — is there
anything going on there? This is an incredibly powerful NDDM proxy to get you
closer. It is clearly not close enough for research clinical medicine. So where
is that?
DR. BILHEIMER: That is not part of this at the moment, but could well be.
This is a very large tent. If you remember that chart about all the different
data sources, there are lots of things that we are not going to have in the
Warehouse, but other people are now popping up and saying, given this
ecological community that is developing of people who are developing data and
people who are using it, can we play, can we be a part, can we provide our
data, possibly in a different way, a different warehouse or through a different
facility.
The answer is, absolutely yes. Things have moved so rapidly in six months.
In fact, one of the applications that I will show you in just a minute is an
example of something that we never even envisioned, but a new way of crowd
sourcing data that is now being used in asthma attack prevention.
So yes, there are all sorts of opportunities. I would be happy to put you
in touch with people.
DR. FERRER: One of the things that oftentimes people don’t talk about is
that from a research competitive standpoint, from a academic medical center
view, you are able to give me the information, we will use it, we will touch it
as long as you are not at a competitive disadvantage, because clinical medicine
is still a competitive environment. So it is good as long as I can still be
number one. So that is why often that people don’t like to bring it up.
DR. BILHEIMER: The point about this is that we are putting the data out
there and saying anybody can use it. We are not going to control what people
use it for. Obviously we have to be very careful about what we put out
therefore.
One of the things that is very encouraging is that the big apps developers,
the Googles and the Microsofts, are extremely concerned about using it right.
They have called us and said, can we have technical consultations between you
and our staff to make sure that we are interpreting these data appropriately.
But once it is out there, we let it roll and see what happens.
One of the things that we are letting it roll, this is the webpage of Me
You Health. Me You Health develops games. They have developed a game called
Community Clash, which some people are saying one of these days will become the
community health farm bill. It is a competition to help people play with each
other to help them understand the health status of their community and what
their community can do to improve its health. Interesting to watch that one.
You can go online and try playing.
Finally, the last of the applications that I am showing is called
Asthmopolis. This had nothing to do with our data at all. This is entirely
separate and independent. What the folks at Asthmopolis have done is to put
little tiny GPS trackers on inhalers for asthma patients. What they are doing
is both enabling providers to track what is happening to their asthma patients,
but through crowd sourcing they are able to identify regions in parts of cities
where the air is bad on a particular day, because they are picking up a whole
lot of inhaler use in a particular part of a city on a particular day. Very,
very intriguing. It is only just started, but they were able to give a
demonstration of how this is beginning to work.
So this is a very wide open tent, and this is fine. This is part of the
Community Health Data Initiative, and it has nothing to do with our data
whatsoever. It is a totally independent data source. One of the things that we
were hearing at the meeting on June 2 was, crowd sourcing data was going to be
a mechanism that we were going to see grow.
MR. REYNOLDS: Doctor, we have got a question.
DR. BILHEIMER: Sorry.
MS. FRANCIS: I just wanted to put the privacy question in there again. On
GPS tracking, possibilities are quite impressive.
DR. BILHEIMER: I am not the government at all.
MS. FRANCIS: I understand that, but it could be combined with government
data.
DR. BILHEIMER: It could.
MS. FRANCIS: And it is also in your slide deck. So I am just interested in
whether they did demonstrate their privacy policies and consents along with —
DR. BILHEIMER: I have not seen their privacy policies and consents. I can
check on their website and see if it is there.
MS. FRANCIS: Well, I’ll help you to do that. Thank you.
DR. BILHEIMER: For CHDI, the next steps for the initiative are what is
called the Health 2.0 developer challenge. Health 2.0 is a company that is
based on San Francisco. They have thrown out a challenge for the development
both of the best ideas for applications for the use of community health data
and also the organization of what they originally called hackathons, but that
got everybody a bit nervous, so they changed the word to code-a-thons, which is
something that the Sunlight Foundation has used extensively to bring together
programmers and web programmers to develop solutions to problems in a short
space of time in an intense environment.
Health 2.0 is organizing code-a-thons and team competitions for ideas and
also for applications development around the country between now and the first
week of October, when the Health 2.0 conference will take place in San
Francisco when the final challenge will take place. That is the next set of
events, but there is a whole lot going on.
Once you throw something out there like this, there are mushrooms growing
everywhere, and it is difficult to keep pace with everything that is happening.
But back on the farm we have a Health Indicators Warehouse to produce. It is
not only to serve the needs of the Community Health Data Initiative. Our
primary purpose originally and still primarily is, is to serve the needs of our
most critical clients, which is Healthy People, CHSI, the county health
rankings and so on.
This is the functionality that we envision for the Warehouse. I emphasize
at the top, there will be no individual level data. These are pre-constructed
national, state and local level indicators. We will have a very user friendly
interface. We will have linkages to evidence based interventions. We will have
web service capability, which a lot of the apps developers are extremely
interested in having. We will also — and this is really critical — have
supporting meta-data to make sure that people understand how to use the
indicators. Actually, constructing the meta-data is about as complex as
developing the indicators themselves, possibly more so.
For us, the next steps first of all is to have version 1.0 running in
December. We get calls, e-mails, around the clock saying couldn’t you just add
this, why aren’t you adding that, have you thought about this. We are having to
be quite hard-nosed about it, saying great idea, and we will certainly consider
it for version 1.1, 1.2 or 1.3. But right now we have to have an operational
website by December.
We are also working on securing the funding to maintain it. We have pieced
this together with contributions from multiple federal agencies, and we are
trying to insure a funding stream going forward. We have to think about
harmonizing indicators to the extent that seems reasonable. In some cases there
will be more than one indicator of a concept in the Warehouse, because agencies
really do have different needs for a different measure.
We have to develop a process for selecting new indicators, and working on
updating and improving. We are in fact having some focus groups this summer to
show people where we are right now and get some of their ideas, particularly
for the web service capabilities about things that they would like to see in
that. But we are very heavily focused on having a product by December, which is
what we are committed to do, and which we have to do for Healthy People.
Healthy People has to have this running by December. Then in 2011 we can start
playing around with some of the bells and whistles.
DR. OVERHAGE: How does this relate to or connect to the CDC’s Wonder
project?
DR. BILHEIMER: It is separate from Wonder. Some of the data that have been
in Wonder, the Healthy People data, will be in the Warehouse. There will be
other data from Wonder that will end up in the Warehouse, but right now we only
have Healthy People in there at the moment.
Please contact us if you need any more information about the Warehouse. If
you want to see more information about the Community Health Data Initiative,
you can go to the HHS open government site and like I say you can download the
videos there from the meeting on June 2nd.
MR. REYNOLDS: Let’s open it to questions.
DR. GREEN: Where is NIH in this?
DR. BILHEIMER: Where is NIH? They are interested in this and would like to
participate more in this. They also would like to help fund this, some of the
Institutes.
We need to have the Warehouse vetted through the HHS IO Governance Board,
which means going through enterprise life cycle processes and having it
certified within CDC, before NIH can participate. For IT projects, they want to
have the HHS IO Governance blessing going forward. That is the other thing that
we are very heavily involved in right now, is reaching the IO Governance
standards for the Warehouse.
DR. GREEN: Can you be more specific about who they are at NIH?
DR. BILHEIMER: There are folks in the National Cancer Institute who are
interested. I am trying to think, maybe in NIMH, but I think primarily the
National Cancer Institute.
DR. GREEN: I would flag as obvious the immediate need that NIH has with
this capacity for the CTSA boards, which are going to go to 60 sites
extramurally all across the country, particularly the community engagement
portions. All of them are required — every awardee has to have a community
engagement component. Several of us at this table are involved in this NIH
initiatives.
This county level data is essential to the community engagement component
in the NIH-CTSA arrangement. The National Center for Research Resources strikes
me as being an immediate need of the capacity and likely to be interested and
likely to be responsive. It would be nice if the CTSA awardees — there are 48
of them now — if this presentation could go to the directors of those 48 CTSAs
for a bit of interest that would come from all across the country from the
research community.
DR. BILHEIMER: Thank you.
DR. HORNBROOK: Just a postscript. Barbara Alving is the Director of NCRR.
DR. BILHEIMER: Sorry?
DR. HORNBROOK: Barbara Alving, A-l-v-i-n-g. The PIs have regular meetings
in Bethesda, and they regularly have a program which you could present this to,
and suggest that all their community programs support this effort.
DR. BILHEIMER: Thank you
MR. LAND: In response to Marc’s comment about the timeliness of data, I
reported yesterday in the Population Committee about some new things happening
in biostatistics. One of those is a plan to develop indicators that will be
published by the National Center, certainly on a semiannual basis and even on a
quarterly basis.
I don’t know if they are going to be at a county level or not. They are
still in the process of defining what those indicators will be, but this will
be the first time that vital statistics data has been available even less than
two years after the event. So it may ultimately go into this data Warehouse.
This is the first time I have seen this. This hasn’t been discussed with my
association but there is that opportunity.
DR. CARR: Thank you. This is a very, very exciting initiative and a great
presentation. This is a topic that has run through a number of our discussions
along the way.
One piece of this is, how do we develop expertise in this type of thinking
and use of data? Where do you see the centers of development for this
expertise, for this workforce who knows how to do this?
DR. BILHEIMER: It is a question that I have been thinking about, too. I
think almost without saying, I don’t think this would have been possible
without Todd Park.
DR. CARR: Without — ?
DR. BILHEIMER: Todd Park. Todd Park came to HHS with a very different
vision and understanding of data and how to use it, and also came from a Web
2.0 environment. He is educating us on a day to day basis on things that we
should be thinking about and how to think about it.
But an awful lot of our learning, and we are learning by the seat of our
pants, this is not a traditional HHS-NCHS type of activity, is through
interactions with the Googles, the Microsofts, the Red Hats, O’Reilly Media. We
had conversations with Tim O’Reilly on his treadmill, discussing indicator
data. So it is a learning by doing.
But I think we do need to be thinking about how we train people in this
dual understanding of the importance of high quality data, the protection of
confidentiality, and also the ability to get data out there so that people can
use it in these creative ways.
One of the things that I am hoping is that over time, we can have for
example fellows and interns working with us who come in with the web background
but not necessarily with the data background, whom we can train and work with.
DR. CARR: Thank you.
DR. HORNBROOK: Two questions. Are you aware of the SEER registry and the
interest of getting the SEER data into this? Is that what they are thinking
about?
DR. BILHEIMER: Right now our discussions have — they are very interested
in county level data. Our discussions with them focused more on county level
data than specifically on the Warehouse.
No, we have not talked about getting SEER in here. It is something that
could be over time, but again it depends. We are not talking about putting
micro data in there again. It would be indicators coming from SEER if that was
to happen.
DR. HORNBROOK: No, I understand, but the issue of giving communities the
ability to track cancer incidence rates, cancer death rates by county, to track
environmental carcinogens and the effect of the kind of environment you live
in, if you live in an environment with heavy radon do you see more lung cancer,
kind of thing.
There is some evidence now showing that an epidemic of head and neck cancer
is starting to emerge among teenagers, related to HPV, oral sex. So it raises a
whole issue of how communities want to deal with the issue of HPV vaccinations
and things, which gets into health behaviors.
The other question is, is the VA involved at all?
DR. BILHEIMER: The VA is not involved, no.
DR. HORNBROOK: Because I can visualize the notion of where are the vets of
the various wars, where do they live, what is their health? Obviously the VA
has had a tremendous amount of health data about veterans that use the VA
facilities, and they can put together county level health indicators for the
veterans that they are treating in that particular county, either that or by
the veteran’s address, if the wanted to do it. I would think that in today’s
world, with the Afghanistan and Iraq stuff going on, you would want to have
some degree of indication of the regional variation of the burdens of the war
on our community.
DR. BILHEIMER: I think that makes sense. The agencies that have been
primarily involved are those that we have been serving all along, because their
indicators are going to be in there. So it has been primarily HRSA and ODHP for
Healthy People.
But yes, there are other agencies involved. The Office of Minority Health
is involved, the Office of Women’s Health is involved, and SAMHSA is involved.
They are the ones who have mostly been involved at this point. But these are
very early days yet. We have to keep reminding ourselves of that.
MR. LAND: One of the cautionary things that I would think about this is
that when government publishes data and overlays the different indicators, we
are always cautious about association versus causation and those issues. When
you have people who are whiz-bang at software, they may not be trained as an
epidemiologist, and they may not be as cautious as what we would be. Certainly
you find that with newspaper reporters and the like oftentimes.
So I just wonder, have you thought through, how can you reintroduce those
cautionary notes as people start overlaying things that may look like
causation, when we would know that they probably aren’t? How do we guard
against those types of things?
DR. BILHEIMER: There is only so much that one can do. Either you have a
free market out there and you put data out there, or you don’t.
We are putting in every caution that we can. We put in as many cautions as
we can in the meta-data. For example, we want to have confidence intervals
around estimates. When you start getting in county estimates, confidence
intervals become very important.
We talk with these guys and they want to talk to us, because they are
concerned about getting the data wrong, too. Those are some of the big guys.
That is not to say that somebody somewhere isn’t going to take the data and use
it for something that would be totally inappropriate. But that can happen
anyway when you put public use files out there. That is just the nature of the
beast. But we do our best, and we try and work with as many people as possible
to try and get them to understand what the data can and can’t be used for. But
sooner or later somebody probably will, I expect.
PARTICIPANT: Just a brief comment and then a question. First off, excellent
work. I appreciate what you are doing. These days I am involved in community
activities, and we are constantly looking for data to track to see if any of
our interventions are working. Obesity is a good example.
Admittedly, I have not followed this issue nationally, but I can tell you
that historically, 20 years ago this committee was talking about linkages from
data set to data set. I assume from what you have said that not much has
happened. We argued over unique identifiers. Some argued for use of the social
security number. But it would be great if somehow we could link indicator sets
from one to another on an individual basis, recognizing that out of privacy
concerns you would then have to aggregate the data. But it sure would be
helpful, and it would deal with some of this timely issue that we are facing
with data that is two or three years old. Just a comment.
MR. REYNOLDS: Doctor, thank you very much. Excellent presentation, very
thought provoking at the right time for us as we just finished a paper on the
next wave of health information and usages of it. Perfect timing. Couldn’t have
been a better time to evoke emotions and enthusiasm and concern. Thank you very
much, really appreciate your time.
DR. BILHEIMER: Thank you.
Agenda Item: Subcommittee Report Outs
MR. REYNOLDS: Let’s move next to report-outs from the subcommittees. I just
listed Quality, Populations, Privacy and Standards. If anybody isn’t ready yet
and wants to go last, otherwise I would like to start with Quality. Since
Quality is run by the incoming chair, that would be a smart thing for me to do,
probably. Would you like to move to later in the program? That is not a
problem.
DR. CARR: Give me a minute, because I was counting on my buddy Matt.
MR. REYNOLDS: Okay. Privacy has been willing to jump forward.
DR. STEINWACHS: Yes, we are always prepared. We held hearings on Tuesday
regarding sensitive information, specifically to try to further our
understanding of what are different sensitive information categories, and
within those categories what information really needs to be considered
sensitive.
I will give you an example of that with respect to mental health
information. Everybody likes to paint mental health information with a broad
brush, are certain meds and the like part of mental health information, when
the reality is that medication may be used for a whole variety of non-mental
health related purposes. So we wanted to try to understand specifically what is
sensitive in the mental health realm. That is just an example.
We heard testimony from mental health individuals, people that are involved
in genetics, child and adolescent, reproductive.
MS. FRANCIS: Domestic violence, and the kinds of information that may not
seem so sensitive but could be really scary if it fell into the wrong hands,
such as physical location.
MR. HOUSTON: Walter calls it contextual sensitivity. I think I call it
situational sensitivity.
But nonetheless, the hearings went very well. As a result of the hearings,
we have been working to put together a draft letter that we really would like
to get out to the committee for approval in September, because this is an area
of immediate interest by the people that are doing the — and in development
and the policy committee and the like.
We would like to try to get this out. Towards that end, our hope is to try
to focus on trying to define these categories in order for individual HIE’s and
the like to take that information back, apply it to their own circumstance,
look at their own state laws. So we are not trying to overlay the whole idea of
what are the other regulatory considerations, but we are trying to categorize
sensitive information so that people can take it back and try to apply it in
their own organizations.
So we have already started working on the letter. Leslie did some great
work on it yesterday, the subcommittee as well as some individuals probably
worked for four or five hours on it yesterday. We are going to do some
conference calls and circulate the letter internally, hopefully getting
something to the committee for its review prior to September, maybe in August,
so that we can try to refine it and bring it to closure in September.
MR. REYNOLDS: Questions for the privacy group?
MS. FRANCIS: I might also add that we are looking forward to having
hearings on governance in September probably.
MS. GREENBERG: I might just ask about that, because I noticed that at your
meeting yesterday. Are you interested in having a hearing in conjunction with
the full committee meeting?
MS. FRANCIS: That would be ideal, but if we don’t have priority for that,
we would work with you to schedule it at a freestanding time.
MR. HOUSTON: I am a member of the tiger team that has been put together for
privacy and security. The tiger team has talked a little bit about wanting to
start to address governance in the August time frame. There have been other
discussions about other groups looking into governance. We certainly don’t want
to step on toes.
Leslie and I had a chance to meet with Devon McGraw to discuss how do we
coordinate activities between the different groups and make sure that we aren’t
overlapping and that we are applying all of our resources that are our highest
and best use. So we are going to make sure we are working very closely with
Devon to make sure that we aren’t doing something that just doesn’t make sense.
MS. FRANCIS: We also have talked specifically with the staff person at ONC
whose responsibility is governance, and they are very interested in hearing
what we have to say.
MR. REYNOLDS: So John, having been involved in that subcommittee for a long
time and a lot of the body of work that we have done, somebody kidded me last
night about the train left the station and it is going really fast. Now the bus
drivers are Google and Microsoft. I am just making a point.
Since you are working on the other stuff too, is it really converging fast
enough that with everything that is going out there, with all the electronics
and everything, that there truly is pragmatic privacy and security? Pragmatic
is the key word that I am using, because nobody has come up with practical
incent yet, nobody has come up with any of these other things.
So where is it? Give me a kind of a state of the union.
MR. HOUSTON: I’ll give you my version of it and maybe Leslie will then give
her perspectives as well.
I think the reason why this tiger team has been established is because of
the fact that there are concerns that there isn’t enough movement on some of
the underlying privacy and security needs. I have been pushing governance for a
long time, because at the end of the day you really do need to have a body or
bodies out there that really do provide the oversight and the enforcement that
is necessary to insure that adequate public trust exists.
There are some other issues that need to be addressed. I am going to let
Leslie speak about this issue of non-covered entities, but I have personal
experience dealing with a situation like that today, and I think that these
HIEs are going to become — and NHIN hopefully becomes an authoritative rich
source of information. Yet absent good privacy and security in dealing with
issues like non-covered entities which haven’t been dealt with yet, I think we
have got some real issues.
So again, the tiger team is intended to try to help further that, but I
think we do have a lot of work in this area.
MS. FRANCIS: Governance itself is an interesting term. When I was talking
with Justine about her interests and so on, I mentioned governance, and she
said, wouldn’t it be interesting to know who all the players are.
Actually, if you think about governance, it can be everything from making
sure that there is a designated chief privacy officer somewhere who keeps the
organization on the appropriate path, all the way to very strong governmentally
related regulatory. There has been some discussion about soft law versus hard
law and so on in that space.
I made a little preliminary list, John has added some already, that I think
I sent you, Justine, and that I sent around to the privacy folks. It is
amazing, the number of people who potentially have a regulatory interest, all
the way from organizational privacy officers to private sector certification,
C-CHIP, trustee and so on, to the panoply of state and federal agencies. So it
may be that what we contribute, at the very high level look at who all the
actors are, but that is the space we are thinking about, yes.
DR. HORNBROOK: Just a couple of questions. To what extent should we the
committee be worried about helping consumers understand where their privacy
risks are? I am thinking about, when people are keying in their personal health
information for their doctor or for their prescription or whatever, they may be
working from home on a wireless network or on a wireless appliance, where and
when are they at risk for information going out to somebody who is listening in
on them?
Then what about the ISP provider? How much of the data that they transmit
through do they actually keep? What are the controls that those providers have
under any data they keep?
MR. HOUSTON: That is a great question. I think the reality here is that
especially in the PHR space, there aren’t requirements per se. There are
standards, there are best practices, hopefully there are fair information
practices that are conveyed to the consumers as to what the risks and
protections are within each PHR. But in large measure it really is up to the
PHR to decide how far it is going to go.
If you look around the industry, there are a lot of tethered PHRs, PHRs
that are aligned with providers or provided by providers. I think you see some
very big players in the industry that are establishing PHRs that are collecting
data from providers, who have made it extremely clear that they are not HIPAA
associates, they are not HIPAA covered entities, and the consumer needs to rely
fully — the consumer has to be able to read their information practices or
policies and make a value judgment as to whether they will accept them.
But I think what we see in the industry is that there is far too much
click-through. They want to get to the site. There are these ten pages or five
pages of small type and they just whip through them real fast and say I Accept.
The reality is, they don’t.
So I think that is a huge issue, one that we have brought up in the past.
There has to be a better more comprehensive, more appropriate way to put
consumers on actual informed notice of how their data is going to be used. That
also includes some of the things you said about the technical, what does the
ISP do, what is the security that is in place. But it really does run the
gamut.
By the way, some of these sites are designed to be open, so there is this
notion in certain sites that you don’t have an expectation of privacy. You are
subscribing so that you can put your record out there. That may be perfectly
acceptable. I am a big fan of this concept that as long as the consumer makes
an informed decision, that informed decision may be that they are willing to
agree to receive advertising, because there is this great site that gives them
all sorts of information and allows them to do all sorts of neat things for
free.
So as long as the consumer is allowed to agree and understands what they
are agreeing to, I think you have to let these sites develop in a way that they
think they want to.
MS. FRANCIS: You may be aware that Google Street recently was found to have
been collecting information that was transmitted over unsecured home wireless
networks, when it was going up and down streets. Google didn’t know it was
doing it. It stored that, so people communicating with their financial
institutions, with their health care providers or whatever over their unsecured
home networks.
Google didn’t know it. This was a rogue programmer who had put that in the
program. I would hope that a starting place for us — because it is all over
the map — would be the governance question. If Google doesn’t have somebody
making sure, or some way to make sure, that Google Street isn’t out there
picking up everything that gets sent to you, that is a huge problem.
DR. SUAREZ: I think the other thing that we found during the hearings that
in my mind was very valuable was, number one, the issue is not just about
protection of the data while in transit, as in exchanges of health information
systems or health information exchanges, but truly the expectation that the
tools that we are going to be using in the industry, electronic health record
systems, are going to have the functional capability to segregate and allow the
user to handle appropriately the data.
So there is a need to see and to pursue that type of expectation, but the
expectation is both at the EHR level as well — the functional capability
expectation will be both at the EHR level as well as the HIE level. But there
are important differences there.
The other interesting part was that this is not just about consumers being
able to control the data and decide whether data can be disclosed to another
doctor for X purpose for payment purposes and operation purposes. This is not
just about that. This is also, we found new dimensions to this, not new, but
new in the sense of the importance of this, it is about also protecting the
data or allowing the protection of the data from parents, to be seeing the data
of adolescents, for example. So it is very different. Or in victims of domestic
violence, allowing relatives and others to see the data.
So it is not just about the traditional way of thinking of disclosures and
patients controlling the disclosure to peers or for other purposes, but it is a
much more encompassing aspect of it.
DR. OVERHAGE: I’m sorry Leslie stepped out of the room. All the points she
made about the Google issue are valid, of course. The wireless networks that
were open doesn’t mean the data was open. Much of those transactions were
secured in other ways over that open wire.
The other thing though that becomes more important is the point you were
making just a moment ago that I wrestle with. It is much like the mortgage
crisis and things like that. How do we create that balance between personal
responsibility? It is a very tough one, and it is a particularly tough one in
the health care environment, where there are so many competing value
propositions, time pressures, complexities and so on.
But I think it is important that we always keep that personal
accountability issue on the table as we have these discussions. I agree
completely. I do it all the time. I read through the quote-unquote eight-page
click-through, and it is very difficult. But I think we ought to keep that
balance.
MR. HOUSTON: I think that is the key. I’m not sure how we make it really
and truly informed, but I think users do need to understand what they are
signing up for. It is so hard, whether it be in a PHR context or even in a
treatment context. Patients, what their expectations are, vary, and very often
they situationally vary. So what is acceptable in one situation isn’t in
another and vice versa. Opposite situations come about because of an emerging
circumstance, so your expectations will change. They don’t want their
information accessible, oh, but only in this context, which they never would
have expressed at the time they provided their previous consent. So it is a
difficult question.
MR. BLAIR: As I think most of you are aware, we have been building our
health information exchange network and making rapid progress, not as far along
as Indiana and a few others.
As we have done so, we have been very, very concerned about privacy and
security. We have hired two certified security professionals to protect us on
that piece. We have hired a privacy officer who has deep experience. We have
been involved with a HSPC project, a DRSA project, we passed state laws. We
have really taken privacy as a life or death issue for the existence of our
health information exchange network. We have worked with the largest health
care facilities in our state that have had very demanding privacy requirements.
So that is my preface. Privacy and security we view as life or death of the
existence of our organization.
One of the things that I would like members of the Privacy Subcommittee to
be aware of, I think many of you are, is that in our state, and we are not the
only state, there are special protections for behavioral health and substance
abuse. The thing about it is, there is no way that we have been able to find —
by the way, we have also called in people from Sandia to find out the security
processes that they use for national defense. We have not been able to find a
way to sequester or mask patient records. We don’t see that it is in the near
future to do so.
So the way we are handling that, because we do have some state laws that
give special protections for certain conditions, we have to require, and we
have done so, written patient consent for access to information over the health
information exchange. We have had these discussions with ONC, because again, we
are not the only state that has had this. We are active on the NHIN. We have
right now, after having the Indian trial implementation projects, we were
active in that, we also have the SSA contract, and that pilot has also begun
right now, so we are pretty much involved in that.
The bottom line is, we can’t share information outside the state unless
somebody provides a written patient consent. With social security they send it
electronically as an image, so we can do it with social security. Within the
state we can go ahead and we can require written patient consent.
So my observation, because I listened to the Privacy Subcommittee, there
were a couple of attempts during that meeting to say, let’s recognize that
there needs to be a balance. That work was rejected. I thought that was really
unfortunate.
So I echo Marc Overhage’s suggestion that balance is needed. I think
someone else also mentioned balance. We are going down the path, but the bottom
line is that whatever the most stringent constraints on privacy is the one we
are following for everything, we can’t parse it out.
So I just wanted to express that. Conceptually, sequestering and masking
and dividing things up might be a good intellectual idea, but please check and
verify that it is technically possible.
MR. HOUSTON: First of all, let me be very clear what we are trying to
accomplish in this letter in September. We are not advocating sequestration.
What we are trying to do is put in place an understanding of what types of
information are sensitive and specifically in what context, and whether there
are subsets of information within that broad category that really are sensitive
and those that are not.
Really, the hope in my mind is to provide guidance to the HIE and others so
that they can then take that information, look at their own state laws or other
applicable laws and their own architecture and decide if and how the want to
try to implement it.
So if an HIE says everything is going to have the same sensitivity to it,
that is great. But I suspect what we are going to find is that as we develop
the NHIN and develop the ability to pass information between HIEs and between
states, that there needs to be an understanding of whether there are certain
data elements that are more sensitive. Maybe there will need to be some type of
sequestration, but that is not what we are advocating. We want to simply make
sure that there is some understanding of data sensitivity as a utility for
people then to use as they see as appropriate.
MR. BLAIR: I appreciate that. I hope that the report comes out with the
caveat that you have just mentioned. The idea of the exercise you are going
through has validity, but I think there should be some recognition in there
that you understand that the possible consequence of that, which may be okay,
is that at least for the near future, many health care facilities simply will
have to go to the highest standard.
That may be fine. That is what we are doing, we are not complaining about
it. We are not complaining about it. But the problem is that it means that
people outside of our state will not have access to our patient data. So it is
a national issue.
DR. MILLETT: Jeff, I am with you. In fact, in our subcommittee we discussed
the difficulty in discussing a lot of the legal principles and laws and
regulations to put in place the requirements for sensitive information. What we
are trying to do is help to make them crisper, what is sensitive, how is it
defined. We talked about the need for further clarity and assistance for the
vendor community, because there is a huge disconnect between the complicated
privacy conversations and what is available to people who are implementing, and
the need to get there quicker.
MR. REYNOLDS: John will make a couple of quick comments, and then we will
move on to Populations. As I travel around, the definition of deidentified is
still incredibly misunderstood in the world, anybody you talk to. I guess the
committee could look back at some of our secondary use stuff, some of our other
stuff, working with the Department to help people continue to get that is key.
Then I think another thing that is going on dramatically in the health
industry is outsourcing. It is becoming a multiplier of ten of whatever it was
before. So a company may have a privacy official, but as people more and more
outsource functionality.
MR. HOUSTON: ASPs are exploding.
MR. REYNOLDS: Not just ASPs. I am talking moving them.
MR. HOUSTON: ASPs are probably the biggest concern I have.
MR. REYNOLDS: So again, I think those are two subjects that could truly use
a little bit of reminders. Deidentified, if you talk to ten different people,
it still is ten different definitions. But they use the word to say I can do
what I am going to do. So just take that as a comment.
Okay, Populations.
DR. STEINWACHS: We spent a good part of our time in Populations fixing the
slides. We do report that the slides are fixed, I hope. We will get to assess
that later.
One of the things that we did, Garland gave us an update on vital
statistics and where the states are and NCHS on that. He said a few things. I
thought it would be worth if you should share with the whole committee, because
it was very good news about the progress that has been made and what looks like
a very rosy future at this point, compared to where we were talking about a
year ago.
MR. LAND: I had reported a year ago that there are some troubling things
going on. The National Center was planning to only purchase what was called
core data, which was a very small set of data that had been available
nationally and previously. The new contract for the states was going to reduce
available resources by about 20 percent. So there was real concern for what the
future was for the national vital statistics system.
In the past year, there has been a series of meetings with representatives
of my association and National Center, and major progress has been made, to the
point that the previous plans, that contract isn’t going to be developed. The
standard 2003 data items will be collected. A process has been put in place to
assess the states that are having problems.
Typically what has happened in the past is, a single state holds up the
whole national data set. That is why there has been tremendous delays in the
Center being able to publish their data. So we have developed a process where
the National Center and our association will be providing technical support to
the states to assist them in getting through whatever problems they might have.
In the President’s budget, it was mentioned yesterday, there was money for
the Center to provide funding to states who don’t have electronic birth and
death systems. So that would complete the few remaining states that don’t have
the current data set. That is the first time the National Center has had those
kind of resources to provide to the states.
So a new contract is being developed right now for implementation in
January 2011. We are very satisfied that we are taking a new look at how to
improve the system for both the timeliness and the quality of the data. What I
mentioned earlier today about releasing data on a quarterly basis is one other
indicator of something we are promoting and supporting.
So, some good news.
DR. STEINWACHS: I don’t know if there are any questions for Garland?
Otherwise I will go on.
We spent our time on discussion of the future agenda. In the slides we have
a very high level view. We were struggling a little bit with how to get down to
what the priority specifics are. We had agreed that we would have a conference
call in early July, maybe another one in August, with the strategy in mind that
we might try to come up with a list of specifics that we see as attractive next
steps, and maybe a few more than what we think we could deal with to bring into
the Executive Subcommittee retreat in September, and look for synergism between
what we are doing and what other committees are doing.
Part of that is the recognition that every member of the Populations
Subcommittee is also on some other committee. Two, most of the things that we
started talking about seemed to be cross-cutting issues. It made sense to try
and think about how to design that next step agenda in that way.
There is certainly a lot of interest in element three. I think from now on
we should just refer to it as element three. This is to protect the content of
what we are doing, right? In part, not only thinking about learning health care
systems, but thinking about learning public health systems, which ties very
much to one of the things in our slides about the IT infrastructure for public
health, as well as thinking about the utilities and applications that can
extend it.
The indicator warehouse discussion that Linda just gave us, sounds like
exciting frontiers. We are beginning to talk about utilities and access to data
for public health purposes. That may be one of the things that we may want to
put on that agenda.
Let me just open it up to any other committee members that might want to
add something. I have exhausted my notes at this point.
MR. REYNOLDS: I really want to thank you guys for your leadership for the
whole event this afternoon. Everybody playing off of that document and being
able to use it as a stepping off point for the future. I thought you guys did a
wonderful job pulling that together, and having gone through it and presenting
it. I think it really is going to set a stage for populations not being a word,
but being something that turns into real activity in many places.
DR. CARR: I just wanted to add that I want to thank Bill for his role as
co-chair of that subcommittee. Bill is stepping down and Larry Green is
stepping up to be co-chair of Populations. So thank you to you for taking that
role.
DR. STEINWACHS: You notice how I deftly as the old co-chair was able to get
the young co-chair to take the lead on that Friday afternoon set of calls. We
found the person on the subcommittee with the passion.
MR. REYNOLDS: That is a good thing to do.
DR. GREEN: I just thank Bill and Susan. When you say you guys did a good
job pulling that together, I think we should be a little more precise.
MR. REYNOLDS: If you had listened to me yesterday, you guys always includes
the staff when I say something, very precisely.
DR. GREEN: At our committee meeting, I think it was Don that assigned to
Susan the attribute of extraordinary patience. I personally believe that to be
a serious understatement.
MR. REYNOLDS: Judy.
DR. WARREN: Standards also spent some of their time refining slides for
this afternoon. It was interesting, in that both Jeff and I started the slide
set, and then Walter stepped in when he became co-chair. We thought we had it
pretty well refined. But then as always, the committee works incredibly well
asking us what we meant by certain words. We realized we had the wrong ones, so
we edited. So we think we are going to be okay.
But then Justine commented that we had too many slides. So I said, we could
come in under the number, but nobody would be able to read the font. So we did
promise we would stay within our 15 minutes, so we are ready for this
afternoon.
We spent the majority of our time talking about our hearings in July. As
most of you know — and we found out it is called ACA 2010, not PACA, the
Affordability Care Act. In that act, NCVHS is mentioned about five different
times on various tasks that we must do and the due dates by which that work
must be finished. All of those things have to do with standards.
So our subcommittee is taking the lead in working on that. As Walter and I
read through the actual legislation and then worked with Lorraine Doo, our
staff person, we really felt that there was a lot of cross-cutting issues in
that. So I had a chat with Marjorie to see whether or not we could invite at
least one person from each of the other three subcommittees to be part of those
hearings. I had mentioned that yesterday. So I am hoping that by the end of
today, so I am hoping that by the end of today we have someone from
Populations, someone from Privacy, and after hearing the earlier dialogue, I
think privacy is going to be really important, and then someone from Quality.
Within our own committee, we have a lot of people that are serving on dual
committees. So Walter and I would appreciate having the final set of names. We
are putting together a set of briefing papers to let everybody know what is
happening on the committee, ready for the hearings.
Justine has asked that we also send that out to the full committee. We are
on a fast time line. We have to have a letter to the Secretary by September. So
it may require some phone calls, at least the subcommittee and the Executive
Subcommittee, so that we can get that letter to you in September and get it
approved.
The things that we are going to be having hearings about are the
development of a unique health plan identifier. This was originally talked
about in HIPAA, it was once more mentioned in the ACA. Given the deadline of
that, that is going to be the first thing.
The next thing is identifying operating rules. These are the business
rules. We are still trying to get a clear hand on the definition of operating
rules. We have to define operating rules for, Walter, eligibility and what is
the other one?
DR. SUAREZ: Claims attachment.
DR. WARREN: Claims attachment. These two standards have the most work done
on them, so it is fortunate that they are the first two that we have to have
identified about, what are the operating rules for these and what will be the
criteria for deeming the status of the organization that wants to write these
rules and publish them. So it is a two-part thing.
There will be more operating rules that we have to do later on, but right
now it is the two. There will also be a second letter in September about the
operating rules.
We also had a very brief discussion about our own strategic planning. I
think we got down to a laundry list and ran out of time, so there will be more
work on that as we progress.
Walter, do you have any comments?
DR. SUAREZ: No, I don’t.
MR. REYNOLDS: I know there were a lot of friendly wagers on the World Cup.
I am putting one down that unless they both stand up there and talk together,
they won’t get their 15 minutes.
DR. WARREN: We will.
MR. REYNOLDS: No, just kidding. It is going to be an interesting time with
the new bill that is out from the standards standpoint. Having been involved in
it for so many years on the committee, I would recommend that, as we mentioned
yesterday, those of you that think you are not part of Standards, it now
reaches out in new and different ways, and it is doing things. So if you decide
whether or not you think you are a standards person, everybody is going to be a
standards person, because it is going to define an awful lot of things that are
going on. So it is a great way to learn, a great way to help, and it is much
needed, with as far reaching as some of these standards are going to be.
So it might not feel comfortable to you, but if you let it happen without
all this representation, it will be very uncomfortable with you, and figure out
how you play that.
Before I move on to Quality, I would like to note that Dr. Simon Cohn, our
most immediate past chair, has joined us. He was a mentor to a lot of us.
Welcome, Simon.
DR. CARR: We had a great discussion, great participation. I will note that
there is substantial overlap between the Populations membership and the Quality
membership. One of the things that we talked about is whether we work together
with a single agenda, being cognizant of resource, time, bandwidth, and also
the fact that we really do all need to understand what is going on in Standards
and actively play a role there.
The conversation that we had yesterday I think put us in a great direction,
and I can’t thank Matt enough for putting this together. I will say that we
worked on our slides too, and I am happy to say that I think they are done now.
With regard to our committee, one of the recommendations for the meaningful
measurement hearing was related to identifying the core set of data to support
quality measurement activities, a critical factor in supporting both quality
measurement and 2013 meaningful use, and broader quality measurement capacity
from EHR, is to assess the components of this core data set in relation to
measurement priorities.
A key way to understand approaches and the impact of the recommendations to
establish and require support of the core data set in EHR is to gain feedback
from integrated and distributed health care organizations and groups that have
taken this approach.
What that means is, we sat together. On one side of the room was Mark
Hornbrook from Kaiser and then Marc Overhage from Indiana, both very mature
programs, both very different. I think they represent an array of elite
programs that have done all of the infrastructure on EHR and have learned a
lot, and have a lot to say here.
One of the thoughts was to hold a hearing for these elite mature programs
and ask them, now that you have this all in place, what got better and why did
it get better, and what did you need to learn or do over along the way to get
to where you are.
So I look to subcommittee members to elaborate on that in any —
MR. SCANLON: I think one of the aspects of looking at both organizations
was the idea of comprehensiveness of the data that they have. Part of the
recommendation in our December letter about a core data set was to build on the
idea that you do need a range of elements in order to be able to construct
quality measures reliably and validly.
The current paradigm, which is, every EHR is going to be able to produce
quality measures, is potentially significantly flawed, because if an EHR and a
single provider is missing key information that could be used or would have
relevance for that measure, then we are getting distorted measures. So we have
to think about how do we overcome that. I’m sure both Marks have examples of
how you miss the point if you limit yourself to a narrow set of information
that just may happen to be available, not that the other information was ever
there, because that is not the nature of EHRs. EHRs were not meant to be
comprehensive assessments of your lifetime health. They are the record of
encounters with the health care system.
DR. CARR: I have to say, now that I have just redone our whole
presentation, it all does tie together. If you remember, in 2004, sa we talked
about yesterday we had 23 candidate recommendations, and a lot has happened, so
I put that into the presentation. But that was all about the building blocks
that we need. What we are doing today is talking about how we use those
building blocks, how we put them together in measures and how we improve health
care. So I see a great storyline finally out of this. We are now taking it and
putting it together, and what can we learn.
Then the key questions, and even tying into the presentation we heard
today, this whole issue of data integrity, what is there, not integrity in
terms of honesty, but the completeness of it, when you have fractured data sets
and you don’t know it. So I think that goes a little bit with the community
data set thoughts that we heard about today. Yet we know that there is great
power in that.
So as we think about these mature programs and think about for everybody
who is now investing so much in their EHR, where do they start to begin to
harvest some of the benefits of this and where do they need to build. So I
guess we will be having a call as well; maybe we will all be having the same
call together to work on this.
Walter?
DR. SUAREZ: I just wanted to point out, there is also work being done by
the Health IT Standards Committee, the Clinical Quality Work Group. In fact, we
are meeting tomorrow afternoon to review the results of a preliminary informal
survey done with a number of health care systems on quality and performance
measures that could be a road map into meaningful use for 2013.
There is a lot of interest now in identifying quality measures that have
some impact from and repercussions on health IT, so that meaning that they are
significantly able to be leveraged through health IT and that they also are
able to show or document how health IT helps improve the outcome issues.
So a lot of work being done at that level with the Health IT Standards
Committee and the Clinical Quality Work Group.
DR. CARR: What I have talked about as I have spoken with almost everybody
on the committee leading up to this, is these parallel universes that are going
on on the same topic, and how important it is that we have a landscape of what
are the activities, who is in charge, to whom are they reporting, and where are
the gaps and where are the redundancies.
We talked in Standards about collaborating with the HIT Standards
Committee, to almost come together and say this is the piece I am going to be
doing, and you are welcome to participate and this is what we are doing, and
welcome your participation. So I think we have got the start of a couple of
landscapes underway in privacy and standards; we need it in quality and
populations now even with the element three. So I think that is a very good
point.
DR. SUAREZ: The importance of cross communication; I think there is going
to be plenty of topics and areas and priorities that each of the various FACA
committees are going t o be dealing with. So there are going to be places where
our committee here clearly is going to have the core responsibility, and then
there is going to be areas where the HIT Standards Committee will have that
type of responsibility. That way we don’t duplicate.
But I think there is also going to be areas where it will be beneficial to
have joint activities more and more. In the case of hearings, for example,
where the information from the hearing will benefit the three FACA committees.
But I think it is going to be beneficial to look at more and more coordination
in that area.
MR. QUINN: I am Matt Quinn from AHRQ, staff to the Quality Subcommittee.
Just to frame this a little bit, we had three recommendations for the
meaningful measures hearing. One was that there be a national coordinator for
quality. Second is that we should start with a core data set which is open,
public, well specified, so that it can be implemented, and that that should
serve as the core for our measurement activities for both meaningful use and
more broadly. Then third, a way to get this implemented is to have it as a
certification requirement for systems. Not that that is the only data model,
but that commercial systems that volunteer to be certified can support that
model.
What we see in the current meaningful use talk is mandating a handful of
measures which will then be hard coded into systems in support of the 2011
requirements. What this presents the opportunity to do, as opposed to starting
early on hard coding the next measures for 2013 and the next five or 15, is to
say here is a denominator of data elements from which we can pick and pull a
variety of measures. The question is balancing how big that is, how small it
is, based on where we want to go with it and what we can support.
We have some examples of organizations taking that step of saying we have
gotten a standard model, standard data elements across organizations to do
this, and in many cases it is not nirvana, it is a rough path. So just to frame
it a little bit.
DR. WARREN: Based on your work on the quality models, there is some work
going on in the VA. Right now they are working on falls, what is the model for
data that you need to track falls that are outcomes, and then taking that,
modeling it to HL-7’s RIM so that we insure interoperability, and then encoding
that in both SNOMED and LOINC.
One of the things that we have been doing with Betsy Humphreys at the
National Library of Medicine is trying to create a library of encoded
indicators so that each institution doesn’t have to do their own encoding. Just
go there, pull it out of the UMLS and be ready to run with it. That way it is
all ready to be translated or transferred or aggregated up in whatever way you
want to take a look at it.
So that group you might want to touch bases with. I can give you name and
e-mail.
MR. REYNOLDS: With that, at 11:00 we are supposed to be — right on time —
turning it over to let Justine handle the plans for September and future
meetings, and then I’ll wrap up.
Agenda Item: Plans for September 2010 and Future
Meetings
DR. CARR: I never cease to be amazed by the richness of this committee and
the shared learning. One of the things that I enjoyed a lot about the last two
days is the opportunity to engage in dialogue and have enough time to share
ideas. We have been doing more of that in recent years, and I think that we
need to always create the time to do that.
A couple of themes came out of this for me. One is that we have a lot of
work currently with 14 members. Don’t feel guilty or anything, Harry, but we
are down to 14. So we have to be very thoughtful about how we prioritize our
work. We also have to all share in it.
I think the work that we are assigned on standards has to be the work of
everyone. We all have to be up to speed on it, we all have to be reading it and
thinking about it with our perspective, whatever our subcommittees are, and
then coming back and putting out whatever your thoughts are.
The time line is very short. We have got to have a letter in September. The
background materials are coming together in the next week, and we will
distribute them. The hearings are in July 19, 20, 21, so we need to think about
whoever will come up with Judy’s request for representation from each
subcommittee. But we need to be in conversations.
I think we had a great example of that, where we missed our February
meeting and yet, look at the wonderful report-out that we will be giving today;
we did it with phone calls and so on. I think we are going to have to be
prepared to do that as well. So I would say that the work of Standards is the
work of the full committee, and we will be on a sprint. It is going to be a lot
of work but if we all work on it together and kick the tires on everything that
we hear, push back on everything that we are trying to say, because it has got
to be our best, best work, and we need everybody’s perspective. Even if it
turns out not to be the right thing, just get it out there.
That is issue number one. I see Standards as work number one. Number two,
what we just talked about, is that each of the subcommittees will put together
a landscape of who are the other players in the particular area, privacy,
populations, quality and standards, and what are they doing. Then we use that
going forward so that we are checking on a regular basis with what the other
committees are doing. As Walter brought up with standards, that we reach out
and have a very specific conversation and collaboration where appropriate.
Then the third thing, I know we talked about it in quality and populations
since we were one and the same, is, we have this big agenda, and we are all
carrying a lot more of the work, that we think about that Populations and
Quality work together on a particular agenda going forward.
So those are the overarching thoughts that I had coming out of today. In
September 16 we will have a half day Executive Subcommittee retreat to talk
more about this and map out how we see ourselves, not just in the near term,
but the next two years. Then we will have a follow-up on that in November, a
continuation of that Executive Subcommittee retreat.
DR. SUAREZ: Is that September 16 now?
DR. CARR: Are we set for the 16th?
MS. GREENBERG: The full committee is meeting September 15 and 16, and then
that would be the —
DR. SUAREZ: It would be the 14th or the 17th. It is going to be the
afternoon of the 16th. We will meet until noon as the full committee, and then
the Executive Subcommittee will have that afternoon, the full afternoon.
Then our last meeting of the calendar year is December 1-2, so we were
talking about November 30 for a full day. I think you said the Health IT
Standards Committee was meeting that day. So Walter is on both of those.
DR. CARR: I think for now we are going to have to stick with November 30. I
think we need some framing before the full committee to be able to report out
where we are headed.
MS. GREENBERG: I raised the question of whether there was any possibility
for a joint session, an hour or two, with the Standards Committee on that date.
I don’t know whether that is something —
DR. SUAREZ: The agenda for the Standards Committee doesn’t get set until
much later, so we don’t know what that is going to be.
MS. GREENBERG: Maybe in your capacity as being on both you could suggest
that. We are not sure where we are even meeting right now, and I don’t know
where they are meeting. It is all going to be in the Washington, D.C. area. I
don’t know whether you might be able to join us by phone, be in two places at
once. I know how much fun that is.
DR. CARR: I think way prior to November 30 you would begin a conversation
with that committee of this idea that we brought forward today, so that by
November 30, if we came together, it would be because we had already gotten an
idea of what we wanted to do.
MS. GREENBERG: Regarding the landscape, we know there are a lot of people
in the privacy and security and standards space, quality. The NCHS DFC has had
a bit of a hiatus in getting new members, new chair and everything. They seem
to be up and going again. When is their meeting?
PARTICIPANT: September 23-24.
MS. GREENBERG: So it is right after our meeting. So I think we will have to
make sure we have our liaisons in place there as well on the population side.
MR. LAND: Marjorie, could you clarify the dates?
MS. GREENBERG: The agenda is the dates that we had had. Originally we had
the November 17-18 dates, but those represented some conflicts. So we
rescheduled that meeting and we did Google everybody, to December 1-2. But I
apologize that this still said November 17-18.
So the meetings the rest of this year are September 15-16 and December 1-2.
Those are day and a half meetings. The afternoon of September 16 is an initial
retreat of the Executive Subcommittee and a full day on November 30 for the
Executive Subcommittee.
MR. HOUSTON: In September is anything going on? Are there any subcommittees
meeting on the 14th?
MS. GREENBERG: Were any subcommittees planning to meet on the 14th of
September? I know Privacy had thought —
DR. CARR: I think what we would like to do is check with Quality and
Populations to see whether we would be ready to meet then.
(Discussion off the record regarding future meeting dates.)
DR. CARR: The plans are set both with the hearing, the landscape and the
communication with the FACA. Quality and Populations will perhaps even decide
today about whether that hearing will occur on September 14. We will set up a
call to get that going. And Privacy is going to have a letter by September.
MR. HOUSTON: Yes, we would like to have that letter available for review in
August. If we need to have a conference call to discuss the letter we will do
that in August as well.
DR. CARR: I wonder if it would be helpful, with the idea of keeping the
whole committee up to speed, to have a background document, why is this letter
important, what are the things we are trying to address, what did we hear, what
did we learn, what are we deciding. That way we will all be up to speed.
MR. HOUSTON: We can try to do that, or maybe on the surface make sure the
letter is clear enough. We will try to give ample opportunity for input and
review in the August time frame. I think that is reasonable.
DR. CARR: And you are doing your landscape both on governance in terms of
who is in charge of privacy from the micro to the macro level, as well as for
the agencies.
MR. HOUSTON: Yes, that is in September.
DR. CARR: Anything else? I think we have a busy summer.
MS. GREENBERG: Also, if you want to have any presentations to the full
committee in September, we need to start working on those. Being a transition
meeting and follow-up to our paper and to our symposium, et cetera, I would
think there would be a meeting in which you don’t have a lot of time for
committee discussion. And of course we will be needing to discuss the privacy
letter —
DR. CARR: We will have two letters.
MS. GREENBERG: — and the standards letter. You will have a letter on the
operating rules and another letter on the plan. So at this point it might be
that we not think about having any outside presentations.
DR. CARR: Maybe one or something like that.
MS. GREENBERG: Maybe one at most.
DR. CARR: We will have an update on meaningful use at that point as well,
hopefully, where we are with the final.
MS. GREENBERG: From Chuck and Karen and Jim, I would think we would have
those.
MS. JACKSON: We were looking at follow-up from this meeting as well, but
you have not seen each other’s PowerPoints. You will see them today for the
first time. We wanted at one point to have the menu folders to have everyone
see everything. We realized that was a bit much. But at the next meeting would
be an opportune time, particularly considering that the Executive Subcommittee
would be meeting directly afterwards, to keep the synergy going, the flow of
all of the work, seeing how much has coalesced, with the presentations we had
yesterday.
This is a lot that is happening that we can get our hands around at the
next meeting. Maybe have one presentation like what happened with this time. It
was very strategic. We had a whole new way of organizing the subcommittee
sections yesterday in comparison to other meetings. So things like that I think
will work for September.
MR. REYNOLDS: In closing. It was kind of fun to sit here and have Ron as
one bookend and Simon, our most recent mentor, on the other end.
One thing I would say. I hear a lot of voices in my head as I go out. Make
sure as a committee — we have been around for 60 years. We have been relevant
for a long time. We will be relevant for a long, long time. Sometimes it is
easy when there are other committees and other things that are going on right
now that look like the limelight might be there, the action might be there, the
relativity might be there. This is a long game with a lot of turnoffs, with a
lot of rest stops, with a lot of other things going on.
So be careful that the committee doesn’t stay out front, because out in
front in this environment, two years, three years, five years. We talked about
the hearings for 2015 meaningful use, starting the meaningful measures. Setting
the plate so that others can pick it up and talk about how to implement it and
what it means has been a real sweet spot as a committee, whether it was under
Ron’s tutelage or Simon’s or John’s or anybody else’s. I think it will be the
same for you. So make sure that you keep that in mind.
Having just gone through it, I have gone through the ONC ramp-up. There are
times when you are sitting and thinking about going, are we really doing okay,
and we continue to do okay. We have got a great team, play that way, play hard.
Jeffrey, I salute you.
I am going to adjourn this and turn it over to Justine. Thank you very
much.
DR. CARR: We’re done.
(Whereupon, the meeting was adjourned.)