[This Transcript is Unedited]
Department of Health and Human Services
National Committee on Vital and Health Statistics
Subcommittee on Standards
February 24, 2009
Hubert H Humphrey Building
200 Independence Avenue, NW
CASET Associates, Ltd.
Fairfax, Virginia 22030
- Welcome to “Setting the Context for the
Evolution of Health IT Standards”– Jeff Blair and Judy Warren
- Panel 1: ONC Update and Strategies
- Robert Kolodner, M.D.
- Charles Friedman, Ph.D.
- Panel 2: IT Thought Leaders
- Don E. Detmer, M.D.
- John Tooker, M.D., M.B.A.
- Carol Diamond, M.D., M.P.H.
- Panel 3: Enhancing Standards Adoption by Users
- Steve Waldren, M.D.
- Alan Mertz
- Rosemary Kennedy, M.B.A.,R.N.
- Rod Piechowski
- Panel 4: National IT Initiatives – Role of Standards
- Douglas Bell, MD
- Gil Kuperman, M.D., Ph.D.
- Linda Fischetti, R.N.
- William W. Stead, M.D.
P R O C E E D I N G S
MR. BLAIR: Hello everyone. On behalf of Judy Warren, my co-chair, and
myself, I’m Jeff Blair. I would like to welcome everybody to the subcommittee
on standards and security of the National Committee on Vital and Health
Statistics. For the national committee is the statutory advisory committee to
the Secretary of Health and Human Services dealing with the topics of health
information policy and standards and obviously this is the standards
subcommittee of the NCVHS.
The first thing that I would like to do is to go around the table where each
person would introduce themselves. As we do that, please speak into the
microphones. We have new microphones today. They seem to be very responsive. We
do have people that are on the Internet and the only way they can hear us is if
we speak clearly into the microphones. As we go around, please identify whether
you are on the committee or not, what organization you might be part of, and if
you are on the committee whether there is any topics that are happening today
where you need to recuse yourself.
I will begin by saying I am Jeff Blair. I am Director of Health Informatics
at Lovelace Clinic Foundation and I have no conflicts that I am aware of.
MR. REYNOLDS: Harry Reynolds, Blue Cross Blue Shield of North Carolina,
member of the subcommittee. No conflicts.
DR. GREEN: Larry Green from Colorado, a member of NCVHS, interloper on the
subcommittee. No conflicts.
DR. WARREN: I’m Judy Warren from the University of Kansas School of Nursing,
member of the full committee and co-chair of the standards committee. I will
report that later we are hearing testimony from the Alliance for Nursing
Informatics and do belong to several organizations that are members of that
alliance, however, saying that I don’t proceed any conflicts.
MS. TRUDEL: Karen Trudel, Centers for Medicare and Medicaid Services, staff
of the subcommittee.
DR. FITZMAURICE: Michael Fitzmaurice, Agency for Healthcare Research and
Quality. I am not a member of the subcommittee. I am a liaison to the national
committee and staff to the subcommittee on standards and subcommittee on
MS. BUENNING: Denise Buenning at CMS Office of E-Health Standards and
Services and lead staff to the subcommittee on standards.
DR. KOLODNER: Rob Kolodner, Office of the National Coordinator.
DR. SUAREZ: Good morning everyone. I’m Walter Suarez with the Public Health
of Spanish Consortium.
MR. BLAIR: And you are a member of the subcommittee and do you need to
recuse yourself for any reason?
DR. SUAREZ: Excuse me, yes. I am a member of the subcommittee and I don’t
have any conflicts.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics committee
MR. SLACKUS: Steve Slackus from CAQH.
MR. BELL: Doug Bell from Rand Corporation.
MS. BICKFORD: Carol Bickford, American Nurses Association.
MS. ZYMAN: Carol Zyman, Markle Foundation.
MS. WILLIAMS: Claudia Williams. Markle Foundation.
MR. CONNELLY: David Connelly, Capital Associates.
MR. WANG: Derek Wang, Social Security Administration.
MS. HELINISE: Marian Helinise, student at University of Maryland.
MR. BURBAN: Adam Burban, Blue Cross Blue Shield Association.
MR. JAFFEY: Charles Jaffey, Health Level Seven.
MS. VIOLA: Allison Viola, American Health Information Management
MS. NATHER: Jessica Nather with the Personal Healthcare Initiative here at
DR. WALDREN: Steven Waldren with the American Academy of Family Physicians.
MS. HOLLY: Beth Holly from the Mitre Corporation.
MS. KENNEDY: Rosemary Kennedy representative of the Alliance for Nursing
Informatics, Chief Nursing Informatics Officer at Siemens Health Services.
MR. REVRENT: John Revrent with the Healthcare Leadership Council.
MS. JOHNSONAM: Hi, good morning. I’m Angela Johnsonam with the American
MR. DETMER: Don Detmer. We’re here by the plug. AMIA.
MR. BLAIR: This is the first day of four days of hearings on healthcare
information technology standards development, standards selection, standards
implementation, and frankly those three are listed just so that we are implying
that whatever processes relate to health IT standards we welcome observations
and comments and suggestions. The topics that will be covered today are only
what we could include in the first day kind of a kick off day. There will be a
second day, which will probably be April 28th, where we will be
hearing from HITSP and from standard development organizations that cover
messaging, that cover mainstream terminologies, that cover other clinically
specific terminologies and terminology mappings.
There is going to be a whole day for different standard development
organizations to testify to us. We also intend to hear on the third and the
fourth days from HIT vendors and other users of HIT standards. Some of those
users can be represented by professional associations and some of them by trade
associations and others by individual entities.
Obviously we are going to try to hear from as many different parts of the
healthcare delivery system as we can, but it’s not going to be possible even in
four days to hear from everyone. So please understand that we will also be
soliciting written testimony beyond in person testimony from many of the same
groups that I have just mentioned. This process, these four days of hearings
will go on. The second and third day will be in April, and we are expecting the
fourth day of hearing to be in the June timeframe. This is an open forum. NCVHS
has a history of listening to different stakeholders to be a forum to get
feedback on how things are doing.
Now I think just about every person that I heard introduce themselves has
been involved in the health information technology standards activities in one
way or another over many, many years. I am going to just place a little bit in
perspective here, and I think that a lot of you will probably resonate to this
because if we look back over the last 10, 15, 20 years, it became clear to many
of us in the industry that standards needed to play a more important role. We
needed standards for inoperability. We need standards for clinical specificity
for decision support. We need standards to make sure that diagnoses are more
accurate. We need standards to be able to measure patient outcomes and
improvements in patient outcomes. Healthcare information standards have been
called upon to play a larger and larger role. Complementing the ones that I
just mentioned is the ability to accurately identify the patient and to also
protect patient privacy and ensure that the data maintains its integrity with
security. That is not an emotional issue. The cough had nothing to do with the
content of my message.
One of the pieces that I would mention as I started to lay out the context
here is that as we have looked at standards we have had a few major milestones.
In 1996 when HIPAA was passed, it called on the NCVHS to select the standards
for HIPAA transactions, for identifiers, and for privacy and security. NCVHS
was asked to play this new role at that time. One of the other pieces in there
was that we were also to begin to set a framework and make recommendations on
standards for patient medical record information, which many of us think of as
standards related to electronic health record systems.
It was either 2003 or 2004, when Congress again asked us to evaluate and
select recommend the standards for E-prescribing and we did that. In terms of
many of the standards that we were recommending they have evolved into an
organization that was called Consolidated Health Informatics Initiative. Those
standards were then adopted by the Secretary for DoD, VA, and HHS. In other
words they became a framework and a basis for standards for healthcare
interoperability within the Federal Government.
One other perspective was that NVCHS back in 2001, felt that we needed
greater coordination of health information policy and standards and to move a
little quicker. We had a report called the National Health Information
Infrastructure. The key thing out of that was of course that we needed an
office, which is now referred to as the Office of the National Coordinator.
With this history as you have been listening to this you have been thinking
that gee, in the 1990s we accomplished quite a bit, but things began to speed
up after 2000. And then after 2004, when the Office of the National Coordinator
became in place and AHIC was established and HITSP was established and the
trial implementations for the NHIN was established. It has been clear that in
the last four years under the leadership of ONC, there has probably been more
accomplished with healthcare informatics standards in terms of harmonization
and getting them to be implemented to move forward. There’s probably more that
have been accomplished in the last four years than may have been accomplished
in the last decade or two. Tremendous progress has been made and of course as
you have all read the stimulus package, the American Recovery and Reinvestment
Act, ARRA, you have seen that we are now being called upon to meet the nation’s
needs even in a more specific direct way.
NCVHS has a forum. Hopefully we will be able to play a role in this, a
constructive role in soliciting feedback from all of you over the next four
days of testimony to help us maybe begin a continuous quality improvement
process so we can take what has been a really good effective foundation that
has been built over the last four years, and it was very gratifying to see that
the law has built upon that foundation, and provide feedback so that we could
improve even more.
That is the perspective that I felt would be useful for us to begin with.
Judy, do you have any other observations or comments?
DR. WARREN: No, I don’t have any.
MR. BLAIR: Well, then let me say we are delighted to have Rob Kolodner from
ONC. Is Chuck Friedman here with you as well? Let me turn it over to you.
DR. KOLODNER: Thank you very much, Jeff. It really is a pleasure not only to
be here, but also to collaborate with you on these four days of testimony. As
you know when we first started to plan this, it was before the bill had been
passed. We weren’t quite sure what would actually go into law and now we have
that so the timing is just right for getting started although as you will see
we have some notional directions. There is still a lot of discussion, decisions
need to be made in HHS and across the government, and hopefully over the course
of these four days that you have spread out ahead we will be able to update you
as activities occur and as decisions and directions are made.
What I would like to do is to frame up both the discussion today, but also
the discussion for the next four meetings by going over a few things. First I
wanted just to do a quick review for those who may not be fully immersed in the
process just what’s happened during the past four years or so that the Office
has been in existence under executive order. And then take stock of where we
are now particularly in light of the Recovery Act that was just passed.
Then I want to talk about some approaches for the next stages. These are
more of general direction actually where I would very much like to get feedback
from you and from those who you are hearing from so that we can make sure that
we move forward in this next stage and build upon what we already have in place
because we have been given an interesting charge as we came to being in statute
rather than just in executive order and that is that basically President Obama
doubled our target in the same amount of time in terms of the Americans who
would have electronic health records by the same timeframe, which is 2014. We
need to find the most efficient ways and most effective ways to move forward to
make sure that we are doing that in a way that honors the privacy of the
information of each individual and yet effectively puts in an infrastructure
that is not an end in itself, but enables the kind of transformation that we
need to have in the healthcare and health systems that we have in the country.
First of all as far as the quick review, sometimes we have talked about this
as being turns of crank and there was a process that was recurring. It started
with setting priorities. Those priorities were then turned into scenarios or
were called use cases, that were harmonized – at least the standards that
would support them were harmonized from existing standards. We weren’t in the
process of creating new standards in this phase. Those standards came back to
the prioritization body, which is FACA, the AHIC, The American Health
Information Community, and were then recommended to this Secretary who accepted
them, and then with a year gap recognized them. That year gap was to allow
industry to test them for them to be refined and industry to begin building
them into the products.
During that time we established the organizations that were necessary for
that so the Federal Advisory Committee as I mentioned, is AHIC, and it had both
public/private representation and was chaired by the Secretary of HHS, a very
unusual move as you know, for FACA.
We brought into being for contract a group called the Healthcare IT
Standards Panel or HITSP, which engaged over 500 different organizations in
terms of the process of gathering together and harmonizing those standards and
presenting back what were called interoperability specification as the
standard. Then there was one more group and that was the Certification
Commission for Healthcare IT. It was formed actually before the process and
then the contract was left to them to act as the certifying body for various
products starting with ambulatory EHRs and going through others as well and now
into a networks and PHRs in the coming months.
There were a number of drivers that we had for having the use of these
standards actually more in the shelf. There were requirements, executive orders
for the federal entities to build it into their products as they upgraded and
to include it in their contracts and some of the grants as they move forward.
There was also an exception that was made to the self-referral provisions,
Stark and anti-kickback, that allowed for larger entities, hospitals to donate
up to 85 percent of the cost of the system to small providers as long as the
systems were certified or actually as long as they were interoperable. The idea
was that there wouldn’t an attempt to lock them into just feeding the patients
and information to donating it – that it was a broadly establishing
infrastructure. That was the process we did establish these various entities.
We did have a series of the cycles of turns of the crank.
On this next slide I show the various use cases that were approved in 2006,
2007, 2008, and then what has been submitted to HITSP for their work in 2009.
It is organized by year. You will notice that in 2006, there were three use
cases, consumer empowerment, EHR which was the laboratory results reporting,
and bio surveillance. On the slide you will see that there are three rows. One
is the consumer activity, and the other is for provider activities, and another
is population health. The use cases were reasonable balanced across those three
perspectives. In 2007, we increased up to four of these cases including
emergency responders, EHR, and medication management for the provider of
Consumer access to clinical information much broader array, and this really
is also the EHR transfer, the use case that provided for a standard for
communicating to and from PHR as well as having access to clinical data from
providers. There was a quality reporting initiative as well.
Those standards have all been not only accepted but they have been
recognized. All of those reach that next threshold under the previous mechanism
were then triggered to be included in products as requirement.
In 2008, we went up six use cases. For consumer it was remote monitoring and
patient providers through messaging. For the provider focus it was
consultations and transfer of care in personalized healthcare getting into the
genomic data and family medical history. Then on the population side, it
included public health case reporting and immunizations and response
management. Those have all been accepted and I am going to show you in just a
moment, the acceptance of those and I will be presenting that also tomorrow to
In 2009, the discussion on the priorities was that there were a lot of
standards out there and as I will get to there was a lag in getting them some
of them in product. Rather than just generating all new use cases there was one
that we thought that needed to be forward and that was newborn screening. The
rest of the capacity that HITSP had for harmonizing, we used by tweaking and
finding those gaps or extensions that needed to be added to the existing
standards. There is a whole list of those that I won’t go through now, but
things such as the general laboratory orders. We had lab results but the
laboratory ordering hadn’t been added and clinical encounter notes and aspects
of consumer preference and long-term care assessments. There were a variety of
things that we were doing there.
On this slide you see the updates to the existing Interoperability
Specifications that were accepted by Secretary Leavitt in January. They
included ones in buyer surveillance, updates on consumer empowerment and access
to clinical information via networks, emergency responder, and consumer
empowerment in access to clinical information via media. Then there were the
various underlying constructs and selected standards that formed that were also
In terms of new Interoperability Specifications that were accepted, we have
seven of the eight medication management personalized healthcare, consultations
and transfer of care, immunizations and response management, public health case
reporting, patient providers’ secure messaging, and remote monitoring. The
other will be moved forward in the next month or so, but part of what we have
to look at is because the process is changing as I will get into, what does it
mean to have those reference standards in view of the Recovery Act and some of
One of the discussions that we had at the agency last June in 2008, after
the first full cycle of standards being accepted and recognized and then the
first certification cycle after a set of standards had been recognized was
looking very honestly at the adoption of those standards and incorporation into
products. While there were certainly movement and a number of those were
incorporated most particularly the continuity of care document, which allowed
the ability to begin to move information and a summary record as you will see,
back and forth. There was still not as robust enough take as I think some
people had hoped and anticipated.
There was a very lengthy discussion that we were able to have with the heads
of HITSP and CCHIT and the members of the committee. They talked about the fact
that it was the absence of incentives to move those standards in. It was really
a barrier that while there were many requirements, most people weren’t using
that information or exchanging it, and when vendors had the choice of do they
build in that capacity, that standard, or do they provide other features that
their user base was asking for. When they asked the users, the users weren’t
saying oh, I want this interoperability because I want to exchange information.
That’s really understandable if you think about the world of
interoperability as it exists today. There isn’t any widespread way to exchange
information. The information is exchanged usually at a fair wise action of
coming together sometimes in a small cluster, but the incentives weren’t
sufficient to bring the providers and the vendors to the table. One of the
points made was that we need to have a better way to do that in order to really
move those standards, get them used, refined, and into product.
In terms of integrating the standards into the federal purchasing and
contracting they had some variable success. There were certainly a number of
agencies that were raised to them and were struggling to use them. Again it was
variable. Part of that again was that we didn’t have all the pieces in place
because the ability to exchange was still limited by the fact that we really
didn’t have any networking capabilities. We had standards and we had some of
the products, but that ease of using it was absent.
At the end of the last calendar year in September and then December we had
one of the key pieces reach a level of maturity that we had been looking
forward to and that was nationwide health information network. There were trial
demonstrations involved in there for 20 various organizations and entities.
This was really critical and by the way one of the things about the NHIN is
different people’s perception of what it is. It is a set of specifications and
standards and agreements. A lot of the NHIN has to do with nontechnical aspects
of trust, a use agreement and governance that are as this point still in the
early formation. The trial implementation was meant to be a first test to see
what didn’t take to put all of those together.
NHIN is very much a collaboration. It’s not a top-down, here are the
standards, you must take them. They in fact started with a set of recognized
standards and even using those that were accepted, but it was bringing together
these 20 or so entities to work together and say what does it take to actually
get to use these?
Let me show you at least conceptually what this was like. We start with
health information exchange organization to RHIOs that would be in various
communities and tied together a variety of entities in those communities. They
also have integrated delivery networks, things like Kaiser Permanente and
Cleveland Clinic, in the government, VA or DoD or any health service. The
promotion of the concept of Health Bank or of PHR support organizations,
vendors, or plans, or others that are offering the PHRs to individuals.
Community health centers. It is very important as we move forward that we not
increase disparities, in fact, health IT has the opportunity to actually help
foster a decrease or an elimination of those disparities if used effectively.
We need to make sure that they are connected to this network as well. A variety
of federal entities that either have connected or are beginning to plan their
connection to the NHIN. State and local governments are critically important
both for the public health aspects as well as for their employees and other
activities that are important within the health arena. Then large entities such
as laboratories, pharmacies also that may effectively be a network as a large
We connect them not by creating a new network of course, but by going over
the Internet. Since we know that while that’s effective, it’s not completely
secure. In order to mobilize the health information network it is that set of
Specifications which are the things that are worked out amongst the
participants that haven’t yet become standards, but will loop back to become
standards to a feedback loop as well as these nontechnical aspects and
agreements to create a secure connection. That is really the NHIN concept that
Over the course of time that I have been in this job the last two years
there have been a number of, more than a number of times where people have come
and said I have the answer for the nation’s connectivity. I listened very
carefully as do my staff and they may very well have the answer, but they may
not. Since not all of those agreed with one another, they were actually very
What we saw the role of the NHIN as being is to create a neutral platform
that allowed the best of the solutions to compete and to move forward and
really to foster innovation and creativity not by creating barriers or not by
favoring one view or another, but by making sure that there was a secure way or
a reliable way and a trusted way of being able to exchange that information.
One that was in essence neutral to the architectures of the entities coming
forward or to the solutions that might emerge. Think of this as kind of an
initial common connection and if by market force is a consumer, providers chose
to use one or another then that really could spread and it could either shape
the future of the NHIN as long as it met the needs of everybody. It wasn’t
dominance or controlled by anyone –- really the evolution like we see on
the Internet where one company or new idea comes forward even though it
threatens other entities that may have dominance in the market. People choose
where they click or what products they use, what browsers they use, and so they
really can compete on the level playing field. I think that is where we see the
role of the government.
We had a number of –- these three different types of networks that we
envisioned and network in the sense with quotes, an organizational network of
multiple sites. As I say the interrelated delivery systems or hospitals chains.
Geographic which are the standard RHIO or HIEs and then these PHR-oriented or
personally oriented support organizations. It really has to be hybrid because
all those exist and it isn’t that it favors one or another, but that it allows
however entities whether they be providers or support organizations or
individuals themselves of how they want to connect that they have various means
that they can choose to do so.
We went through just prototyping that this would be capable in our first
phase that was 2006/2007. NCVHS played an important role of evaluating that and
selecting from that. What was the core set of services that were common across
other various solutions? We then had the trial implementations last year and
then very successfully demonstrating these activity by exchanging information
of all patients, test patients across test systems, but they were located
across the country using all of the constructs.
Then starting this year of move into what is called a path to production is
really starting with limited production pilots which allow us to begin putting
this together and make sure we get the necessary infrastructure that wanted
these in order to really go into production, things such as monitoring the
network making sure that we have an ongoing vulnerability detection capacity
and even testing constantly to make sure that the new technology is used by
hackers, can’t break what we have and keeping it very secure and going forward.
That is what we plan to start this year and we have been really very pleased by
the participation and the interest.
The last year with the NHIN, we demonstrated the technical capability. We
created these initial Interoperability Specifications. We showed could work
across these different organizations with very different underlying
architectures. We were able to organize these organizations into a cooperative
that work together and that included both the public and private entities and
lay the foundation for the usage agreement began to explore and identify the
long holes –- things that we needed to work together on and really
completing those trial implementations and comparing for the limited
On this slide you see the 21 entities that participated. They include about
six federal agencies and two integrated delivery networks and the rest being
community HIEs and RHIOs from across the country, some smaller, some larger,
some early in their existence, and some that are extremely well established. I
think it was something where the experience of having it come together was one
that was a very important one to teach us because it had to be where we have
set the direction and the goals top-down, but where the real solutions had to
meet the needs of all of those entities to participate.
One of the things that we had early in typical techy style was we had an
opportunity before the demonstrations for the entities to get together. We of
course chose the garage, the appropriate setting where people would be
comfortable, and brought the entities together in August of 2008. We had
clustered them into groups of three or four around different scenarios of
summary records. Those groups were able to exchange the records amongst them.
We brought them together and said okay now communicate with each other. Despite
the very detailed specifications that came out of HITSP and the work that we
had done at that point in time, they were not able to connect across. What that
taught me is that if you rely on groups testing standards and refining them and
they do it in a parallel way or separate from one another, they will succeed,
but they will come up with the necessary refinements and tweaks and resolutions
that in fact won’t foster the interoperability on a nationwide basis.
What we needed to do was to bring a sufficient number together and make sure
that they continue to spiral out and involve more. That it was through that
process that we were able to identify those last few specifications. Sometimes
groups would interpret the standards slightly differently or we found that with
one architecture something wouldn’t work quite as well. We are able to then
resolve it so all of them were able to work together with the idea of being
that those improvements then would be fed back with feedback looped back
through the standards process, updating of those standards, or for things that
were these specifications, the ones that went beyond what was part of the
initial interoperability standards of things that allowed the networks to
connect together was most of the standards were developed for use within a
single network. This is the feedback loop that is needed for refining the
It goes very much to what Carol Diamond had written about and her colleague
in health affairs that we had a dialogue about, which is you don’t have
standards by creating them by a committee and then having them on a shelf. They
need to actually get used.
The last part of that constant we were putting together was having a network
that allowed entities to begin collaborating together and beginning to use
those standards and really refine them so that they are then ready for movement
outward. That is important for I think how we go forward as well.
The key thing is that in addition to the 20 organizations coming together
and the movement of the summary record –- let me show you – I believe this
is the last slide in the NHIN pack. For September we were able to use just a
summary record. It doesn’t have a lot of great detail of a huge number of
standards, but it allows you to move a readable record with certain parts that
actually are interoperability and computable. That was a very powerful first
step and we were able to show multiple entities using it for emergency care,
for transfer of care. The VA/DoD collaborated with private sector and showed
treatment of wounded warrior that went back and forth across federal and
private sector. We also had something supporting the consumer and finally an
innovative business use that was there.
The innovative business use was actually Social Security which could turn
around the gathering of the information in minutes and therefore give the
evaluation in a week or so compared to the usual 6 months or more and the fact
is that Social Security spends half a billion dollars a year collecting the
information for its couple of million disability claims each year. That’s a
real potential savings.
The key thing here is that we were using an approach that said we wanted to
foster the widespread adoption of the standards by using that acceptance as a
driving force. That acceptance meant that it should be in products. It should
be in the feds and therefore all of these things were kind of moving forward
and would be adopted. I think that what we saw in this is that while that made
sense, it didn’t work quite that way. That what we want to do is to build on
the things that we learned and the good things that are there and begin to move
forward in something that allows us to be even effective in this next phase.
I think as Jeff said, in the last four years there has been an awful lot
accomplished. And compared to anything before that and Jeff ran through the
history, it is as if we are in phases. This may be the third phase as I will
outline in just a second. Now with the passage of the American Recovery and
Reinvestment Act of 2009, still looking to see how it’s referenced in its short
form, but the Recovery Act is sometimes used as a referral. That it includes
significant funding -funding that we have never had of this magnitude into
health IT. It creates an opportunity as well as outlining certain requirements
that will allow us to move forward.
I think one of the key things and it was a question certainly that was
either spoken or unspoken at the end of last year, was would all of the
activity of the last four years cross over? Was it going to be embraced by the
new administration or was it going to simply be that’s nice we’ll set it aside
and we’re going to start over or do something different? I think that the
language of the bill, the actions by the new administration really showed that
we have crossed over. It is the best evidence that health IT really is truly a
bipartisan issue and that it is recognized that for any solution we put in
place for solving the healthcare crisis that we now have, health IT is an
essential enabler not the answer, but an enabler that has to be there in
addition to other important steps.
There is a detailed section in the bill that regards health IT. It is called
the Health Information Technology for Economic for Clinical Health Act or
HITECH Act, and just a few of the things that relate to this committee and what
we are talking about here wanted to establish ONC in statute. The position is
now appointed by the Secretary of HHS and reports to the Secretary. It makes
explicit reference to the federal health IT strategic plan that we released
last June in collaboration with our federal colleagues. And actually charges us
with updating that through collaboration not only with public entities but with
public and private entities.
Given the 9 months that it took to get it through public, it will be
interesting to see how we expand this and make sure that there is adequate time
not only for input by the private entities, but for the engagement because the
input from the feds was not just a you should do this. It’s we’re going to do
this. One of the issues here is how do we make sure that there is a we as part
of that feedback from the private sector as well.
The AHIC as you know was a FACA. It was actually sundown in November and the
nonfederal –- transition to a public/private entity and I believe you will
be hearing at the Full Committee tomorrow, from Laura Miller who is the interim
Executive Director with that. There is some language there about that entity
because the bill established to FACA. There is some wording about what this
public/private entity can do and is still being explored and awaits a new
Secretary as well, for making some of the decisions.
It took one FACA, the AHIC, and created two FACAs. One a policy committee
and one a standards committee, and defined the roles for each. It also required
that the standards implementation specifications and certification criteria
which is its full set of standards that it is referring to be adoptive through
a rulemaking process not through an acceptance or recognition by the Secretary.
That has important implications as well as you know from your experience with
HIPAA and rule making.
It also requires that an initial set of standards and implementation,
specifications, certification criteria, be adopted by December 31st
of this year. It allowed that to be through interim rule making so the interim
final rule so you can’t actually achieve that date, but that is still a very
aggressive date to achieve and especially because it added in some new
standards that hadn’t yet been fully included in the previous standards such as
making requirements for race, ethnicity, primary language of the patient be
included as well and a variety of other areas that were outlined in the bill.
It does allow us to use the ones that have already been recognized and have
been out there so we will be building on that. I will be talking about that in
just a moment.
In terms of where we are at as I have mentioned, the NHIN as a place where
entities can begin to work together using these specifications and start active
exchanging not on a point to point, but as cluster or of among multiple
different entities. It allows us to actually have a venue or crucible to begin
taking the initial set of standards that have come out of HITSP and get that
experience together rather than just in a laboratory.
With that being kind of where we are at let me talk about the final section
and the final things I want to talk about which is really where we go moving
forward in review what is the next stage. I think we need to think about this
as stages as they really are. There was a stage where we didn’t have standards
where we were trying to bring up various health IT activities. There were
stages in the early development of standards where we had multiple standards
organizations and I know Mike Fitzmaurice and others, were involved with an
attempt to coordinate across those standard development organizations with
We then had CHI, which NCVHS had been involved in, with Consolidated Health
Informatics, where the feds got together. We were saying the feds should just
do it. The feds got together and for themselves outlined the standards they
would use, but use the NCVHS as the public sounding board to make that they
weren’t doing something that was really at odds with where the rest of the
private sector industry were going. It wasn’t a matter of forcing that on the
private sector and then the most recent, which was the last four years with
Office of the National Coordinator.
We are now entering this next phase. It is one with significant resources
attached with a slightly different mechanism for how standards are – now the
term is “adopted”. What we want to do and what I hope that we will be
able to have as the dialogue ongoing in these next four days of testimony, are
how do we learn from that, learn from the past, learn from other experiences
and move forward in a way that really effectively accelerates and the adoption
in use of standards in a much more systematic approach, one that actually will
speed their adoption.
We are faced with the fact that the standards of adoption with two FACAs in
rule making does have incentives. It has one of those things we weren’t having
in the same way during the science phase. It is a longer process when you have
two FACAs instead of one that have to pass in a rule making process. What we
are looking at is viable means to make sure that we have interested parties
clustering together to use and test the standards that come out of the
harmonization process and feed it back in a very tight loop so that the
standards can be updated and further refined. We also need an approach that
really matches this very aggressive 24-team goal that is outlined in the
Rather than taking a huge number of standards and moving it forward, what we
are considering is that the goal is really to get good enough standard and a
set of good enough standards that are simplified in terms of how they are
promulgated and that can be reconfigured more easily to meet slightly different
uses that may not align perfectly with the use cases that were helped to
develop them initially but maybe close enough that they don’t need new
standards. They just need to be recombined in slightly different manner or with
a slightly different template.
To support these high priority needs because we are not going to be able to
get everything we need to use the 8020, and began to say what is needed to
support the health reform and to foster that widespread adoption, the pervasive
use of health IT in an interoperable fashion. Again it may not leave out in a
secure manner and one that honors patient preference and patient privacy.
In essence what we are doing is going from a model where the government was
going to have a process that created or resulted in standards and used the
recognition as the forcing function to get them used. To one where the actual
adoption of standards is a solar process because of the rule making because of
multiple factors. That the use of the incentive to share information went
through either Medicare, Medicaid dollars that are looming in 2011, or by other
By keeping out that barrier which held very capable organizations from
exchanging data and we all know of premier organizations that may be sitting
right across the street from one another that technically could exchange data
but have no business case for doing it. What we are doing with a variety of the
health reform as well as some of the dollars that are going out for Medicare
and Medicaid incentive by defining meaningful use to include exchange of
information. That is no longer the showstopper. We believe based on a number of
experiences, and I would be interested in those of you who are on the community
giving feedback about those as well, that this then will unleash the voluntary
adoption of standards by many entities that will work together because now that
the disincentive for sharing information has been taken away, they now will
have the reason to make sure that that information shared does meet their needs
and is something that is effective in where they can get efficiencies and when
they are being held accountable for the quality of care they deliver and the
efficiency of care that will help them to hit those targets and compete with
another by showing how much better quality and much more efficient they are
We have an opportunity between when a standard is harmonized rather than
looking at trying to get it immediately adopted there may be an interim between
that initial harmonization and the ultimate adoption to rule making that is in
fact and I think we need to figure out what term we might use or whether the
feedback over these four days is whether this makes sense. It then says if you
are going to exchange information, here are the ones used. By the way if you
are using to give feedback in the process to help refine it, so that in fact it
is used well before the five multiple entities before it actually reaches that
adoption phase. The adoption becomes almost a cleanup rather than the driving
of front edge.
I think the bottom line then and I will be concluding in just a moment is
that we really are talking about how do we do incremental advancement,
iterative incremental advancement, learning from those, and that what was
missing in the first three years of the last phase is we didn’t have a network
that would allow that. The most that could be done were some pairings maybe two
or three entities, but what we saw from our experience with the NHIN a trial
with patients is that doesn’t accomplish the ability to really refine the
standards for nationwide use because what would happen is each of those groups
would get it to work just would be get getting it work in a desperate way
rather than something that is synergistic and allows the clusters to move
There are other activities that are in the HITECH Act that we are charged
with that really go beyond the focus of this group with things having to do
with adoption. The recent NRC report, I think pointed out some very important
things about the functioning of health IT that we need to make sure that we pay
particular attention to and make sure that what we do with the dollars that we
have in the HITECH Act. Two billion dollars is a huge amount of money, but
compared to what it is going to take to change to get all of the adoption, it
is actually still a small part. We don’t expect the government to pay for all
of it. What we need to do is I need to figure out, with staff and with
colleagues across the feds, and with input from the public, is how we get $10
worth of impact for each dollar we spend and how we use it to enable innovation
and creativity. How we use it to make sure that there is real competition among
the products and awareness by providers of what their choices are and
comparisons of those products. Rather than just saying oh it took X dollars for
us to install system up to now or to bring up an HIE in X amount of time, how
do we actually use that as a pilot and say now, you had to do it again if you
had to make it into something that is production ready that you are really
moving forward, how do you drive that down that cost while maintaining your
increasing the quality of the install of the product to speed up the time that
it takes or decrease the time that it takes to install it or for community to
bring up? It may be that that’s by the viral spread by one group that is
successful helping another rather than a new group starting from scratch.
Whatever ways we have we are going to be wanting to get input to trying to
find the best ways to do it so that we really can and using these dollars to do
that, using these dollars to get the most effective spread that we can.
Standards are vital. Once we start going back and trying to put standards in
to things that go out, we need to make sure that the whole system is
upgradeable. I know that there has been concern that if you put out a system,
is it going to become the beta-max? I think that we know enough about how
systems can upgrade and that there are paths forward that we make sure that the
requirements for those systems will allow an evolution of those systems or by
having the ability with standards, prevent vendor lock so that providers can
switch at least ambulatory can switch from one system to another easily without
additional cost or without some long process that we again foster the ability
to make sure and reassure the users that they have a system that will last into
I look forward to the collaborations that we have with NCVHS. I think that
the areas that I will bring up probably tomorrow with the larger committee
where the charges are for E-prescribing in the statute for E-prescribing in
HIPAA, are important areas for us to look at how we collaborate across the
three FACAs. At this point let me stop and open up for questions or comments. I
look forward to the dialogue.
DR. WARREN: First question goes to Jeff.
MR. BLAIR: At the beginning all of the committee members traditionally lined
up indicating if they have any potential conflicts of interest and I was
thinking in terms of either standard development organizations or standard
development processes when I indicated that I wasn’t aware of any conflicts of
interest. But there is one that I do need to mention and that is –- I am
not aware of any conflict of interest with respect to any standard development
organizations or the processes of developing standards, however, I am the
Director of Health Informatics at Lovelace Clinic Foundation. Lovelace Clinic
Foundation is we represent the State of New Mexico and we are one of the nine
states that has the contract with ONC as part of the NHIN. So we are a user of
the standards that are involved in that and have been part of that process.
That may mean that as we go on through these four days of testimony there may
be times when I will need to recuse myself from discussions if it touches on
MR. REYNOLDS: The question I have is as we are going to be listening over
these next number of days, and a lot of us have watched this whole process
mature and really start to deliver functionality, I look at this in a number of
ways. One when you look at standards, some are robust, some have limited
adoption, some are being developed, and some are immature. Then you look at
things like summertime sensitive, because as we are trying to drive towards
certain things so for example if you took E-prescribing now that it is for
example, CMS is going to make some difference in payments going forward and
some people have adopted it then obviously the next set of standards are very
important to have out with that so it is more robust and it does deliver a
total picture. You have time sensitive. You have some that are controversial
because you have lots of people deciding what is the right thing. Some needs
study and some are ready now.
Then you have this whole arena that we are -– and one of the reasons
that we are doing this and we have so many different speakers, is that there
are a lot of people working on this as you know. Some of them have been working
on it for a long time. You can use the X12s and some others as examples. Others
have grouped up around use cases and some of the other things.
Out of this I would love to be able for us to help – people want to grab
this and go. People want to grab it and go before there was money, but they
really want to grab and go now that there is some money on the table so people
get really interested. Trying to map some kind of thinking on how do the people
that have been established and we are most familiar with NCPDP. I am not
selling anything today. I am just using examples that we are very familiar
with. How do they continue which they have done an excellent job of moving
forward in the industry, but how do we map all this to all these factors that
we are talking about and all these processes that we are talking about so that
we can help the entire industry see how to play, see how to engage? Once
everybody understands how to play then I think the game goes a whole lot faster
than it even is right now. That is a thing that I think you guys are going to
be able to help us with.
I hope the other speakers are going to help us with because if we could just
put that together, if we could just put that thinking together so that people
know how to really play and how to make a difference, I think we can go a long
way. That is one of the things I think you could be very helpful as you think
about because I think we are really at a switch point right now. Gone from a
lot of use cases as you mentioned. We got people together and used the word
garage. Well now it’s going on the street and now it’s going around the world
and some other people have been working in garages for a long time. That is the
one thing I think if you can have any comments on that now and then as we go
through this process. The other speakers can help us see this because that is
the thing right now that I think can escalate this to really make a difference
or just maybe create continued confusion for people as to how to play.
Then most importantly the people that want to take it and go with it really
want to see. When you pick a standard, is that real? If it is real, then I can
invest in it. I can engage in it. I can put it out there. That is kind of what
I know I’m trying to think about on this group. If you could help us continue
to think about that, that would be helpful.
DR. KOLODNER: At this point I can’t give you a specific answer and my
technical staff would kill me if I tried. It is really more of a direction that
we have been talking about. If you think about the Internet, it was when we got
a core set of standards that were solid that people could use that they know
might evolve but they were – and they were passed forward. What we are
looking for this December within the constraints law has some specific things
that we need to put in is come out with that core set of standards that is
entities on top of that that will allow them to exchange starting with
something like summary record, but it will have the other necessary components
that the discovery of or matching of patients and other security aspects that
they can now get on the net now. In terms of capacity we may not be open for
business in December, in fact, I would say we probably aren’t going to be open
for business because we still need to get some of the nontechnical things up
and we need to make sure that we have a solid operational plan and some things
that don’t have to do with the standards but has to do with responsibility
operating a network. We have to grow very carefully. As fast as we can to do
that with protection of this area, sensitive important information that we
have, we want to start moving out, but entities we know that they can start
buying products next summer. This has that set of standards and will be good
enough that then allow already decentralized clustering as I keep saying that
they will know what the core is that they need to be back to and with the idea
of being that the real innovation and work occurs having the periphery then as
those are refined that gets brought back. We want to be careful not to have I
groups working on the same thing because again they can optimize to something
that doesn’t work. If they start that way early on they need to come together
so that they become a consortium of certain information.
DR. SUAREZ: Thank you. Thanks Rob and Chuck for coming and thanks for your
leadership on this issue. I think it has been an unprecedented decade where
health information technology and we are about to embark in probably what would
become even more unprecedented one with the leadership of this administration
and your leadership in expanding and really pushing for health IT to be right
in the center of the discussions.
I have several questions around larger issues on health IT adoption. My
question right now focused specifically on the standards side. We are in the
standards subcommittee. It deals with the challenges that we face in the
previous decade what I would call the HIPAA decade in the 1990s when a lot of
the expectations were built around the voluntary adoption of standards and how
the industry will embrace them and the industry was creating them and the
industry was developing them. It took basically an act of Congress to actually
make it happen. After that it took an act of federal act both of the Secretary
to create regulations to establish the adoption and to establish the process by
which entities will in fact adopt and use and be required to do so.
With health IT standards and with the standards related to the exchange of
clinical information through electronic health records, I think we started to
test a new approach with what had been done in the past 3 or 4 years through
the executive order and so the expectation or the understanding that this type
of adoptive standards are accepted and recognized standards will be used by
entities that are receiving funding or doing contract with the federal
In this stimulus package there is quite strong language around the voluntary
adoption of these standards. In fact establishes some language or provides some
language for the fact that nothing in the act requires private entities to in
fact adopt the standards and use them. I’m somewhat concerned about that
because after living through 5 years of working as many of us around here and I
salute every volunteer that had worked on these issues and identifying and
evaluating and recommending a series of organized standards. I see the
challenge of actually seeing it be used. Even though it has been incorporated
in certification criteria for products that will then be certified to be used
by healthcare organizations, it seems that things are left really to the
volunteer adoption of all this. I’m concerned that we are going to spend a lot
of time making the case for the volunteer adoption while we already had an
experience with previous standards. We are actually the day to day operations
of healthcare and those are the administrative transactions that were
standardized by HIPAA that every health organization provider and payor had to
everyday and there were significant benefits noted by the same organizations
that later only had to be required under statutory requirements, federal and
regulatory requirements to adoption.
I am intrigued by some of your earlier comments about the adoption itself
and how the shift in now creating a regulatory process by which there will be
an MPRM and a final rule adopting as it is mentioned in the regulation, in the
stimulus package, the standards. There will be the adoption officially in rule
of this, but there is no requirement. It’s for volunteer acceptance.
I wonder if you could comment on that and if you see any expectation in the
future to push a little harder on the intent of ensuring that that adoption
DR. KOLODNER: There is one section that talks about voluntary adoption by
private entities. I think the point that I would make is every provider doesn’t
have to abide by HIPAA. They only have to if they did this billing in Medicare
as a start, therefore, there are providers who are in fact non-HIPAA covered
entities because they have a cost that’s special. This says that if you don’t
take the money, if you don’t do business with the government, if you don’t do
other things, then you don’t have to use the standards. At $20 billion is a
start is all of those are tied with using the standards in a meaningful use
depending on which flavor of the dollar we are talking about and therefore
those entities will move forward. I think there is reasonable evidence that the
quality of care that we want in treating patients with complex illness can’t be
achieved reliably in a paper-based system then those entities are going to be
ones that we may promote. Here’s somebody who has that kind of health records
and here’s the quality that they have. Again, this is where the health reform
separate from the health IT comes as further stimulus. It isn’t that if you do
work with the government, you are still voluntary. You crossed the threshold at
DR. FRIEDMAN: Thank you Walter for that question and I think many of us who
have read the bill have passed our eyes many times over that paragraph as I’m
sure you have. I will give a response and a compliment, Rob’s, I think, and I
agree with everything he said, at the same time going to a very high level of
approach in this very, very difficult problem. At the end of the day we need to
offer products defined where products include standards that people want to
use. I think many of us had the experience of trying to offer products of
various types that people did not want to use in part because they simply
weren’t up to the level where they saw themselves. I think the challenge to all
of us and I hope many good ideas to this end will come out of these 4 days is
to make everything that is being offered for adoption including standards good
enough that like the Internet itself for many of the purposes for which it is
used, people will want to use it. I’m not saying that the access you introduce
is not important for voluntary versus mandatory access that’s there. We have to
pay attention to it. Wouldn’t it be nice if we could get to the point that I
think we all want to reach where the things we offer have such incontrovertible
value that voluntary versus mandatory is kind of off the table.
DR. WARREN: Mike.
DR. FITZMAURICE: There has been an emphasis on accountable progress in use
cases addressed in the past 3 years and this progress has been successful as
the number of use cases and corresponding interoperability specifications has
grown each year. Of each interoperability specification, however, has
recognized and listed gaps that are still to be addressed. Some data elements
are not well specified. Some SDOs need to be approach to request new standards
or expansion or modification of their existing standards to anticipate that
these gaps will be addressed at HITSP in the coming short term.
DR. KOLODNER: I expect, Mike, I can’t say yes or no. I would expect yes, but
I think what your question and what Harry’s question raised is that there’s
ecosystem out there. It isn’t just HITSP and that as a first step we were
harmonizing the things that were there because there was a lot there and just
harmonizing it got a first step. We are now beginning, not because we covered
all of that, but because some of these areas do require standards to be
developed further that haven’t yet been fully fleshed out that there is this
whole ecosystem and we need to figure out how does that relate to all that we
are doing and what is the impact that we have on those other organizations and
what our dependence on those other organizations? I think it warrants a
discussion to broaden the view and I would be interested in the thoughts that
the subcommittee has or that those who are testified have about what we should
be looking at as we go into this next phase.
DR. FITZMAURICE: It would be advantageous to get $10 of value for every
dollar in spending on health information technology. I find myself wondering
how will ONC pursue getting value for investment? Maybe through choices of the
investment, choices of who will implement the standards and information
exchanges first, choices of priorities of sister agencies and federal
departments, choices of benefits to the taxpayers and others. What I’m leading
into is the question will there be an evaluation component to our ARRA HITECH
DR. FRIEDMAN: I’m going to have to move to another meeting in just a minute.
I have to apologize for excusing myself, but before I go I have to respond as
someone with a long-standing interest in evaluation to say that we absolutely
have to do this and we have to find possibly very innovative ways of conducting
these evaluations. I think we have to evaluate prospectively to the extent that
time allows. We have to study our options before we make decisions and I think
decisions having been made we are going to have to study the consequences of
what we did.
I have a concern that we don’t have the methods available to us that are up
to the level that they need to be to give us the information we need to guide
the decisions that we need to make so any advice that anyone would have for us
about what would be good value of approaches to use would be greatly
appreciated. I understand that evaluation that is useful is very, very hard. We
want to evaluate well, and we are going to need advice and guidance about how
to do that.
MR. BLAIR: Thank you both Rob and Chuck for giving us all this information
and the benefit of your judgment.
DR. WARREN: Our last question will go to Larry.
DR. GREEN: Dr. Kolodner, I want to thank you for coming back and express
admiration for your fidelity and your durability. My question is on the basis
of the last couple of years of work at the NCVHS and the hearings and what was
seen here, I think we’re all really clear that there are very powerful
interests at play here and that the discussion we have heard again this morning
actually demonstrates that yet again. Is there a statement in the plan about
what the problem is 302 million Americans who dare become patients want us to
Is there some galvanizing focus that takes it all the way down to human
beings that we are supposed to be serving with this health IT infrastructure,
this National Health Information Network? Somewhere in the statutes, in the
offices somewhere is there a sentence maybe even two sentences, that says here
is what we want to accomplish for the American patient?
DR. KOLODNER: Larry, thanks for the question. I think it raised both
explicitly and implicitly a couple of things that I think are worth making
explicit. One is the concern that in some ways the process we have had the last
four years is dominated by a particular segment of health and healthcare and
the concern that that wasn’t necessarily patient oriented or patient focused or
patient centered. I think we need as we forward in this next phase, to make
sure that the involvement, that the setting of the standards, that the setting
of targets, that the setting of priorities that reflects all the stakeholders
and is not dominated by one particular stakeholder or stakeholder group and
that we need to look at what we can do to foster the involvement of the other
stakeholders particularly the consumer stakeholder group that sometimes it
doesn’t have the resources or isn’t as well organized as others.
In fact the bill actually has some language about the covering views for
consumer involvement in some of these things.
The strategic plan that is referenced in the bill actually has as its first
goal – I’m trying to bring it up here so that I could just read it. It is
basically enable patient-focused healthcare, enable the transformation to
higher quality more cost-effective patient-focused healthcare through
electronic health information access and used by care providers and by patients
and their needs.
In fact you will find in the language here that it goes beyond healthcare to
health and that the IT itself is an enabler for the health reform. It’s the
kind of statement that you are looking for I would see as being in the health
reform as we go forward and the health IT is focused to enable that
achievement. In the language that we use for health IT, were very explicitly
about this being something that’s about individuals and communities and
improving their health and well being of individuals in communities as its
higher level goal that all of this activity is focused around.
DR. GREEN: I think the answer to the question is no.
DR. KOLODNER: In terms of bringing it down to the specifics of what an
individual will get out of the health IT? Health IT by itself is not enough.
Health IT is a technology. It’s a technology. It can be used for good or bad.
It’s the context that it’s in. Our role is to make sure as government, is not
moved towards something -– we represent the consumer. We represent the
public. The population health is the second part of this. All of the things in
the language are in that direction but the specifics while I can give you
scenarios, we do scenarios as we talk about this. What it means for my mother
who is 93 to make sure that all of her doctors have access to the information
when she wants them to have it so that they aren’t causing the problem that my
brother who is a doctor and I had to detect when one doctor did something that
caused a worsening of a condition by another doctor. That gets down to being
real and we can talk about those as examples.
The health IT is very powerful for making it be something that serves the
individual in the community. That certainly is the motivation that many of us
who have been in this for a long time have taken it. If we haven’t stated it
explicitly and we need to, I would be really appreciative of where you think we
need to state it, how we need to state it, —
DR. FERRER: A good bit of effort has been in the standards world. I think
most of us are concerned with the gap in adoption by multiple providers. Is
your office going to do -– where is the development for that clinical
utility, improvement of the system that we can look –- let the market sort
of develop the best of breeds. We have done that for a number of years. What is
the federal government’s role in maybe serving as a catalyst or enhancing the
capacity of the free market so that when physicians are evaluating electronic
medical records for their offices, they actually have some really good
performing clinical information systems that are not only cost effective, but
they don’t have to sell themselves so hard like we do today. Where are we with
DR. KOLODNER: That is what Chuck was referring to a little bit earlier. I
think the NRC report came out just a couple weeks before the bill or just as
the bill was being introduced. We think that there are some areas of the bill
that are broad enough that they allow us not only to foster the adoption, but
if you look at the extension centers that are meant to model at least as I
understand it and waiting to hear more, after the agriculture extension
centers. They actually are meant to touch the frontline clinicians, feedback
the experience, and get innovation, at that phrase “fostering innovation
and creativity”. That that’s what we are aiming for as we put this out. We
are going to have to learn how to do that, where we can foster that, where the
dollars may be able to go. There are a number of centers here. There is an
information exchange center that NIST has funded. We need to look at those to
make sure that it’s not just exchange, but it is that interface with the
clinicians that makes the exchange possible.
MS. GREENBERG: Thank you, Rob. I very much appreciate your testimony and
hate to follow it with mundane information here, which has nothing to do with
technology but I guess it does have to do with security. Probably those of you
who came in this morning know there have been increased security procedures in
the building. You had to be escorted up to the room, but we also have been
informed that actually if you are to leave this room you must be escorted
wherever you go if you are not a federal employee. Now the members are special
government employees. As long as you have your ID badge and please do carry it
at all times, you will be okay. As I understand it our nonfederal participants
if you wish to go get some coffee or use the restrooms or whatever, we will
need to escort you because you are otherwise at risk of being escorted out of
the building apparently. Now we’ll have to see how this works. Marietta is
available and others as well. I know we have several people here. This is what
we call the buddy system I guess. Thank you for your forbearance.
DR. WARREN: On that happy note we want to thank you, Rob, for your testimony
and we will take a 15-minute break so find your buddy.
DR. WARREN: We are getting ready to start our second panel and we are going
to change our process just a little bit. Dr. Detmer has a plane to catch so
after his presentation, he will take questions on his part and then we will go
to the rest of the panel and go back to our pattern of hearing all the
testimony and then questions. With that Don it’s on to you.
DR. DETMER: Thank you very much. It is great to be back with the NCVHS. I
apologize for a couple of things, one having to interrupt the flow of things
and run off afterwards as well as not having my handout very everybody. You
have it on this and it will also be on the AMIA website.
I want to cover quite a few things and in fact to speak to some of things
that Larry and Rob were talking about near the end of the question and answer
period during Rob’s excellent session. These are the expressed goals of the
hearing as you sent them to me so I feel compelled to do those, but I really
want to go really off of that bullet that says, what should the model be?
Because I want to talk to the model. I think that’s where a lot of this is.
Also I’m not basically a standard’s person. I don’t really have the
personality for it, but I’m a great admirer of the people who are and I
certainly think the topic is important so to add a little more moisture to this
topic that some of us consider at times to be a bit stultifying, I put in some
pictures from my farm in the snow. We also haven’t had much snow so it relates
to that as well.
Actually as has been mentioned, there have been a lot of recent activities
in the HIT standards development area and these are just some of them. I won’t
go through them in great detail, but clearly I think the list itself speaks to
a lot of this. There is HITSP implementation specification, structured product
labels for food and drugs. A lot of promotion which is excellent on the
international adoption of standard terminology from the International Health
Terminology Standards Development Organization, plus more coordination among
ISOC and HL7 and the group I just mentioned, and CDISC. Proactive expansion of
LOINC to include genetic test, newborn screening test, the new version of the
surgeon’s general family health history tool with SNOMED HL7 built in. With
AHIC working group on personalized health standards matrix that detect
conditions. I just came from the American College of Medical Informatics
meeting where I saw some of you, and the whole session was on personalized
healthcare. There is a lot more work to be done in that area and then a lot of
important international standards have also been published by ISO on the EHR
and on privacy and security. The point is there have been both national as well
as a lot of international work.
What we have been talking about we touched on just briefly, is the standard
HIT processes are actually changing. We are shifting from what had been, what I
think more of a bottom-up to an emerging top-down. The point is there are
pluses and minuses with both of these so we asked us to describe the standards
Part of the plus on the bottom-up is that they tend to be well vetted. You
are dealing with experts who really are committed. One of the downsides is you
end up with probably too many standards and nobody says this will be the
standard. It is also slow to achieve global harmonization. One of the major
downsides is having the funding and the site for both maintenance and access.
You may also have limited visions because you are tended to focus by the folks
who show up. The good thing is, the devil you know as opposed to the devil you
don’t know and where that may head. And the limitation of both of these
actually is that idealism can reign over pragmatism, but that probably becomes
a little bigger threat if you work on the top-down side.
On the top-down side actual standards can be set. That’s a major plus on
that. You do have I think hopefully more potential for globalization, certainly
hopefully, for more potential for stable funding on the maintenance and access
Some needed standards may not get considered or may not move forward for
political reasons. You have the wrong cooks stirring the pot. Sometimes you can
have problems plus or minus on the vetting side there and I’m almost certain
there will be unintended consequences just because the massive nature of how
this plays out going forward.
One standards development selection and implementation challenges remain. If
I just talk in terms of HIT, some of these you’ll be hearing more about this
today from looking at the list of people you will be hearing from; decision
support, personalized care, population health support. There has been progress
made on interoperability. Still work in semantic interoperability, ties
SNOWMED-CT to record structures. Clinical knowledge models reflect clinical
best practices. Then of course you have a set of selecting challenges. The
probably most immediate one device terminology and identifiers, but there’s
more coming later.
What should be the model? It says here for development adoption and
implementation of HIT standards for the nation as a whole. At the process
model, clearly Bill Stead will mention this I’m sure this afternoon, label data
at the source. It is hardly something – he is the only that’s talked about it.
I think the fact is the closer we can do these things at the point of get go
the better off we will be so whether it’s LOINC on test kits, outfits from test
devices labeled with LOINC, RxNorm available when a drug is approved and SPL
released. I think those are just one comment about the process.
The point is what model is needed and I added this implementation of HIT and
HCT standards for the nation of a whole. This maybe a little cute but I don’t
think it is. I want you to bear with me and hear me out on this.
I think NCVHS needs to develop an outcomes model that uses complementary HIT
and HCT, health communications technology standards. I think this is going to
be important if we are going to reach desired societal values. The point is
health information and health communications aren’t the same. We need both and
when we just say health IT we tend to not to have another piece of our brain
that tends to light up when certain words are used not tend to light up.
Standards basically are explicit representations that reflect our view of the
world and hence they then relate to what we choose to recognize and value.
Values obviously are the moral attitudes and habits by which we live.
Standards inevitably lead our thinking toward certain measures. That’s part of
what their purpose is. The problem is obviously this is not a trivial linkage.
In Elephant Man, unfortunately I couldn’t find my copy of it on the bus coming
up last night from Charlottesville, but as I recall there’s a dialogue where
the surgeon and elephant man are speaking. The surgeon says, what possible
difference does it make what my standards are? An the elephant man says it is
by your standards we live by. I think that’s why this is really important and I
think the NCVHS needs an outcome model for standards development. I really do
think that right now we are too often ignoring a key consideration. I think we
focus far too disproportionately on standards for IT and ignore far too
much-needed standards for CT.
I’ll give you an example. A difference is a difference if it makes a
difference, is one of the comments that I once heard. In advanced directives
for end of life care we had a discussion breakout group in Florida. HIT end of
life standards would describe templates to accurately express information
regarding personal choices. But HCT standards would assure the currency of
these directives and that current directives are actually delivered to relevant
caregivers in a timely manner. I think the point is it selects a difference
between quality measurement or quality improvement or quality performance
improvement. Otherwise as Renee DuBois said, that which is major drives out
that which is important. That’s a little cynical but it’s not totally off the
The point is what vision does America really want to seek or to achieve for
its people? What road will we follow? I would argue that we really want to be
healthy people living in an altruistic society. An interesting thing where we
focus only on health and the interesting thing about vital statistics in NVCHS,
it does have a health as well as obviously, a healthcare mandate. That is its
50-year plus, 60-year history probably by now, Marjorie.
The point is social determinants of health, nobody knows exactly, but that’s
probably about half of health status, meaningful employment, shelter,
education, safe environments, and then healthcare. That it’s equitable,
efficient, effective, patient-population centered, timely, and safe, are really
I think what we would like to have. This is the road and the question is where
does it head?
I don’t think today and I’m not on a screed here, but I don’t think we are
achieving neither better health nor great altruism with the way we clearly have
rise in unemployment. In fact I think I noticed yesterday that the stock market
is where it was when I chaired NCVHS over a decade ago. Rising loss of homes,
rising numbers for children, rising healthcare cost with rising poor system
performance and comparison international standards, rising imbalance among key
social goods and I just mentioned that last point. Obviously kids need food.
That’s obviously not a trivial matter.
What is required for America to achieve these values for its people? I think
we need standards but standards that create a better balance among a set of
competing social goods, altruism, freedom, healthy individuals, healthy
communities, personal privacy, personal autonomy, and useful knowledge.
How might we do this? I think a new model NCVHS has approved, HIC and HCT
standards, obviously in the context of all this structure that we are talking
about today moving forward. We would be talking about value-driven care that
has value-driven standards as defined by a set of perspectives, as defined by
the patients. They have the capacity to have free access to personal and
altruistic options as defined by the health professional based on relevant,
evolving knowledge as reflected in the report that Bill Stead will be talking
about later this afternoon, and as defined by equitable, safe, efficient,
timely, patient-centered, effective standards as the Chasm Report talked about
HIT and HCT standards would be compatible with care that comes out of the
IOM Education Summit in 2003 that is evidence based, delivered by a team,
patient-centered, continuous improving quality, and utilizes informatics. That
is where the HIT and CT piece comes in to technology fit.
What would be required to achieve these? This model would really enhance
standards across each of these domains of equity, patient-centeredness,
timeliness to match the growing focus that the country is having on efficiency
and safety. It’s not like we weren’t making great strides on some of these
things, but we aren’t necessarily looking at the standards very generally. For
example, do we have standards that will look at under insurance as well as lack
of insurance, as just one example.
I think we need a new model. There’s also today’s dance support for
American’s to choice between altruistic worth versus self or selfish, I’ll
leave to you, goals. Today by a rising number of reports HIPAA as it was,
limits access to useful knowledge through legitimate biomedical and health
research. Examples. Today no simple consent generic procedure allows citizens
to share their personal health data for legitimate scientific research.
Increasingly expensive unfunded mandates keep being enacted without clear
health benefits and they are likely to further hurt biomedical and health
research that is already supported by public funds in NIH. The new regulations
on the identified data and while follow under the minimum necessary data is too
early right now for me to say, but it could have some down sides as well.
Few standards I think explicitly support the solely community benefit and so
whether you look at the AAMC report that I think is discounted as being
surveyed a long time ago, but the more recent AAHC and now a new IOM report, I
think this argument is pretty solidly made.
Now you can see I am starting to bring these together. I think we ought to
think about it that way. What HICT standards development selection and/or
implementation channels just remain? We have some of those that we talked about
earlier of course, but we are not looking at them in a little different
What should be the model? I think the model to really help the nation as a
whole. I think it needs to be an outcome model and it needs to reflect all of
these important reports including some from the international community that
relate to quality, safety, public health, privacy, health information,
communication, technology, and use computer based health records, real health,
and research. There is a whole set of domains that obviously need to be in
consideration. Then I think the processes for developing the relevant HIT and
HICT or HICT standards may better reflect core American values and also create
global markets for products that adopt them. I think this has both practical as
well as important.
The question for you folks is do you need a new working group like the NHII
working group to take on this new robust HICT standards model? That’s obviously
the impertinent thing that I dump on your laps here as I come today.
Who are the players? Who should be involved? I think whoever is hungry plus
whoever else shows up. That’s nice about NCVHS actually. It does try to worry
about that span. Who shows up? I think the fact is -– but you also need to
worry who maybe wanted to show up and couldn’t quite get there.
Obviously these comments I did send on a short notice out to my board to
solicit advice and also tap a couple colleagues internationally. The AMIA board
as well as the staff and I think AMIA is nearly 4,000 other implementations
would rather stand ready to help you in this challenge if you chose to take it
Again, I thank you for the honor of the invitation. I will post a
presentation on AMIA and of course it will obviously be posted here as well,
but if you do want to reach me as well later my email is there. Detmeramia.org.
I’m sorry I spoke so fast. I thought that I actually would be pretty tight on
time, but in any event thank you very much and that essentially concludes my
DR. WARREN: Questions?
DR. FITZMAURICE: The question is as we look into health communication
technology as well as health information standards, what is the role for other
industries? Or maybe I should turn it around, should we look to other industry
for healthcare technology in communication standards? Should we start adopting
more global transport mechanisms of things like wrappers of packages that are
more uniformly based? You can fit a lot of different messages in them. Should
we start looking for a national or international set of XML tags so that we can
start talking the same language with inside these decision support systems?
Should we adopt EDIFACT or something broader? Should we kind of bridge some
gaps between some of the standard developing organizations? We have a lot on
our plate just looking at the US, just looking at healthcare and decision
support. I hear a call for you to be broader in our thinking. Should it go as
broad as I have just described?
DR. DETMER: You said it. I think absolutely. Again, I think it’s the
limitation because if anything in the last 10 years, CT has strolled on the
stage. When I chaired the IOM report in ’89 to ’91 on a computer-based patient
record, the Internet wasn’t –- Bob Kahn and Ben Surf worked on it and
playing around with the concept, but in the last few years particularly
increasingly, I think an awful lot of this is going to be mobile communication
kind of things.
There is both the content domain but there is also the technology domain
that relates to this. I think only if we explicitly were to try to frame it out
as both IT and CT, willl we then have to sort through what might in the right
mix or not? I have actually started to work with Bob Kahn and in fact I think
there’s no question that some of what you are talking about needs to be in this
MR. REYNOLDS: Don, thanks. It’s good to see you again. Tell me a little more
about the word communications in your HCT. How broad is that?
DR. DETMER: I think it has to be obviously the thing about standards they
ultimately have to land on the ground. I think that part of this –- we all
know the difference between a flowery oration and something the person carries
meaning. I think the key thing is information. I think for example, if I looked
at electronic health records, docs really are not interested in records. They
are interested in communications. That’s what they see patients for, to care
about people and try to help them.
They need to have an echo of that communication and a record. The record
clearly has a lot of meanings. I think there was like 150 users and 160 uses
when we looked at it in ’91 and Jeff was part of that. The fact is some of
those are information, but some of those actually are trying to get echoes of
prior communications are important. Again the closer you get to
patient-centered care patient-centeredness is communications. They want the
information and sometimes that’s all they want. If you look at how do people
develop trust in clinical situations, privacy based on data mechanics data is
way down the list. Having somebody who actually can relate to them and that’s
in a communication sense. It’s number one. Having somebody who knows what they
are doing, who is competent is number two. Having somebody that will go in
their corner and fight for them when they need some care and they may have
trouble getting it is the next one. Privacy is number four when it comes up.
The point is I think this communications issue is really number one even
before the information to know what they are doing. I think it is really
important. The distinction as I say can seem subtle, but I think it is quite
DR. WARREN: Marjorie, you are next.
MS. GREENBERG: Don, you never disappoint us. It’s good to have you back. I
guess since I know you are going to be leaving I want to ask if you will
continue to engage with us on this because I am very intrigued. I haven’t been
able to concretize exactly what you are proposing because you are always ahead
of me here. I am very intrigued by what you are saying. Particularly as unlike
you is someone who seems to enjoy standards for some perverse reason though I
love your pictures too.
DR. DETMER: I’m sure you are the normal one. I’m not.
MS. GREENBERG: This idea of value-driven standards is just very fascinating
to me. I think it resonates but a lot of what you talked about we are coming up
on our 60th anniversary of the National Committee and in fact we
started observing it. We are doing a series of things so I would love to engage
you with that and particularly on this topic, which of course it resonates very
recently with the work that Larry and others have been doing about the
patient-centered medical home or actually the National Committee’s focus right
now is on person-centered health. It resonates very much with that.
It also goes back to the work that as you suggested, we did on the NHII and
the 21st Century Vision for Health Statistics, which we are
revisiting also starting later this week, and the influences on health that you
mentioned. All of that. There are a lot of connections. If anyone was sitting
here wondering well what is the National Committee going to do now that there
are two new federal advisory committees – I think we realize we will be busy
enough. I think this is something we can really contribute to that.
I wondered, given your experience with other countries certainly very
directly with the UK, I think the first time you ever came to the National
Committee it was to talk to us about international experience before you were a
member, before you were the chair. Are there other examples of countries or
other places where you feel they are taking this kind of vision or approach
better than or in a way that maybe we aren’t or that we could emulate?
MR. DETMER: Not in these terms. I think the talk I have given has been very
US kind of grounded. On the other hand I think the conversation would resonate
everywhere. In some respects it is interesting because we are a very consumer
driven kind of culture. I compare it certainly to Southern Europe, the
Mediterranean for example. Northern Europe, I think is more like the US in that
context. I think some of this would seem pretty unusual. On the other hand, I
think that as I say I think there are clearly dimensions to that that are also
echoing internationally as well.
Relative to your question about my own availability, as you might get a
sense of I am eager to see a little more of my farm. My wife has been leaning
on me a bit so I plan to spend a little more time there starting after July,
but obviously I am deeply interested in these issues and Washington is not far
down the road. Definitely I would be delighted to keep it up. Like all of this
it always remains the work in progress. I don’t care what we have said or
worked at. This is never going to be a done thing and what I’m saying is what
we just need to try to make sure is that we don’t underestimate really what we
are in, the business we are in when we are talking about standards. That’s the
echo I was really trying.
MS. GREENBERG: It really reminded me of that -– if you remember Kathy
Colton and Lisa Izony’s camera image. The main thing is what are the questions
we want answered? What do we want to communicate? In case you think you might
be escaping to us by mentioning the farm, we are talking about an event in
Charlottesville with former chairs. We’ll find you there.
MR. BLAIR: I’m just echoing everybody else’s appreciation for your coming. I
have this engineering personality and I listened to what you say and I figure
how do I build it or how do we build it, the whole healthcare informatics
community? When I hear the word outcomes, I’m assuming that you are focusing on
patient outcomes and the improvement in patient outcomes. Am I on target on
DR. DETMER: I’m not so much maybe saying that. I think what I really want to
do is pull us out of just talking about processes as though we really thought
about the universe that mattered. I think we really need to pull back and think
about what are the ends in view that we really have in mind with our standards.
I don’t know how that fits into an engineer’s topography.
MR. BLAIR: I was ready for the alternative. The other outcomes that I think
of are the outcomes from a labor report, a radiology report. You can tell I’m
struggling to understand what outcomes you are referring to. Is that closer to
the outcomes or am I completely off the mark?
DR. DETMER: I think we probably need standards that relate to variables.
What you are talking about are people who have made it to getting a lab test
and a result. I’m saying we really need to be looking at this from a whole
societal perspective. I think we are adding according to Commonwealth what
14,000 uninsured a day now. The point is that we have to look at the all those
that are hungry not just those that show up for the food. I think that is part
of what I am getting at. It’s not like what we are doing is wrong. I just think
it’s insufficient and I think we need to really be looking –- essentially
this is the national information infrastructure and I think I would like to say
communications venue for really talking about these highly important kinds of
very complex and very important interfaces and interactions.
MR. BLAIR: My last piece is if I understand what you say then healthcare
communications, the HCT, actually builds upon the HIT. Doesn’t it?
DR. DETMER: Sure. As I said and at times they maybe one and the same, but I
don’t think they are the same. As I say I really do think patients want
communications. Yes, they want information. They go to the web and get some,
but they really want to communicate. I think that if we don’t talk about
standards that let that be timely and let those work, we are missing the verbs
of all of this. We are missing the action. I think informatics is more about
the action side of this than I think it typically is a more probably about the
communication side than it is purely the IT side.
MR. BLAIR: Thank you.
MR. DETMER: I hope that helped.
MR. BLAIR: Yes, helped Jeffery think outside of the box a little more.
DR. FERRER: With regards to the task with NCVHS approves, would you see
standards perhaps expanding its scope a little bit broader as opposed to
starting a new work group which will require a considerable amount of learned
lessons and so forth as opposed to there is a standard governance. Would you be
comfortable in also maybe contemplating that and what task would you initially
give them like Marjorie said, as you have authored and pushed this forward?
Would you be willing to serve? Getting the people in the room probably is not
going to be the issue, but getting them in on Capitol Hill(?) and the
DR. DETMER: Your question reminds me one time when I was testifying at Ways
and Means and one of the congressman said are you trying to tell us and I said
sir, you are the one that got elected to represent the country. I’m not on the
committee. You have to decide what you want to do with it. I think you probably
need elements of both personally. I think the NHII working group really did
help us think outside the box, but ultimately it became the box in a way. All
I’m saying is but it wasn’t all the box we needed. I think that was the point.
One of my mia culpas when I was working on that is that I didn’t have research
in education in the original model of the NHII. Well I was totally colossally
dumb. At the time though I thought wow there is so much even if we just try to
get around the care space that it could be confusing and such. At least that
was my excuse later. I think the point is though research and education are
hugely relevant. That’s a big piece of communications too. I guess my point is
yes I think you would find some of your current structures would allow you to
put the C into the IT, but I think also this is a fairly substantial mandate if
you were going to take it on that I also think you probably need to be doing
some of that kind of thinking through a working group.
DR. WARREN: Jim.
PARTICIPANT: We have been looking at some of the issues involved and just
representing knowledge in health information technology standards. There is
always this box that you build for yourself and you never get it all. For any
given use case I have seen there has always been the need to send the note
along with what is in the database. We have already now I suppose we have
technology that you could document the phone call and digitize the phone
conversation. But then the question comes how do you search these things now
that contain both this very unstructured data and stuff which you can basically
run like well done standard work that is widely supported like in an industries
like -– queries or objective-oriented queries. I’m aware of research work
that is going on in departments of information systems trying to tie the two
together, information retrieval and database queries.
In terms of operation ones in this, where are we in being able to query a
database? In essence if you wanted to reconstruct this for patient care
purposes, are we talking about developing software like Oracle databases or IBM
databases under systems that would be needed to support it?
DR. DETMER: I’m going to punt that to Bill this afternoon mostly because I
think he actually is trying to engage that kind of question as he tries to meet
the spirit of the report that just came out at NRC. It basically says again
this is going to be an iterative sort of thing. I do say as you were talking it
did bring me back a little bit to Michael Fitzmaurice’s question at the start
and may well be some other communities that could well change the dialogue a
bit as they got engaged in this. It is way beyond my expertise. I know it just
enough to see it as a real issue. No question about that. It’s a good question.
DR. WARREN: Our last question for you will be by Rob.
DR. KOLODNER: A couple things. One is we did have conference with our EU
college last year. In the conversation it looked like they were talking about
ICT and that at least in terms of what we are talking about it was the same
thing. We were using health IT, but in fact a lot of the components had the C
in it probably not as much as where we push it. That is part of the discovery
and the peeling of the onion and there are things that you say you would like
to include. This is iterative and each time where we get there you are able to
see a little farther and add the next component. The sense to have is that we
have some C in our health IT and it is a matter of making that more explicit
and bringing that forward. The issue of the role of NCVHS in face of factors is
one that I think we’ll be working with bringing that up tomorrow with the
larger committee and I think that there are certainly roles especially given
the NCVHS was there before. It will be there after. It’s kind of the long view.
While you charged NCVHS with taking this on, we have this very fast activity
that is being charged to the FACAs. The question is how would you parch that
out. How would you make sure that something that’s mature enough –
DR. DETMER: The thing is maddening. Obviously we now have resources in this
country on a scale that we have never dreamed of having to put to some of these
things. On the other hand the kinds of trajectories toward the future that we
are talking about these are very serious immutable social problems. I think we
can’t get so caught up with the pile immediately in front of our door that we
are losing track and also keeping it from getting out the back door. I don’t
know where that metaphor is headed.
The point is I think maybe getting trapped and I think the point I guess I
would make is for example I really feel like it’s really critical. One of the
things 11, 12, 13 national educational organizations, AAMC, ACN, the
pharmacists, the dentists, the psychologists, the osteopaths are working
together with AMIA to try to in fact create an educational structure that would
be interprofessional for team-based learning focused on continuous quality
improvement, timely patient-oriented care, using informatics. Essentially if we
can do that whether they will all be learning together, but if we have a
competency based curriculum that says I have heard of that. I understand it. I
can do it. You know those levels of confidence. There will be timelines of
these professional educations or such that if we don’t start that now, we will
be forever in remedial education trying to keep up with how do we ever change
the dialogue. I think we are going to have to tap our head and rub our tummy at
the time. I don’t think we have any alternative but that.
I do come back to the C in the IT. I think this is not –- S.I. Hayakawa
Language in Thought and Action. It was a pivotal book in my own intellectual
kind of development. He talked about directive language. It is directive
languages is language which actually directs your thinking. I think the fact is
HICT tends to direct your thinking in a different direction and it goes with
just HIT. I admire and I admit that you clearly are looking at both of those
things, but I think it now with Internet I think we were mostly indirective.
Fifty years ago it was vital statistics recession and we tried to change the
name of the committee and we thought oh my goodness what will happen if we try
to do that. I’m happy with NCVHS. What I think the fact is as you say and as
Jefferson noted from Virginia. I got to make a Jefferson quote. The past always
has to give way to the future. I think that is where we seem to be.
I’m glad to hear actually because I wasn’t aware at the EU level that some
of that was as visible as it was. I clearly interact and mostly I would say
with my colleagues. In fact I have a paper coming out from the Nuffel(?) Trust
this week, that is a review of what we think we have learned from this whole
space of electronic health records over the last few years. It does speak to
this to some extent. I do think it’s in the dialogue out there. Anyway a bit of
a rambling response.
DR. WARREN: Thank you, Don. John, you are up next. While John is pulling his
slides up, I did want to let everyone know that Dr. Kolodner will be spending
the day with us and hearing all of the testimony and then a clarification, he
had mentioned his strategic plan. For people who want to look at his strategic
plan it is on his website for ONC. It is right as you come on and it is almost
in flashing lights, up on the right hand side which means I have been there far
too often. Whenever you are ready, John.
DR. TOOKER: Good morning and thanks. It is nice to be back testifying before
you again. Let me put this a little bit in context. I do really look forward to
the conversation that we are going to have once I finish my remark.
It is a privilege to be here. My perspective in being here really is not as
an expert in either health information technology or medical informatics and
the relationship between the two. That obviously will become apparent as we go
through this conversation. Where my grounding is in all of this work is really
my role with American College of Physicians. We have established policy in all
of these areas that we are talking about today from a healthcare delivery to
health information technology to quality and safety and particularly around
models appear that are patient-centered care. In looking at Larry and of course
as Steve Waldren who is your next panel we have all been working together to
try to develop a realized environment out there to develop a delivery system
that will improve the quality of care. In our opinion the adoption of health
information technology not just the adoption the development of the necessary
standards, but even on top of the necessary standards and specifications and
implementations of those standards is solving other real documentation problems
on the ground.
Towards that end I have been very active to be involved in the nationally
health collaborative. Quite briefly, you may recall that the AHIC, America’s
Health Information Community, had the foresight to develop its successor
organizations so that there would not be an interruption in the work they had
worked so hard to achieve. And so Brookings and LMI were charged with
developing that successor organization which we affectionately called the
AHIC’s successor Inc for a long, long time.
I was privileged to chair the governance committee for the new organization
and along with John Glasser who I think you know and with John Perlin who I
think you know, were the other two incorporators for that organization. We
worked very closely not only with the Brookings LMI but of course with the
Office of National Coordinator and Rob was enormously helpful along with Ellie
Cronan and Mary Joe Deering throughout this entire process.
The NeHC as we now call it. The National eHealth Collaborative was formerly
established in July of last year. Between July and November 12th,
the organization was fleshed out in my detail including the development of a
board. On November 12th we presented our report to then Secretary
Leavitt, at the final AHIC meeting. The next day on November 13th,
this new organization began. We are relatively young, starting in November, and
taking this up now to where we are today.
The operative aspects of this organization are all about connecting health,
connecting information, and improving health. Obviously we have stolen from
every other organization to find the right name for this organization because
it really is as Don talked about. How do you communicate among the necessary
parties in order to deliver effective healthcare? From the very beginning it
was apparent that all of the stakeholders had to be engaged in this
public/private process. I want to emphasize that the National eHealth
Collaborative is not a federal body rather it is a voluntary consensus
standards body similar to the National Quality Reform. It has a broad
stakeholder representation that I will get to in a moment, but the emphasis
from the very beginning has been to recognize the importance and as Rob
mentioned earlier today, the ecosystem that we are all in. From the ecosystem
we need to derive the stimuli that will drive our work. While it is very
important to have advice from higher authorities, it is really those people on
the ground that are most affected by the healthcare that need to help drive the
direction of this organization.
We have been working very closely with the Office of the National
Coordinator as I have mentioned. As the board is currently constructed, the
Office of the National Coordinator is a federal liaison through that board. In
addition to that there are two other federal liaisons and Secretary Leavitt
himself was one of the members of the NeHC not as member of the board, but as
one of the liaisons from the federal government, and in addition to that,
Secretary Peake from the VA. The point is very high-level representation at the
level of federal government.
We have been asked why the federal government could not be members of the
board itself. It turns out that the direction from the Office of the General
Counsel and Health and Human Services is that the federal government cannot
occupy a fiduciary seat on the board, but rather they have to serve in a
liaison capacity which maybe not news to you. I would say that it took a little
while for us to get to the point where we were for sure that that’s what the
direction would be.
In addition to the National eHealth Collaborative we have a very close
relationship with both HITSP as well as with the Certification Commission for
Health Information Technology, CCHIT. You may have noted that in February that
we published a common road map in business statements with HITSP and with CCHIT
that made public what we think our working relationships ought to be. Then as
already been mentioned today is the importance of the National Health
Information, a collaborative and the NeHC serves on the work group for the
government structure for the –-
I want to emphasize the importance of bringing together all of these
stakeholders and there are a lot of them. There are what we call healthcare
providers and that’s a compromised term because most physicians would rather be
referred to as doctors and providers. We are talking about all people and all
institutions that provide the healthcare.
I need to particularly focus on the importance of consumers now. It was
recognized early on that consumers are both desired and needed to be recognized
in the development of National eHealth Collaborative in all of the work groups.
There were three at the covenant’s work group and the sustainability work group
and the membership’s work group. The consumer representation was substantial.
The only two seats on the board that are slotted seats by design are for
consumers. We have Steve Finley from the consumers union and Michael Lardiere
from the National Association of Community Health Centers has members of that
board. Their perspectives have been very important. In addition to that
industry employers, other payors, quality improvement organizations,
By government now I think we are realizing more and more that in addition to
the clinical importance of the federal government the state governments are
really latching on to the adoption of health information technology and that is
where we are seeing a lot of the barriers identified that they need to have
solved. The eHealth alliance with the National Government Association, the NGA
met earlier this week, has been very vocal and testified before our last
meeting about that they need help in overcoming these barriers to
implementation of electronic health records at the state level. Although they
are challenged financially it is a remarkable laboratory when you have lots of
states that are innovating all at the same time in making those innovations
public for now.
I mentioned the government population health research and academic health
centers and of course the vendors. We think the market is very important to
driving work of the National eHealth Collaborative.
How is that done? Well you know a lot about the use cases that were derived
from America’s health information community. Building those use cases we are
talking now primarily about value cases now. Value cases recognize as others
have already said here today that we need to understand what values should be
derived from the development of the use case that will impact those
stakeholders who need to have their problems solved. The value case concept
encompasses the concept of healthcare communities. This is a real
public/private partnership, which together identify the opportunities to
accelerate the adoption and use of interoperable information technology. There
aren’t 10 of these. There aren’t a hundred. There are thousands of those
possibilities out there that will be a challenge to prioritize and assign
resources too. Important of their tentative facilitative course information
exchange. As Don said when we were talking about patients, you can’t neglect
the importance of communication. I just want to give a real life example of
what this may mean because one of the challenges I recognize in more and more
times in these forms that I start to talk more like the forms rather than what
I really talk like as a physician now.
If you take the Medicare population now and that is challenging enough let
alone the enormous number of patients who are either underinsured or uninsured
right now. The underinsured are becoming more and more important because they
are withholding healthcare because they can’t afford copays and others. These
problems are mounting out there.
If you take a typical Medicare patient that Steve and Larry might see, 25
percent of those patients have five or more chronic conditions now. They see on
average 14 physicians. They have over 40 office visits a year. They are
hospitalized multiple times and are on multiple medications and you can imagine
the complexity of health information exchange to take care of a single patient.
In a practice that has 2,000 patients and understand what a single physician in
a single practice the barriers that need to be overcome for health IT adoption.
On average for every visit 70 percent of the time patients will at least have
medications that they need to be reconciled or changed now. It just takes time.
We know in terms of care coordination that the most common problem that
leads to either missed opportunities to improve healthcare or as importantly
leads to issues of patient safety for patients that on average that medication
and reconciliation is the most important challenge for us at this point and
time. It is one that is relatively soluble if we can develop not only the
standards and specifications in electronic health records, but the
implementation processes that allow for this to be resolved now.
In addition to the patient safety issues we want to talk about how to
decrease the systems integration cost. If you think about the Recovery Act,
this enormous push toward adoption of electronic health record, one of our
concerns is how to affectively adopt an electronic health records many of which
will not be very interoperable with wish to adoption now. Quality and produce
concept care for patients and to general meaningful value propositions to
healthcare stakeholders. Hence the bottom-up approach that I was talking about.
What would be the ordinary goals of a value case, a broad stakeholder input?
I might go through a short diagram in a minute in which I will highlight that.
As important as it is to identify the valued case, it is to identify how are
you going to solve the problem with that value case represent? These are the
breakthrough strategies that need to be part of any value case strategy. Of
course in a value case proposition or value case the scenario the value of the
proposition needs to be understood by all of those who are participating in
solving that particular problem.
There are multiple types of value cases that are currently in vision. One is
more traditional that is around standards harmonization which was the focus of
AHIC and the value cases, the use cases that AHIC developed and of course with
NCVHS now. There are also other models of processes of care. What would be the
model of the ideal care coordination process? As we have done the work to
develop the medical home, we realize that the largest single barrier to
improving healthcare is the coordination of care or even the single patient
among all the healthcare entities that are providing care for that patient.
Best practices. For example we now know that there is a mandate to adopt
describing starting with the perhaps a carrot, but increasingly becoming a
stick. How do you identify the best practices so the practices are able to more
rapidly adopt and not just because they are forced to because they will see the
opportunity to improve not only healthcare, but the stability to practice as
well, frameworks such as with health information exchanges. There are many
types of opportunities that are available.
A typical value case prioritization process and I’m not going to go through
this in detail is that the strategy for all value cases will be developed by
the National eHealth Collaborative and is currently a value case steering
committee that is chaired by Paul Tang that you all know well for NeHC. Once
the strategy for the value case or prioritization process is established then
there will be a public call for value cases. This means that the public at
large representing all the stakeholders will have the opportunity to represent
themselves as having particular problems that they need to have resolved. Value
cases would be then submitted and there is the steering committee would oversee
the submission process as well as the bedding of these value cases in
soliciting public input and getting to a point where there is a go or no go
decision as to whether or not this is a value case that should go forward. The
value case might require an establish of new standards. It might require
improved use of existing standards and it might require no standards at all
depending on what the nature of value case is that would need to be solved.
Ultimately once the value case goes forward then of course it will be assigned
to wherever it needs to go. In the case of standards of course to HITSP and
then leading of course to the development of the specifications for
certification into electronic health record.
Going forward I think that everybody has been consumed by the stimulus bill
known as the Recovery Act. You have already heard from Rob in considerable
detail about the importance of this act and from the Office of National
Coordinators point of view a how that money might well be spent. I have been
impressed with the quality of the oversight that ONC is thinking about
providing to make sure that those monies are spent wisely and fairly. I would
say that everybody is following the trail right now. Where is the money going
Importantly for us as National eHealth Collaborative is the section in the
legislation that delineates the two federal advisory committees. In different
parts of the legislation there is different language that reflects on the work
of the National eHealth Collaborative. One is that the Secretary may designate
the National eHealth Collaborative to be either the standards body or the
policy body, but assuming that that is the case then the National eHealth
Collaborative will have to change to become a federal advisory committee and
what that means in part or in whole we don’t know yet. We are looking forward
to having that interpretation. We are also looking at the likelihood while
wanting to cooperate in every way we can with the federal government, the
National eHealth Collaborative is not necessarily dependent upon being a
federal advisory committee or not. We will know more about that presumably in
the next few months, but we are developing our own options for moving forward
because the business sustainability model is not dependant upon being a federal
Also interesting in the language is the way that the two committees were
described. In the case of the policy committee the language was quite
descriptive as to the membership and how it would be determined. Whereas in the
case of the standards committee the language is much more permissive and the
make up of that standards, a body looks very much like what the National
eHealth Collaborative is right now and even provides for revenue to the
standards body due to sources. The model that we are currently envisioning of
course is in part a membership dues model and at our last meeting we approved
the 10 organizations that wish to belong to this organization.
What will be the next step? We have a very strong relationship that I have
already talked about with ONC, HITSP, CCHIT, and the developing NHIN, the NHIN
collaborative now. We intend to fully sustain those relationships and build on
them as we now have.
Being stakeholders across the healthcare committee because we have such a
broad stakeholder base, it is relatively straightforward for this organization
to convene stakeholders. They both understand from their points of view what is
needed in terms of problems to be solved, but also to engage them in the
development of the value cases I talked about as well and to foster the
public/private libraries that we are talking about here and working within the
plan of the federal government to help identify and the sets of priorities that
are related to interoperability standards and initiatives as working with HITSP
and CCHIT to do that. Where there are nonstandards based areas to accelerate
progress we expect to identify and to address those as well.
My last point is going to be about the importance of coordination. It seems
to us that there are two key areas of coordination in which this organization
can be particularly helpful. One is the coordination among stakeholders
themselves to have an important national form where all the stakeholders for
health information technology and the related medical informatics can come
together in a common place. The National eHealth Collaborative can provide
Two is to assist in the coordination between all of these entities that have
been involved in health information technology. Again with the leverage that we
have in terms of those who are currently serving in access to those it would be
helpful. We think that we can play a meaningful role in the coordination among
organizations seeking to address these problems as well.
Any questions please let us know. The website is the nationalehealth.org and
of course I’m available to you. I just wanted to mention that the last time I
was here talking in a generalist’s tone it was about the patient-centered
medical home. While it may seem odd to be talking about that the last time I
was here and now the National eHealth Collaborative at this point in time. It
really is all about how we are solving these problems to address the important
needs of patients. With that madam chair I will close and thank you.
DR. WARREN: With that we will move to Carol Diamond and then after she is
through, we will have questions for both of you.
DR. DIAMOND: Thank you. My name is Carol Diamond and I am the managing
director with the Markle Foundation and I’m also chair of Connecting for
Health, Markle’s public/private initiative in health IT.
I don’t have slides today but my remarks are available to the committee and
so is the article that was referenced earlier by Dr. Kolodner. Those materials
are available. I’m going in my comments really riff off. I think the question
that Larry teed up earlier in the widening of the aperture that Don’s remarks
made us all think about because I’m very much in the same place.
First and foremost the enactment of the American Recovery and Reinvestment
Act portends profound changes to healthcare and health IT. There’s just no two
ways about. I really appreciate the opportunity to address this committee on
this topic at this critical moment.
These are unprecedented public investments and they have to change the way
we think about health IT and about achieving the goals of the new law. For
years our public policy in health IT has really been about trying to motivate
and stimulate voluntary and consensus based action in the private sector. Today
is a result of massive public sector investment in IT. We find ourselves in a
different environment in which government must use these investments to respond
to dire economic crisis we face, to deploy these investments to ramp up a
nascent health IT sector, and to really live up to the public demand for
accountability that ensures that this large taxpayer infusion achieves the
health quality improvement and cost reduction objections of the new law. In
short we have to get health IT right. Now that the legislative battles are
settle the successor failure will all boil down to how well we can implement
The federal government and health IT really changes the landscape and we all
need to open our minds to new opportunities rather than searching for ways to
preserve the old. These times call for new thinking. Health IT has now become
an area of significant public investment. I bet if I asked each of you in
October if you would be willing to wager money on whether or not there would be
$20 billion invested in health IT and federal legislation by February, none of
you would have won the bet including me. I think what this means is we have
essentially doubled the size of an industry by federal tax dollars and we
cannot meet our new and more urgent challenges if we operate under old
assumptions and approaches.
This kind of spending if done wrong can also have the negative mark of
consequences of interfering with rapid innovation by locking into these
processes or technologies which although well intended came about without a
systemic view and we risk locking out the very innovations we need for
meaningful health information sharing to support better decisions. I think
that’s really what Don teed up as well. This is about supporting better
decisions. It is about communications that support those decisions. That’s
really what health IT has to do.
Our goal must be to achieve real health improvements, create jobs, and
reduce unnecessary costs. It should not be primarily the creation of standards
or the certification of software. Rather standards and certification should
support measurable health improvements. Health improvements are not achieved by
installation of software. They are achieved through the effective use of
information for better decision making.
Effective use is what keeps the patient from suffering a medical error. It
helps the doctor prescribe the right treatment at the right time. It allows a
caring team to provide the best possible care in the most cost-effective way.
These are the expectations we have for health IT. Make no mistake the American
public will not cheer for all we have done with 19 billion as installed
software. There’s either unused or siloed so that it cannot be used to improve
health. The public will not understand the importance of this investment. If
administrative costs aren’t cut or knowledge in use of best practices is not
improved and they will not support this investment of patients and consumers
aren’t more engaged in their own healthcare or if they worry about whether
their information will be kept private and secure.
In fact we risk the same crisis of confidence that have plagued so many
other government investments in IT. The risk now is that we can make standards
progress on paper and we can stand up certification regiments who assess
compliance with those standards and it will move us very little toward the real
objectives. The real objectives are to improve healthcare quality, to help
contain costs, and to create trust across health information networks.
I have five points to make and I will talk first about objectives and then
how about under the new authority and provisions of the Recovery Act the
government can help insure that standards and certification support these
The first point is really about setting objectives. The right measure of
success is not how many physicians you certify health information technology.
It’s whether outcomes are better. This measure of success can only be achieved
if physicians and other caregivers are using the right health IT that enables
them to improve outcomes. Are there better health outcomes as a result of
improved adherence with medications that are proven to be cost effective? Is
there a reduction in unnecessary or avoidable hospital readmissions? Is blood
pressure better controlled? Are there fewer unnecessary medication errors?
These are the kinds of questions we should be asking health IT. The basic
criteria for setting such targets should be to boost cost effectiveness,
improve health, and quality of life, and reduce cost in the long run. I think
the recovery outlays these requirements out clear in a statement of purpose.
The government is both the major funder and investor in health system we
have today. This gives it enormous leverage, but that leverage can be wasted if
the government sees its role as dictating the adoption of particular tools or
applications. Rather it should identify high value health improvement goals
against which IT hardware and software enterprises should compete and make its
health IT investments in ways that can improve health sector performance. We
definitely need technology to support these goals. There’s no question about it
and we need it to help measure these goals and we need to facilitate rapid
learning and feedback loops so that we can continuously improve.
The electronic health records purchase with these stimulus funds by
hospitals and doctors need to have these requirements to find up front. We
can’t deploy the technology and then thinking about the requirements. This is
true for lots of issues.
An example of how this can work. Consider New York City’s efforts to
encourage small practice providers to adopt electronic health records. The city
focused on a thousand primary care providers in hospitals, community health
centers, and more than a hundred small practices caring for more than one
million patients. This project succeeded by setting clear health objectives at
the onset. The entire process focused on using technology to make concrete
improvements against population health and prevention goals. During the
software procurement they selected EHR vendor and enhanced his preventive care
functionality to meet the city’s health goals. In addition providers
incorporated new workloads as a condition for participating in this program by
adopting shared public health goals and quality improvement activities
including decision support tools.
Second point, and this one is on information sharing and standards. What is
needed now? These kinds of goals will be critically dependent on the private
and secure information sharing and the industry will need to achieve the
ability to make such sharing possible. A few successful examples of health
information sharing we have today come about when implementers set an objective
and then deploy technology to achieve it. Sometimes by using a software
application and sometimes by just moving critical information to the user.
When the government built the federal interstate highway system, it did not
set the standards for cars and trucks just roads and similarly with the basic
set of standards that enables the sharing of critical information among
authorized users, we can lay the groundwork for innovative products and
services in a competitive market that meets these requirements by getting
robust technology in the hands of providers and patients.
What about standards? The truth about standards is that in large complex
environments they are not created. They are adopted. Standards are adopted in
several stages, evolving as users find more reasons to need them. We need to
start with just a basic set of standards for interoperability using the
Internet and other standards will evolve as products and software using them
provide value and mature.
In health we face a situation where information sharing is the exception not
the rule. Most systems today lack basic connectivity to the most fundamental
pieces of information that a provider critically needs. I think John alluded to
this in his remarks, lab results, medication history, access to the care
summary of the last physician who saw the patient or the provider who saw the
patient. It is always tempting to list all the current problems and to
provision a single massive solution, but past evidence from projects like FBI’s
virtual case file, FEA’s air traffic control modernization, or IRS’s electronic
for a detection system show that it attempts to provision sudden massive
upgrades are higher risk ventures that often fail to deliver these imaginative
A particularly relevant example is the Government Open Systems
Interconnection Profile called GOSIP. It mandated the adoption of a particular
form of networking technology that would replace poor Internet technologies. It
became a federal standard and despite being federally mandated, internally
consistent, conforming and interoperable, GOSIP was largely a failure because
few organizations in the field actually adopted it. After 10 years of work the
standard was rewritten to allow the use of standard Internet protocols. We
cannot afford that fate be fall error.
In order to succeed IT procurement and development of standards has to be
incremental. It has to be incremental. It has to concentrate on the most basic
issues first which is making it easier to share data in its current form while
protecting privacy and security. Standards for moving information between
authorized caregivers are the most basic problem for two reasons. First sharing
information is the key element for improving basic health outcomes. Second
improving the ability to share helps create the motivation for improving
underlying data quality and adopting more standards, but critically the
converse is not true. Standardized data creates no incentive to share.
The initial focus should be on only a critical set of standards for data
sharing, the way it moves from point A to point B over the Internet. This would
involve initially specifying interface, transport, and security standards
rather than standards for all data expression or the behavior of local
applications because the critical predictor of good outcomes and cost
effectiveness is whether or not data can move between a person’s various
By concentrating on interfaces, transport, and security a network of
participants will be able to interoperate while preserving significant space
for innovation and matching EHR applications to a wide variety of needs. This
is the same recommendation we made in 2004 connecting for health roadmap. It
remains the case that the majority of the providers today do not have EHRs that
most data being exchanged even with the enactment of the Recovery Act will
inevitably be unstructured. I think the Recovery Act sets forth standards and
policy advisory groups that will under FACA rules aim to focus on standards
needed for the most critically uses of health information directed against the
health objectives. These bodies can facilitate rapid investments.
Commerce has also foreseen a role for the National Institute for Standards
and Technology and testing standards and also advisory role for NCVHS. These
are all very important opportunities. The processes we have had in place in the
past like HITSP can be integrated into this new set of requirements. They are
nascent enough that they can adapt and contribute along with many important
bodies that have contributed to standards and certification in other areas of
Once the health objectives are set our focus on standards should be on
adoption. Many standards exist that can be deployed to serve the end of
innovative improvement in health outcomes. It is ultimately the users who
ratify standards through use.
History is ripe with examples where proposed and ratified standards were
never implemented because the users didn’t adopt them. In the appendix to my
testimony there are several examples from other sectors for those who are
The web took off when innovators like those at University of Illinois built
a browser through their own innovation to meet the objective of making the
Internet a platform for sharing information for communicating not because the
government specified standards and certified the browser. Utility of PCs took
off when Microsoft innovated and created an operating system that met its goals
of individual users having computing power at their desk to share information
and collaborate now in the government ratified standards.
My third point. There is a new urgency, urgency for results. We must focus
on moving information when and where it’s needed to those who are authorized to
receive it. If we are going to learn history, incremental standards that solve
one problem well are better than complex standards that are constructed to
solve all problems at the same time.
Such complex standards are in general hard to implement by the average
health IT department. I provide examples in the appendix also of soap and XHTML
in the appendix for those who are interested. They are very instructive stories
to look at.
It is far better to have healthcare institutions sharing data that is not
perfect than perfecting data that is never shared. This means we have to orient
our thinking differently from a hope that idealizing expressions of health data
will solve the use case for sharing into a much narrower but more tractable
problem of addressing sharing itself, taking advantage of the subsequent
impetus that it will provide to improve data quality and adopt more and deeper
I think this is a very different approach than the road we have been down,
but to be effective high-tech standards adoption must happen this way. The
sooner organizations see a reason to share information the sooner they are
likely to embrace standards that make the exchange of data easier.
Let’s create the reason to share by setting up clear health improvement
goals for the use of IT and focus on good enough standards to transport
information initially so that rapid adoption in innovation cycles can take
place in this new market place.
Fourth point is that a new bar has been set. Technology will have to
implement new privacy policies. Trust is another critical aspect of information
sharing. Information policies that protect information are essential to trust.
I have testified many times before this committee on that topic. These policies
must influence the implementation of standards and the selection of technology.
As I like to say standards don’t push the send button. People who trust each
Markle’s connecting for health common framework articulates the need for
comprehensive set of privacy policies that should be implemented together. This
set of policies has been articulated in the new law including audit and
consumer control, unbreached notification, and several others. The law also
requires that health IT policy committee operating under FACA rules to further
specify policies that protect privacy and security in the implementation. These
new criteria are critical to getting health IT right. Without them choices
among technical standards and architecture make de facto decisions that
determine how personal health information is discovered, how it will be
accessed, and where it will be stored and aggregated.
The new committees both the policy and standards committees are
opportunities to marry policy and technology implementation more closely than
we have in the past. As a Markle Foundation is connecting for health
representatives to emphasize before many federal bodies for many years now,
policy and technology must be addressed in tandem. New oversight and
accountability mechanisms will require these privacy policies to be implemented
in the technologies that get funded with taxpayer money.
Last point. Accountability for results. The administration has promised
unprecedented accountability and transparency in the spending of the stimulus
money. When it comes to accountability in the health IT standards world and
health IT governance, the standard for accountability has been ratcheted way
up. The administration will ultimately be accountable to demonstrate results
and those results need to include health outcomes, health improvements, job
creation, and cost savings particularly in administrative costs and innovations
in healthcare delivery.
Government sanction standards and certification efforts must bring in
competing voices and competing players in the IT market. It is critical that we
don’t lock in today’s processes and technologies, which were designed for very
different markets than the interoperable exchange of health information and
vision by the law. I know Bill said he is going to testify later, but in his
report, the National Research Council’s report, he emphasized how current
products and deployments fall short even in the aggregate of what is needed to
support the IOM’s vision of quality healthcare. Ossifying the current
electronic health record systems or using them as the measure of what users
should be purchasing is a strategic mistake.
What are the next steps? First and foremost measurable health goals must be
established. I think I have made that point several times. The Secretary of
Health and Human Services should have the ability to establish and prioritize
these goals. Examples could include eliminating drug interactions, reducing
unnecessary or costly diagnostic tests, improving health outcomes for patients
with chronic disease.
Once these goals have been established the federal government should work
with NIST and the HIT standards committee as designated by the law to propose
basic standards for transport and sharing for critical information required.
The most important specifications would be for the secure transfer of data
where the range of appropriate data types who standards could be and should be
tightened over time. Again, this is something we wrote about in 2004 in the
connecting for health roadmap. I think it still rings very, very true. Rather
than seeking semantic interoperability, all data types universally coded in
precisely the same way. The implementation specification should support basic
data liquidity using commonly used Internet standards like XML with subsequent
effort made to observe and harmonize standards with real world use and vice
Second, how should we validate the use of standards? It needs to be an open
market based approach. When certification is necessary the method must be
market ready, low cost and nimble so that certification itself doesn’t become
the bottleneck investment of this scale.
Once these certification criteria identify there could be a plurality of
private certification organizations including ones like CCHIT to compete for
public and private sector business. HHS could issue an open request for
multiple vendors to certify in conformity to the NIST criteria. These could
include application vendors or independent bodies. This model is widely used in
other e-gov approaches. NIST has performed similar functions for other
Concluding I think out of what is clearly a shared desired to see our
country emerge from this deep economic crisis we are contemplating the largest
federal investment in health IT in US history and that brings with it an
unprecedented public scrutiny to deliver on the promise that health IT can be
used in a way that will improve the health of patients, save lives, save money,
and create jobs. People will want to see measurable improvements of this
investment. I think the law established a good framework. We can get IT right
and if we do, it will justify further public and private investment. I thank
you for this opportunity.
DR. WARREN: With that Rob was asked to ask questions first.
DR. KOLODNER: I need to step out but Carol as usual lay out the issues.
Hopefully as we saw when we wrote the articles in healthcare that our areas in
agreement far outweigh the areas of disagreement. You have heard that we are
intending to embrace the approach that you outlined and look forward to letting
us know where you think we veer from that.
DR. DIAMOND: Thank you, Rob. Let me say I’m being articulate. I need to
thank the whole Markle team for helping me prepare these remarks including Clay
Sharke and the entire Markle staff. Everyone had a hand in it so I don’t
credit. I look forward to working with you.
DR. WARREN: Jeff.
MR. BLAIR: I guess maybe I’ve been around for a long time and we are
constantly seeking new ideas and perspectives and broader understandings in
thinking outside of the box.
DR. DIAMOND: I try to deliver on that.
MR. BLAIR: The thing that I’m thinking of is that when I step back a little
bit in look at it over time it’s less of a new approach and more of an
evolution because it seems like we actually are not throwing away the building
blocks that we have, the technologies that we have. We are just recognizing
goals that now we have the technology to be able to meet that we didn’t 5 years
ago, 10 years ago. I guess that’s my speech is that I’m tending to look at this
more of an evolution than a change.
The other piece is I am so glad Dr. Tooker and Dr. Diamond are both here
sitting next to each other because as part of what I perceive that evolution to
be when we move forward into trying to understand how to wind up supporting the
concept of a nationwide health information network, we had to come up with new
organizations and processes and those are only like 4 or 5 years old. Now we
are winding up looking at those in terms of new goals, broader goals because we
had additional resources in addition to additional technology. One of the
things that Dr. Tooker was mentioning was moving from the concept of use cases
which has been useful in the last several years to moving to a broader
perspective called value cases. My question is to both of you. Dr. Tooker, you
have heard now from Don Detmer who has introduced communication technologies
and who has introduced the idea focusing on outcomes. Carol, you pointed out
new goals that are broader. Dr. Tooker, could you tell me your perception as to
the degree to which, it may be partial, the degree to which the idea of value
use cases can accommodate the broader perspective that Dr. Detmer and Dr.
Diamond offered. Carol, could you maybe give a reply to that in terms of your
perception of the degree to which the concept of value cases is beginning to
move evolutionary towards providing a structure that could accommodate the
things you articulated?
DR. TOOKER: I want to be clear by saying that the National Health
Information Technology will be set by the federal government as the purpose of
the policy committee. Within that framework though we need to recognize the
contributions of those who are closest to the delivery of healthcare and by
that I mean consumers, patients, and employers in more regional and local
areas. To that extent our sense of the value cases is that the ideas and there
are hundreds and thousands of them out there need to be encouraged and sought
and developed as to how this large enterprise in health information technology
is going to solve their problems. I think to date and it certainly is not a
criticism there has been a sense that those who are not directly experiencing
these issues in healthcare are making policy decisions as to what should be
done. I think building on a lot of the work that has already been done I don’t
want to certainly scrap that, but we need I think to have a bottoms-up approach
to how ideas are being generated to solve this healthcare crisis.
A point to be made I think in the context of the Recovery Act is all of us
are so and I’m certainly one of them that are caught up in what the
implications of this act are and this is the end of February. I don’t want to
confuse the Recovery Act with healthcare reform. It is complementary to it and
there may even be changes in the current legislation because it was as always
there are revisions to legislation once they are adopted in such a hurried
manner. In speaking to the Ways and Means Committee yesterday they expect to
have a Medicare bill and the health reform bill this year. Our end goal is
healthcare reform to get to the goals that Carol was talking about. Our primary
interest is being how health information technology contributes to healthcare
reform. In the value case notion is to derive the stimulus from the environment
in which patients and all of the institutions that provide healthcare live to
generate those value cases.
I think that we have learned a lot from the AHIC process. I having
participated in the AHR work group within the AHIC process I always had the
feeling that it was pretty constrained by the framework of the AHIC process
itself now and that there wasn’t as much as an opportunity to get the
bottoms-up that we needed to get the ideas that we needed to solve these
DR. DIAMOND: Dr. Tooker and I have worked together a long time. He is a
member of my steering committee. I really admire his efforts to try to make
sense of where all this is going and I think we’re struggling with it. The act
has only passed a week ago or so. We are all trying to understand what this
means. I guess what I’m saying is that the urgency of the crisis we face now
and the reason for this act coming about to begin with, and the significance of
this investment, warrants a fresh openness to I think the question that Larry
teed up at the beginning of the day, what of 302 million Americans really need
and want and what should happen? What should our goals be? I think what I loved
about Don’s question that he asked and his set up was that he really teed up
some questions that you don’t normally hear in these use case or value case
conversations. I think we need to go there. That is what this law forces us to
do in my view. We have to be open to that.
I know with all change people worry about what it will mean for the efforts
that have come before it and I would just say that -– I agree, Jeff, with
your comment about evolution. This is an evolutionary process and we will take
what’s good and run with it and what isn’t so good we should shed and move on
because I think there are lots of stories and I listed some of these in the
appendix as well. There are lots of stories where you try for a really long
time to fix what isn’t good and you end up wasting more time and money. I guess
what I’m saying is let’s open up those questions in a way that this law
requires us to do.
MR. REYNOLDS: Thanks to both of you. The thing that I continue to struggle
with and I would really like you to help and then I hope the panel the first
thing this afternoon can help further. As somebody who spends a lot of time up
here and then spends a lot of time at home implementing standards and then
spends a lot of other time at home talking about investing in things. To play
off of Larry’s a little bit, I don’t hear us talking especially as
interoperability. I get it. We go on to E-prescribing. I see that and I think
that’s actually been a great story. I tell that story regularly. As I look at
EHRs and I think of some of the other things that are going on, I still don’t
ever hear the voice of what makes a different in every doctor visit. What makes
a difference in every nurse touch? What makes a difference in every home health
visit? What happens is there’s a lot of information that’s going to be flying
Dr. Tooker, I will take your example and I use it a lot, the 24 percent of
the Medicare people who have five illnesses and up to 14 docs. They show up for
a 10-minute visit to a new doctor or a 20-minute visit to another doctor.
That’s the confluence to me. I think we have seen with E-prescribing we have
seen some of the decisions support and how you can deliver things. This whole
idea of a huge set of records and my father is going through exactly what you
were saying and every doctor that touches him. We don’t even know how to tell
him what’s going on. We don’t know how to tell him who is seeing. That to me
when you talk about Don’s communication and the other things you said. I just
don’t see that getting much attention. The interfaces get attention. The
interoperability gets attention. So good you just backed a dump truck up to
that doctor’s office and just dumped all the records out electronically so they
went into a nice bin, but what does that doctor do as they are trying to care
and then the next day when another doctor sees him. As I wear all my hats,
that’s the one that I can’t see. If you have any indication on that and then
this is a pre-announcement to the panel this afternoon. It’s going to be the
same question. At the point of care is where we are going to make a difference.
That point of care is changing more and more. It’s not just doctors and
hospitals. It’s nutritionalists. It’s pharmacists. It’s anybody that’s going to
touch this person and this philosophy of whatever medical home we might have.
DR. DIAMOND: Let me try to respond to that briefly. First and foremost I
hope I was painfully clear that when information technology pays dividends, it
is when people use information to make a better decision. That is when it pays
dividends. I think many of our constructs or ways of thinking about things have
focused on what I will unfairly I guess label as transactions. Even
E-prescribing is a transaction. The value of sending a prescription
electronically is not actually just sending it electronically. It is when you
as a prescriber have the information to get the dose right, to know what other
meds the patient is on so you don’t give them a drug/drug interaction to know
if they have refilled the medication they are supposed to be on in a timely
manner so that you can help them or ask them about compliance. That is where IT
pays off. I think we have to force ourselves to ask the question about whether
the transactions or the focus on transactions is going to get us there. That is
why I really pressed on the idea of what is it that you want to do with the IT?
What is it for? If you can answer that question, you come to very different
conclusions about where standards are needed and how technologies deployed.
DR. TOOKER: That really is the cardinal question. I think that all of us
struggle who spend a lot of time in this town of how do you get from here to
-– I think that we are actually in a better state than where we were even
2 or 3 years ago. Let me begin with not so much as calling it the medical home,
but what the inside of the medical home should look like.
You may know that -– I’m sure you do know that in order for a medical
home to be a medical home or the purposes of testing the model that the
National Committee on Quality Assurance has developed a set of criteria. The
criteria are designed to know from a payor’s point of view that I’m willing to
pay for this experiment if I am assured that what the capacity of that medical
home is to deliver care and the ability to handle health and healthcare
information is one of the requirements of the medical home. Notice I didn’t say
electronic health record because it begins at tier 1 and it may go up to tier 3
depending on who the payor is where you would have a fully developed electronic
The second important piece about the medical home is that it is specifically
designed on the Wagner Chronic Care Model. The key feature and it is hard to
overemphasize is the Wagner Chronic Care Model, and I will be happy to send
everybody lots of material about the chronic care model, is this is a
partnership between informed and activated patients to the extent that they can
be with the practice. In this case it’s a practice team. We are so used to
thinking about providers even the word provider. It sounds like it’s a
unidirectional process when in reality with the patients at the medical home
with Wagner care model is a process of interaction and empowerment of the
patient within the practice. You can’t do that without the patient having not
only the information, but the ability to act on that information and make the
decisions within the practice. Although it is early on there are some I would
say experiments within integrative health systems that have been able to derive
dividends from these approaches. One of them that we have talked about all the
time is Geisinger that has had electronic health record for a long time, but
its integrative system in which the patients and those providing care are
integrating within the practice. They have been able to demonstrate higher
quality scores, increase base in satisfaction which means from a patient’s
point of view these are really experiential measures which says that the
patient not only appreciates the process of care, but to Carol’s point the
outcomes of care that they got what they thought they wanted out of that
process where the challenges really are moving from these well-integrated
cultural systems to where the real universe of care is and where understandably
we are not seeing the dividend yet. We know almost nothing about the small
person practices out there where most of the healthcare is provided.
While I share this sense of urgency and frustration of not moving there
faster, I think we are learning how to do this and a big question I think for
this body right here is whether you are going to be focusing more on virtual
environment in which there are communities of scale for patients and practices
and not, I can’t say allowing, but not creating environments in which you can
continue to have fragmented care out there that is not delivering the quality
of cost reduction safety that we are talking about.
DR. DIAMOND: Can I just make one follow up comment? I just want to make a
brief follow up comment. I just spent two days at Gracinger recently. I think a
lot of people hear about these systems and think that somehow it was because
they had an electronic health record. In fact what they would say is that the
electronic health record was very transaction based and to achieve the health
improvements and health goals they had, it was enormous reengineering of
processes and technology and the EHR in order to get there. I think that’s the
critical message and it’s why I made such a point of saying that we really
shouldn’t ossify the systems we have today or use them as the bar for what we
are trying to achieve. Let’s set the right goals and challenge the systems to
DR. WARREN: I have to say we are way over and we staying for lunch and if we
want to finish our things, what I would like to do is call and break for lunch
and the three of you who are on the list if you could ask your questions at a
later time or grab people. If we could be back here about 1:20, I think that
will give everybody a chance to find your body and go up for lunch and then
come back. Thank you very much for your testimony.
(A luncheon recess was taken at 12:40 p.m., to reconvene at 1:30 p.m.)
A F T E R N O O N S E S S I O N (1:30 p.m.)
DR. WARREN: This is our third panel for the day. The theme for this is
enhancing standards adoption by users. We have tried to pull out
representatives from various stakeholder groups. The first one to go will be
Dr. Waldren from the American Association of Family Physicians and then Alan
Mertz from the American Clinical Laboratory Association. Followed by Rosemary
Kennedy from Alliance for Nursing Informatics and then by telephone, Rod
Piechowski from the American Hospital Association. Dr. Waldren if you want to
DR. WALDREN: On behalf of the 94,600 members of the American Academy of
Family Physicians I want to thank you for this opportunity to offer this
testimony. I did provide you a copy of this testimony. There are no slides.
Like Dr. Warren said my name is Steve Waldren and I direct the Center for
Health Information Technology from inside the American Academy of Family
It really is an honor to speak to you today about health information
technology and how it increases value to physicians and to patients. The role
of standards in industry should be to help drive innovation in that industry,
to make products and services cheaper, better, faster. Innovation can elevate
what vendors and users compete on and it can lead to new products and services
that increase the value of the industry as a whole.
It is also important to realize these standards are not generally and in of
themselves, but rather they are an ends to a mean, a means to an end. They are
set of outcomes or results such as lower cost to increase efficiency, greater
convenience or faster response times. Generally speaking, there comes first the
desire or need to attain these ends and then in pursuing these ends to the
market place, the participants in that market find ways to standardize. This is
true of standards in the computer industry which allow components from multiple
participants to be assembled and configured dramatically lowering the cost of
personal computers and to increase the speed at which these products are
delivered to the market and into the hands of users.
It is true in the chemical industry where standards for piping, for design,
and for various substrates allow for almost anyone to assemble components into
the system for the production of starring for example.
It is true in the communications network we call the Internet where
protocols and standards allow for many different kinds of software to plug and
play relieving the end user to worry about whether or not their email or their
audio file would be interpretable by their receiver at the end which maybe at a
very remote node on that network.
There is a saying in standards that Dr. Diamond mentioned earlier, which is
that standards are not created. They are adopted. There is a big difference
between proposing a standard for use and its actual acceptance supply groups of
users or by an industry. It is important to understand the predictors of
success as well as the various standards of adoption of one in which is to
accelerate the adoption of standards.
I would like to discuss some of those predictors of success and barriers
relatively to health IT standards and inoperability and computability
particularly from our experience working with the continuity of care records
standards or CCR over the period of 2005 to present.
Based on experience working with our physician members have they adopted and
used health IT in their practices over the last 6 years. Our experience with
the development and deployment of the CCR standard and other standard there are
few predictors of success for standards adoption that we have been able to
observe. I want to talk about five of those.
First, market demand. Aligning incentives to create business models that
derive demand for standard is the greatest lever to accelerate adoption.
People, businesses, operational units of organizations. Only with good reason
to exchange health information before they do so on a large scale using
CVS minute clinic is using the CCR standard to send patient visit summaries,
information back to their patients, primary care physicians, tens of thousands
of times per week. They needed to find a way to efficiently through the use of
an XML clinical messaging standard because their business model required them
to communicate with the patient’s medical home. This was essentially a promise
they made to their customers and to their partners. These messages in CCR
format, now are traveling to SureScripts RXhub network. This is the same
network that is used for E-prescribing, but now those messages are doing at a
much lower cost than they would have using pay performs and a postal service.
They are also offering patients’ ability to upload those same CCR XML files
to their accounts on Google Health, Microsoft HealthVault or the Cleveland
Clinic Personal Health Record. Without these business case drivers, we doubt
that CVS medical clinic would have standardized this set of processes or would
they continue to look to reduce their cost of data exchange.
Number two, keep it simple. Lower in complexity will increase the ability to
adopt a standard. A strategy of building on early, easy wins that can be built
upon will go much farther than creating a more complex comprehensive standard
at the beginning. We believe that the relative ease and use and simplicity of
design have contributed to the early uptake and adoption of the standard across
a wide variety of projects and programs in the US. Google, for example, chose
the CCR standard for its use within its Google health personal health record
application large on the basis of familiarity with its XML schema and tagging
conventions to nonhealth care experts and its engineering team.
Another example is Dr. Kevin Peterson at the University of Minnesota. He is
using the CCR standard to extract data from multiple disparate EHRs located
different practices and primary care to aggregate that data to answer new
clinical research questions. Also the AAFP has a practice-based research
network that is a funded project through AHRQ called DARTNet, which also uses
the CCR to also aggregate that clinical data.
In each of these cases simplicity of the exchange standard has been key to
the success from programmers and developers who are working in an otherwise
very complex environment which large numbers of proprietary systems connect.
Number three, strong clinical involvement early. A standard must satisfy
clinical need for the clinician. It must take into account the care of delivery
process and work flows. It is only possible with clinical involvement
throughout the standard development and deployment to achieve this. Many of
these standards recently created in health data exchange and the quality phase
has been created in the lab. By that I mean a group of talented engineers and
informaticists create a very sound work product from a technical specification,
but still the standard may not fully take into account the utilization of the
standard during the care delivery process.
Having strong clinical input in leadership from the early phases of
standards development, which carries through testing and real world
utilization, leads to a strong clinical focus for the standard which increases
the value of that standard. We have certainly seen this with the adoption of
CCR standard. Clinicians lead the effort from the beginning in collaboration
with nonphysician technologists. Now companies like Google, Microsoft,
Medtuity, Sandlot, visionary medical systems and many others are using the CCR
standard today in production in large measure because of the clinical validity
and the continued championing of the CCR by physicians and nurses. Some of
these companies are also using the CCR standard to interoperate between EHRs
like Allscripts and Nextium into personal health records and to regional health
Number four, real world testing. Any standard that is not developed from a
proven implementation in a market place must undergo real world testing that’s
functionality and usability. Mandated standards that do not work well and which
are released into the market without testing will be detrimental to further
standards adoption and possibly create wide confusion. Real world testing must
reflect the actual characteristics of our healthcare delivery system. Over 80
percent of healthcare visits in this country occur in medical practices of 10
or fewer doctors. Over half occur in very small practices with three or fewer
physicians. Testing standards in these environments is very difficult and is
challenging, but one cannot and should not assume that the standards test in
the large group practices or hospitals will necessarily function the same level
efficiency in much smaller settings or in the community.
We have learned from experience within the CCR standard that the early use
of the standard was accelerated by helping to organize groups of companies or
vendors whose product offerings were very similar such as personal health
records, electronic health records that help support software, assisting
developers in each of those classes of applications to collaborate with one
another, to test the standards for specific use cases to that class of
Number five. Standardization is not the end goal. As the nation moves
forward with health IT we must not be myopic and solely focused on the
technology and the standards. I think this point was brought home by Dr.
Diamond earlier and by her article that she published in the August 2008 Health
Affairs Journal she distributed today.
We must focus on the decided outcomes we all want such as higher quality,
improved safety, lower cost, increase efficiency. If we only achieve adoption
of certified standardized HIT products but do not allow us to reach those
goals, it lost a great opportunity.
It has been a pleasure to speak to you today about this very important
topic. We are encouraged to see the language of the health IT portion of the
American Recovery and Reinvestment Act of 2009 which articulated that the
Secretary of HHS and the National Coordinator for Health IT must consult with
NCVHS. I strongly encourage you to exercise that opportunity to provide
leadership and insight making sure that the nation’s health IT goals are met by
this unprecedented investment in health IT. Again on behalf of my members thank
you for allowing us to provide this testimony.
DR. WARREN: Thank you, Dr. Waldren. Alan.
MR. MERTZ: Dr. Warren, Mr. Blair, members of the subcommittee, thank you for
the opportunity to have me testify today on behalf of the American Clinical
Laboratory Association, ACLA. ACLA represents the entire breadth of the
laboratory industry from national labs, regional labs, and local laboratories.
I’m Alan Mertz. I’m the ACLA’s president and I appreciate your interest in
enhancing standards adoption by laboratories. I wanted to testify today because
of the importance of this topic. It is extremely important to our members. Our
members have a tremendous amount invested in electronic records. It really is
the way we do business now. That is why I wanted to be here. I’m the president
of the organization. I am not the world’s expert on IT as Jason and Ken can
attest. I did bring some experts with me if you throw a couple of acronyms at
me that I can’t understand. I may pull them up here to answer some questions.
It really seriously is important to us. I wanted to be sure to be here.
As I alluded to our members have an extensive history providing the nation’s
hospitals and physicians with leading edge health information technology to
streamline the laboratory test requisition process and even more importantly to
speed the deliver of results. Sometimes it is as little as a couple of hours
and for a nursing home most of our tests the results go to the physician the
next day. Electronic delivery is very important to us. It is this longstanding
experience of providing connectivity to our nation’s healthcare system that
informs my testimony.
I would like to focus on three specific areas concerning enhancing standards
adoption by healthcare providers. Number one, past efforts that development and
implementation of clinical laboratory test health information exchange. Number
two, remaining standards development and implementation challenges and three,
national coordination of IT efforts and the model for future work.
First I would like to discuss past efforts to developed standards for the
electronic exchange of laboratory test results. I will focus my comments on the
EHR lab interoperability and conductivity standards, E-links project and the
Health Information Technology Standards panel, HITSP, electronic health record
results reporting, interoperability specification. As some of you are aware, in
the spring of 2005 the California Healthcare Foundation at the request of the
former national coordinator, Dr. Braylor, kicked off a new initiative to
develop a national standard for the real time reporting of lab data to EHRs.
Adoption of a national standard would help simplify the use of EHRs. Physician
practices would reap tangible benefits early in EHR implementation and reduce
cost with EHR system installation and configuration.
E-links work closely with other national or international efforts to develop
the standard to ensure widespread adoption. Participating organizations include
the Certification Commission for Health Information Technology, CCHIT,
connecting for health, the Markle Foundation, the eHealth Initiative, EHI,
docket, the CMS, Integrating the Healthcare Enterprise, IHE, the Public Health
Information Network, CDC-PHIN, Health Level 7, HL7, and ACLA.
Fast forward 3 years later when in September 2008 the California healthcare
foundation awarded a second round of grants to independent physician
associations, community clinics, private practices and other ambulatory
providers across California to provide technical assistance and up to $15,000
each over 12 months to implement a new version of E-links. This effort was only
possible because of broad, equitable stakeholder involvement and the
incremental real world approach that E-links projects took when developing the
standard. The result was a standard that healthcare providers could easily
adopt and implement within their existing health IT systems and one that
improves access to accurate and timely laboratory results.
In our opinion this success is in stark contrast to what the HITSP
experience has been. Sam Karp of the California Healthcare Foundation may have
put it best in his testimony to the Institute of Medicine Board on Health Care
Services and National Research Council Computer Science and Technology Board
when referring to HITSP he said, “Not a single data element has been
exchanged in the real world healthcare systems using data standards this
process has developed or deployed.” He went on to state that “Greater
emphasis is placed on ideal standards and less on what can be feasibly
implemented in the short term.”
One example of this dichotomy of ideal versus real world standards is the
inclusion of the Unified Code for Units of Measure, UCUM, in the HITSP
specification for laboratory results reporting. While UCUM is used within the
field of radiology, no laboratory not even the most sophisticated currently
uses UCUM to electronically transmit laboratory result data in the United
States. The reason HITSP has produced futuristic outcomes is that it operates
under a governance model in which the team with the most people at the table
wins and IT vendors have consistently had the most people at the table. We have
seen this elsewhere not just here. Unlike IT vendors, however, most labs do not
have the resources to devote full-time employees to HIT standards development
activities. By contrast CCHIT should be commended for ensuring more equitable
representation among stakeholders, which has resulted in a more practical
incremental approach to certification criteria for EHRs. For example, CCHIT has
used a flexible mixture of HITSP and non-HITSP standards and specifications in
its certification criteria and acknowledged that UCUM is not ready for prime
I would now like to briefly talk about the remaining challenges to standards
development implementation and adoption of electronic laboratory data
information exchange. Outside of results reporting, which I have already
discussed, there are two fundamental issues which remain to be addressed.
Standard laboratory order codes and as mentioned in prior comments before NCVHS
in our 2006 testimony updating the Clinical Laboratory Improvement Amendment,
CLIA, regulations to accommodate and promote the transmission of laboratory
Developing standard order codes will prove to be an effort exponentially
more difficult than E-links ever was. We know this because a number of
ACLA-member laboratories have already attempted to do so to no avail. Without
delving into the efforts that have taken place and why they failed, I only
caution that any successful effort will need to be incremental. The other
requisite for successful outcome is that it often even bear representation on
the standards development organization charged with coming up with a standard.
We need efforts are ongoing within the labor of medicine as well as HHS’ Office
of the National Coordinator to develop standard laboratory order codes and hope
that they will follow these guidelines for a successful outcome that
laboratories and other providers can get behind, adopt, and implement within
their current IT systems. We are concerned, however, that the HITSP general lab
orders’ working group is already proceeding with an ambitious agenda to
harmonize lab order standards for an extremely broad use case including
ambulatory care, acute care, public health, order status queries, and
electronic order updates under the same governance structure that allowed IT
vendors to dominate the outcome of the lab result use case.
Standardizing lab order messages will bring tremendous value to labs and the
healthcare system generally by resolving many issues that negatively impact the
delivery of lab results. However, in the process for achieving this important
advancement should not repeat past errors, but instead acknowledge that
adoption will only be maximized if those most directly affected laboratories
have an equal voice in its creation and have confirmed its operational
The other remaining challenge I would like to raise has to do with CLIA.
Pursuant to CLIA when a test result report information is disclosed by the
physician to another provider such as a Regional Health Information
Organization or RHIO, the clinical laboratory is still responsible for the
content and format of that report. In addition, when an EHR vendor changes the
test result report that is provided to the physician, the clinical laboratory
is still responsible for the content and format of that report in accordance
with CLIA requirements.
Taking this one step further laboratories would then be responsible for test
report information in the RHIO despite the potential for reports to be modified
several times over each and every manipulation made to them when an EHR vendor
modifies the content and for each and every time the content was used, for
example, primary care visit, office visit to a specialist, hospital admission,
nursing home stay, et cetera. This regulatory burden needs to be addressed in
order to facilitate the exchange of electronic health data for treatment
ACLA has proposed to address the situation. In order to address the issue of
EHR vendors modifying laboratory result content we propose the following
solution. Amend the CLIA interpretive guidelines or the CLIA regulations to
clarify that the laboratory’s responsibility ends once a CLIA compliant report
is received by either the client or the vendor or other contractually obligated
Second, to address the issue of information shared with other health
providers aside from the ordering physician, the clinical laboratory’s
responsibility for the test result should end once the result is provided to
the ordering physician or vendor. Interpretative guidelines should make clear
that the laboratory is not responsible for subsequent disclosure of test result
information made by the physician.
The last topic I would like to address is the need for greater coordination
of health IT efforts by the federal government and the model for future work.
While I spent much of my time today discussing clinical standards adoption, it
is important to note that a number of administrative standards will be adopted
over the next few years, claims attachment moving from ICD9 to ICD10 et cetera.
In addition with the recently enacted stimulus bill providers must implement
significant changes to their privacy policies. Between the administrative and
clinical standards mandated by the federal government and newly enacted privacy
provisions, providers will be continuously updating their practice management
and health IT systems over the next several years. The potential for these
efforts to overwhelm providers including laboratories is very real and great
coordination by the Office of the National Coordinator is badly needed.
Finally, regarding the model for the development adoption and implementation
of HIT standards for the nation as a whole I will reiterate a number of points
I have already raised. Future work needs to be incremental and based on real
world healthcare systems. E-links is a great example of how the process should
work. The initial version of E-links establish standards for result reporting
of approximately 80 percent of perform tests. A little over a year after that
effort started clinical care providers and laboratories in California were
exchanging lab results using E-links through the use of EHRs. The success of
this effort was due largely to the fact that it was incremental and operated
under an equitable governance model. Standards adoption will only be maximized
when ambitious goals are tempered by operational realities and equal
shareholder representation. Thank you very much for the opportunity. I will be
glad to answer questions.
DR. WARREN: We are going to keep our questions to the end. Next Rosemary
MS. KENNEDY: Dr. Warren, Mr. Blair and other members of the subcommittee, I
thank you for this opportunity to present testimony on behalf of the Alliance
for Nursing Informatics. I am Rosemary Kennedy a registered nurse with
approximately 25, 30 years of experience in healthcare implementations. I spent
much time building interfaces and implementing systems and designing systems.
Until recently I was also a practicing nurse in a telemetry unit and gave that
up just recently for doctoral studies.
I will review a little bit about the alliance for nursing informatics and
speak specifically to some of the success stories that have occurred in terms
of standards development and adoption and then I will quickly move into an
answering the questions as put forth by the committee. The alliance for nursing
informatics is a collaboration of organizations that represent a unified voice
for nursing informatics. It provides a single point of connection between
nursing informatics groups and the broader nursing in healthcare community at
large. Members of the alliance for nursing informatics include standards
nursing terminology groups and also includes other organizations at the state
and regional level.
There are about 26 organizations representing thousands of nurses that are
practicing out in various organizations in terms of developing standards,
implementing standards as well as educators in various universities. These
nurse informaticists are domain experts in standards development and more
importantly in the deployment within the electronic health record today. They
are engaged in all levels of EHR standards development, testing, and
integration and many of them are responsible for driving the implementations of
the applications in healthcare delivery organizations. They assume various
roles from CIOs to programmers to analysts to knowledge engineers working on
applications such as documentation, CPOE, data warehouse, home care, genomics,
oncology, all facets of clinical implementations.
I will move now into some of the standards achievements and enabling
processes. The enabling processes for the Alliance for Nursing Informatics is
equally supported by HIMSS and AMIA. They can reach out to the various
stakeholders and experts. The stakeholders and experts that are practicing and
providing care delivery in the healthcare environment to proactively get input
on the standards as it goes through its cycle. An example of the Alliance for
Nursing Informatics outreach was to the AHIC’s successor organization earlier
that resulted in membership, nurses being in the membership in communication
committee as well as the value case committee. Nurses are also represented in
the nationwide health information network. The Alliance for Nursing Informatics
recently appointed a nurse representative to HITSP that was elected as co-chair
to the care management and health records technical committee.
The outreach for the Alliance for Nursing Informatics nationally can get and
gather input from clinicians and for domain experts. Why is this important? It
helps to facilitate the use and reuse of multiple interoperable specifications
across all venues of care because they are involved in the specification as it
is developed and then they take it out operationally into their day-to-day
A nurse serves in a lead role as the HITSP standard implementation technical
manager. Also nurses are involved in Health Level 7. There is nursing
representation on the board, the patient care technical committee, structure of
document technical committee, and the clinical interoperable council. The work
that is done on these various committees extends beyond just standards
development. They collaborate. They do research. They take it back to their
various organizations and entities and actually use some of the standards and
measure the impact that it has on outcomes.
Integrating healthcare enterprise, patient care coordination, nursing
subcommittee is working on functional assessment and nursing documentation
profiles using standards and efforts coming from SNOMED as they structure the
interdisciplinary plan of care using nursing terminologies. This includes
cross-disciplinary and organizational support. The connectathon for which many
nurses have built and working today as we speak in Chicago includes HITSP
participation. There are 77 companies registered to test 126 systems. These are
real systems testing real life cases in terms of patient care delivery and the
role for the standards.
In order to outreach and influence and have the standards adopted within
healthcare we need to reach out beyond nursing informatics. The technology
informatics guiding educational reform has done that. They have reached out to
other domains of nursing and have infused standards into day-to-day
implementations. They have provided a form for education on what the standards
are, tutorials, awareness campaigns and more or less they are infusing the
standards and the infrastructure behind it within the elements of the
professional practice such that not only are nurse informaticist is responsible
for moving the national health information network forward but practicing
nurses are equally responsible in their day-to-day practice.
The ISO Technical Committee 215 Health Informatics Work Group 3 provided
input into the ISO 18104 reference terminology model. That model is in
electronic health records and is currently being measured in terms of the
impact that it is having on care delivery.
Within SNOMED there are nursing special interest groups. The International
Health and Terminology Standards Development Organization play an important
role in the ongoing development of the SNOMED CT clinical terminology. Nursing
is represented on quality assurance and content committees.
The Nursing Terminology Summit started about 10 years ago was an annual
event to promote the development and integration of a reference terminology for
nursing within applications. The summit brought together educators, vendors,
and developers of terminology, providers that use the terminology. It started
out as a think tank to identify how we could move forward using standards to
come up with ways and means for using the standards. The providers and the
vendors then went back to their respective organizations and proactively were
able to integrate the standards within their applications. I can speak
specifically to that because I was able to integrate ICNP using the ISO
standard, mapping it to clinical care classification that is live in an
application. The nurses on the frontlines know what we are doing and they are
measuring the impact that that has on quality outcomes, patient outcomes, and
general nursing care. Nurses are also represented in standard subcommittees
within the American Nurses Association, AMIA, HIMMS, and CMS is doing coding
work in long-term care.
I now will move on to the challenges. We have made a lot of progress in the
past 10 years. I really don’t like to use the word challenges. I more or less
think of this as future work or of the state of where we are right now. The
remaining challenges are grouped into four major areas: the complexity of
information exchange unique to care delivery, the implementation of standards
within providers’ organizations, standards development in testing, and then the
points of intersection between standards and the broader context of care
deliveries. I will walk through each of these challenges and give concrete
The complexity of information exchange is unique to healthcare delivery. I
think the prior panel members talked about this. Information may seem like a
simple exchange in terms of a discharge document going from an acute setting to
an ambulatory setting, but in real life it is very complex. There are various
work flow processes, many different operational procedures. Much that goes on
in terms of clinical decision support. It’s not just the standard that needs to
be infused and moved out there. It is workflow processes, decision support, and
other factors that play a role outside of standards if we are actually going to
improve the quality of care delivery.
Implementation challenges from a provider perspective. In a lot of the
organizations that are members of the Alliance for Nursing Informatics
represent those groups regionally and locally that everyday are infusing and
working with the standards and putting them into their implementation efforts.
These are some of the challenges that they are faced with. Creating standards
that fit the natural dialogue of the clinician. There is a mindset out there
that once they standardize and it goes on to a template that maybe some of the
natural dialogue and we talked about that earlier. The communication that
occurs between a nurse and a nurse and a nurse and a physician may be mitigated
or something may be lost in the standardization process, also a mindset that
everybody wants to standardize.
There is an awareness, an openness to it. I get many questions about what
standards should be integrated into the implementation project and this fear
that it will increase the documentation time and take them away from patient
These organizations have to leverage exiting information technology by
moving towards interoperability. They have many different systems in various
stages of development. There are certainly migration challenges. Developing
implementation plans that show a value on the investment. I’m seeing an
increase of awareness as information technology plans are being laid out from
the sea level to the staff nurse. They want to develop the value proposition.
So looking at it if you will top-down as oppose to bottom-up. Top-down view
maybe that they want to achieve and improve quality outcomes. Looking at the
quality outcome then they take a step back and say okay what are all the
standards that we kind of need to help us achieve this quality outcome.
There is a great need for repeatable best practice adoption methods to
reduce the implementation cycle times. We have many facilities across the
nation all trying to achieve the same thing and a tremendous overwhelming need
for educational programs for all lovers of the provider organization to
facilitate standards adoption. It is a lot to keep up with and they want to
keep up with it. There’s a need for NIST to keep up with it and they keep
asking for educational program.
In terms of standards development and testing in and of itself there is no
consistent way to access or define some of the data elements although there is
a tool that is currently being developed by AHRQ that will help us be able to
search the elements going across various standards. I say that with all due
respect to that effort because that will remove that challenge.
The existing models use different frameworks making harmonization labor
intensive. It’s just a reality. There is a need for methods and tools to obtain
input from all members of the clinical team. The Alliance for Nursing
Informatics has reached out to these clinicians and they want to provide input
and feedback. We are looking for different methods and tools to help them
provide feedback on a 140-page document in a way that has a short turnaround
time. It is costly too for them to provide feedback and to participate in the
standards development. They want to participate. I think with the Alliance for
Nursing Informatics we have created low-cost tools and methods to get their
input into the standards and have a whole network for reaching to them. Some of
the terminology and the language that we use are different than what they use
so we have learned what terms and codes to use in order to get their input.
Communication between the standards committees. There are so many of them.
In communication what goes on in one committee impacts another committee so
ways and means to kind of facilitate that and communication is needed.
Also and we talked about this with the panels earlier today. Points of
intersection between standards and the broader context of care delivery. If we
could develop a way or a method to create a balance between standards adoption
and the future function necessary for meeting quality safety goals. Standards
play a role in that, but there are other variables that also play a role when
an organization is using technology to meet quality and safety goals. Standards
should liberate, therefore, making it easier to develop healthcare information
technology and I think we have seen that and have success stories of where it
has been made easier.
Developing tools for measuring the clinical effectiveness of standards
adoption. The way that organizations are they integrate within their
implementations that they can identify and quantify the impact it has in the
clinical effectiveness. Organizations are starting to ask for these tools and
these methods as they forward for all the reasons that they are under pressure
to adopt and use the standards.
The future model for healthcare information technology standards development
and adoption. Connecting information within decision support and workflow. I
think connecting standards to the broader aspect of healthcare delivery would
be a huge step in the right direction. A process for developing standards
within the context of care delivery. We create a balance that moves the nation
forward making sure the adopters of the standards are moving right along with
the developer of the standards. This real result in real implementations,
standards that work, and potentially can generate new knowledge that move
standards forward. The circle here shows the process of standards development.
That has been outlined and laid out by the nation. If we infuse in that a
mechanism for combined collaboration between all stakeholders, input from all
stakeholders to reduce the turn around time for the standards development, and
also if we can create tools and methods to conduct pilots and do pre and
So we have a tight connection between the standards that are developed in
care delivery and maybe come up with new creative models where it is the
private/public sector, it is academic universities and providers of healthcare
coming together to conduct the pilots to measure the impact that standards
We also need to integrate this within the education curriculum so as
clinicians come out and practice and provide and nurses in particular, we are
convinced that it needs to be part of our professional practice. That it’s not
just something that is driven by standards, developers, or information
technologists, but is part of the professional practice.
The stakeholders are many, physicians, nurses I think consumers. We need to
pull the consumer in because obviously as we use these standards we will be
interacting with the consumer, pharmacist, quality organizations, and also
content vendors. There is time being spent mapping if you will the
evidence-based sources of content coming from the vendors into the electronic
health record and the need for standardization on all aspects both the content
as well as the use of that content within the electronic health record.
A future model is a model that integrates standards within the broader
framework including decision support and work faux processes. The contextual
representation of data within templates impacts decisions and ultimately what
is transmitted. We are seeing too that based on the level of experience of the
clinician the data should be presented maybe in a different way on the template
or on the screen. You can have standards in terms of terminology, standards for
sharing that information, but then the real meat of the matter comes to it in
terms of the template and the decisions that people make off of how the
information is grouped and presented on the template.
Quality and safety outcomes should be integrated within the testing and
implementation of standards so we could look at the broader context of
improving safety and quality. Prioritization of standards that enhance
cognitive decision making and processes of care delivery can facilitate the
total solution impact because many of these providers is deploying technology
are really looking for the total solution impact in terms of quality, safety,
effectiveness, and financial performance.
The fundamental framework in terms of what needs to be behind the standards
is our policy, our quality standards, our evidence-based practice, and clinical
effectiveness. That is the framework that should drive the information
technology standards. It’s the framework that facilities use to drive the
implementation, deployment of technology within their organization. That would
help align the standards with the overall quality goals that we need to achieve
as a nation.
Concluding remarks. Nurses are affecting change in the development of
adoption of interoperable systems across the nation showing quantifiable impact
on national health. In general a combined collaboration between all
stakeholders is needed to develop, adopt, and integrate standards into
practice. The healthcare environment is ready to adopt standards and they are
seeking tools, methods, and success stories, best practices in order to move
forward in improving the health of the nation.
Thank you for this opportunity to testify on behalf of the Alliance for
DR. WARREN: Rod, are you on the phone?
MR. PIECHOWSKI: I am. Can you hear me?
DR. WARREN: Yes, we can so go ahead.
MR. PIECHOWSKI: Good afternoon. My name is Rod Piechowski. I’m the senior
associate director of policy at the American Hospital Association. On behalf of
AHA’s nearly 5,000 member hospitals, health systems and other healthcare
organizations as well as our 37,000 individual members I really appreciate this
opportunity to speak to you and your colleagues this afternoon of the
importance of and the progress in the development of standards in health
Next slide please. The current process for developing and adopting health
information technology standards resulted from the recognition that with a
multitude of standards to choose from healthcare providers would be reluctant
to make a substantial investment in technology that could be rendered obsolete.
If the system selected did not use standards that would ultimately emerge as
dominance in the market place, providers would have to replace the entire
system or continually invest to adapt it to new standards.
In the creation of the American Health Information Community the Health
Information Technology Standards Panel, HITSP, and the Certification Commission
for Health Information Technology along with the creation of the Office of the
National Coordinator for Health Information Technology watched the focus
government-led effort to resolve many of the issues surrounding health
information technology standards.
The AHIC now the American eHealth Collaborative established a commitment
from HHS to remove barriers to adoption of HIT through the creation of a first
set of use cases for which previously conflicting or nonexisting standards had
to be either harmonized or created in order to support the use case.
Where we have come from is roughly analogous to if you imagine trying to
drive across the country with each state and local community having its own set
of driving rules. If you imagine that in Iowa you have to drive on the left
side maybe in Florida as well, but stop signs might be red octagons in five
states, but be blue triangles in others. If you never cross outside of those
boundaries, those rules are easy to work within. As a local set of standards,
they are fine but when you start to cross boundaries and reach out as we are
trying to do now with the national system of interoperability, the local rules
are much more difficult to deal with. We need to continue to see
standardization move in the direction that we are moving in.
What we have done is really created a basic workflow and structure for
better or worse. AHIC has created a set of use cases from which HITSP takes a
look at which standards are available or might need to be created in order to
match and enable those use cases. They have done a lot of work in harmonization
of that process. The Secretary of Health and Human Services then recognizes
those standards and sometime later well actually between the harmonization and
the Secretary recognizing them sometimes can pass between his acceptance and
recognition. Those standards are then used to certify products by CCHIT, which
has to date, certified probably over a 150 different products. To some degree
we have seen a deal of success if you consider where we came from and the fact
that this actually quite chaotic. It has become a little more focused and we
have an opportunity to reexamine where we have been with that and see if we
need to move forward.
Another outcome of this on a very broad scale has been that we have managed
to engage especially through HITSP and CCHIT hundreds and hundreds of different
stakeholder groups. Alan Mertz made note earlier of the fact that his
organization feels and his organization is not the only one that feels this way
that there could be broader representation and that it is somewhat heavily
weighted toward the organizations that can afford to continually send people to
these meetings and spend the time to do. I have attended meetings of HITSP, for
example, where some of the topic of discussion has been centered around burn
out of the volunteers who are participating on this. I completely commend what
HITSP has accomplished. They have done an amazing job of organizing the complex
work that needs to be done in order to harmonize all of these standards. It is
quite true that a lot of the people who volunteer and spend their time doing
this also volunteer and spend their time doing other things so many of the
people who participate in HITSP leave that meeting in Washington, DC and
perhaps fly someplace else and show up and meet again only under the banner of
HL7 or another organization.
While that is good, we have to also be very cognizant of the fact that we
cannot burn out the same people nor can we have the same people making all of
the decisions all of the time on this and that is one area where we need to
work to broaden the engagement of different groups of people.
I think in short we have really managed to at least build a very basic
foundation for this process and we can at least show a mapping of how we are
able to link goals to measurable process for progress.
On a very large scale we have over the past several years since the
establishment of the Office of the National Coordinator at least been able to
establish a sense of top-level leadership. We have established a process to
harmonize technical standards and in the market place certainly from CCHIT’s
perspective and the role that it plays we have been able to establish a level
of trust that never existed before 5 years ago. In short there we are very much
more intelligent about what we have learned about how to do this.
This is really we are a very nascent point in the development of standards
and harmonization. We really should note that standards like the technologies
that they enable will continue to evolve over time and the processes that we
create must be able to manage that evolution. What is commonly used today as a
standard may in the future serve little purposes except in rare instances, for
example, at one time three and one half inch data diskettes were big
improvements over the five and one quarter inch counterpart which in turn were
much easier to carry around than the eight-inch version. Every new computer at
that time came with drives that could read the new format. That becomes
dominant for a time before being replaced in turn by smaller, faster, and
higher capacity media. We have and we will continue to see similar changes as
health information technology evolves. We see these kinds of things everyday in
our lives and we take them for granted.
The telephone system is another prime example. It is still possible in 2009
to attach a rotary dial phone to our highly digitized system. The standards
that are used in communicating voice still allow for that technology. We still
have not pushed that completely out of the realm of existence. I don’t know
anyone who has a dial phone. I have one in the basement, but I don’t think we
will be hooking it up any time soon. Mostly it serves as a historical document.
Where we are at using the traffic analogy again I noted that this is still
quite nascent. We have basically just devised a process that helps us determine
that all stop signs shall be red octagon across the United States. If you live
in a state that uses red octagon, you don’t have to make any changes, but if
you live in a state where the octagons are green or perhaps they have put the
stack kinds on the other side of the road you have to change and no one likes
change. It is difficult. It creates situations where we can make mistakes and
creates disharmonious situation if you will as we get used to the new
As far as remaining challenges I wanted to go and take a look a little bit
higher than any specific standards and look at beyond EHR. Our numbers look out
into the future and tell us that some of things that really need to be worked
on that aren’t necessarily considered associated with an EHR system really do
have a lot to do with the overall ability to function in a health information
technology enabled world.
Quality matrics is definitely an area that some progress is being made on,
however, it really needs to be simplified and there needs to be an alignment
process created. Currently hospitals are responsible to respond to hundreds of
different requests for quality metrics. In many cases they are duplicative, but
they are created differently, reported to different agencies at different
times, and it creates single-thread nature where each one needs to be responded
to individually and it is quite costly and cumbersome and inefficient. If you
look at the highest level of what we want to do with quality metrics, at some
point we would love to be able to aggregate those and measure them and compare
apples to apples, but if they are not coming in the same format that is much
more difficult to do.
The same is true for data standards. While we are making progress on
vocabularies and some data formats, I could foresee a day where all data could
be exactly the same for a particular piece of information. If that were the
case that would severely reduce the number of customized interfaces that we
need to create and it would make interoperability that much easier without any
of the interpretation required.
Something that isn’t often considered to be part of this discussion are
supply chain type standards. A lot of progress is being made there currently.
The FDA recently came out with another request for some comments regarding the
unique device identifier on medical devices. While there has been some
discussion over the past several years on how to go about doing this, they are
getting closer to coming up with a proposed rule and this is a piece that
really is a key connector to a lot of what we talk about as helping us really
achieve a technology enabled healthcare environment. If we know which device is
connected to which patient or where something is within a hospital, we increase
both patient’s safety and the efficiency of our materials management. It is
especially from a patient’s safety standpoint imperative to do something like
this when we talk about product recalls and those do occur. They are not as
rare as we would like to have them. Once we have that kind of information it
can all be interconnected into the electronic health record and you will know
exactly which infusion pump was connected to which patient and whether or not
it is one that has been calibrated for pediatric use if that is what you indeed
need. Because of the growing sophistication of our system, it will be able to
give you a warning that perhaps you shouldn’t be using that device.
Another issue that has nothing to do necessarily with electronic systems,
but is a big issue for hospitals as they try to manage and make the transaction
between paper-based systems and electronic flows. An example of that is what’s
going on with the DEA. Currently we have incentives in place to use
E-prescribing yet the DEA forbids the use of E-prescribing for the prescribing
of controlled substances which means that if you have a healthcare provider
that is moving forward and becoming more modern and using technology, they
still need to maintain two separate data streams for prescriptions. One being
electronic and then the other being paper based for controlled substances. It
is extra work, extra maintenance, and an extra opportunity for mistakes.
Hospitals are all about providing quality of care and what to avoid errors at
I wanted to point out that Alan Mertz again kind of set the stage for a lot
of this. This really calls for a high level of coordination not just at the
level of which standard we are choosing to use for interoperable activities,
but which agencies and organizations have a hand in some of this and what are
they currently doing and what are their schedules and plans. How does it
intersect with what the state of the art is today? What we see with the recent
passage of the stimulus bill now the question shifts from how am I going to pay
for this given if the credit markets open up to how do we go about doing this.
I think it will expand the kind of thinking among providers have not done this
yet and they will start asking much more difficult questions about how do these
various requirements all intersect? It’s not always about technology and
standards, but it goes into coordination and other levels of policy and our
This is one piece of the puzzle that our members believe it is imperative to
solve and that is how are we going to match patient to the proper set of
records in our electronics system. It has to move records among a variety of
organizations during the course of providing care. For over 10 years we have
been pretty much at an impasse over the creation of a unique patient identifier
as mandated in HIPAA legislation. The issue has been reframed as a privacy
issue when it is in fact about improved patient safety and better quality of
care brought about through technology speaking as from a technology viewpoint.
This is not to discount in any way the importance of ensuring that electronic
transactions remain secure and private within the healthcare domain. That
requirement is paramount. Privacy is a concept, however, and it cannot be
enforced by nor can it be guaranteed by technology alone because it is a
cultural and policy issue.
The goal here should be to create an environment in which technology can be
leveraged for what it does best and in which policy is leverage for what it
Technology cannot manage behavior nor can it eliminate human factor. A lot
of the privacy breaches that we hear about are the result of someone leaving an
uncrypted set of disks or backup tape in the trunk of a car and when they go
into a store and come back to find it missing. The technology itself can’t do
We also have a very long history of what can happen to and what does happen
to our information on a daily basis. The social security number has long
migrated into other areas of use and people are very wary of that. Our
financial information is bought and sold everyday. It’s a complete industry to
buy, sell, and rank credit. As far as I know I’m not allowed to opt out of that
process at all. I think when people see what is coming down the road as far as
information, it is very easy to paint this as a similar situation where our
information will be used for nefarious purposes.
Technology is very good at managing information efficiently. It can encrypt
data. It can audit events. As far as how relational databases work, IDs are the
essential elements of how those pieces of data find each other in a very
accurate way. If we are going to avoid having some kind of an identifier, then
we are going to introduce uncertainty, risk, and threaten patient safety
because at the highest level that identifier is what draws everything together
in a database. Without it we know that pieces of information belong together.
We’re just not that certain if we don’t have that identifier to which
individual those pieces actually belong.
That doesn’t necessarily mean that a single identifier is the only answer to
this. There are other methods that are being investigated.
Another issue around this is that when people talk about a national patient
identifier the mind immediately goes to government ownership and issuance of
this and government control. From a technology standpoint when we talk about a
national identifier we are really talking about the scope of that identifier’s
uniqueness and it does not necessarily have to be a top-down issuance. There
are methods no doubt available that would allow us to create a way of matching
patients accurately with their records that do not have to be issued from the
With that said technology is one part of the equation and the social,
cultural, and policy elements are another side of that equation and we need to
have some very insightful long-ranging discussions about what that really means
and find some balance in the discussion and figure out exactly what we do have
in common and what we hope to achieve because we would like to apply technology
to do what it does best. In a perfect world of course none of these issues
would exist if we didn’t stigmatize disease or use information to deny or
In our continued process government leadership is required. It has helped us
get where we are at today. We still have a long way to go. We need to increase
the wide range of participation that is already and very, very helpful.
Continue to engage in knowledgeable healthcare leadership. This is
essential. People who have been on the ground implementing and doing this on a
daily basis know what the real issues are. This is not something that can just
be decided without having had that kind of experience.
The best way that we can really accelerate this on one hand is continue to
push forward to do the types of standards that are required to make the
connections given the systems as they are today in our current state and on the
other hand we need to make an effort I should say to look far into the future,
15, 20 years and see where it is that we really want to go and identify what
the future healthcare delivery system can look like. Then we can migrate our
systems over the course of the day to day and month to month and year to year
evolutionary process toward that visionary goal and along the way we absolutely
must maintain a sense of patience and collaboration because this is really not
about the technology in and of itself as much as it is about what we are hoping
to achieve through use of technology.
In summation it really can be kind of outlined in a very simple way. Just
think long term, create a plan, coordinate the tasks and layers, and that
includes planning for implementation, transition, and training. The privacy
issues that have been introduced in the stimulus bill really will add a lot of
work and if these systems need to be certified in order in time to achieve and
receive some of the incentive payments there is a lot of work that needs to be
done and we also need to prepare people for the culture shock because
technology changes the way everything happens. Tell that story, measure the
results, and tell the story again and continue to evolve.
What we do locally can scale up to what we need to do nationally. Thank you
very much for the opportunity to present this afternoon. It’s been my pleasure.
DR. WARREN: Questions?
DR. FERRER: Dr. Waldren, I had a question for you. In speaking we heard both
sides. We heard level of depth of HITSP specifications of the standards. It was
so granular and so deep that at times there were no real life with very few
real life exchanges whereas Dr. Waldren you related to the CCR was a relatively
successful exchange of data. It had high clinical utility, but it wasn’t as
prescriptive as the level that HITSP has gone. Could you comment as to at what
level of standardization do you draw the line to say enough is enough or we
need to kind of do more of?
DR. WALDREN: It’s a great question and I think it’s one that we continue to
struggle with as an industry. What we have tried to talk about instead of
talking now about interoperability we really take that and divide that into two
concepts: portability and computability. The first is to be able to get the
decision maker the information in a timely manner that they can manage. The
next is actually having a computable that a computer information system can
interpret that information and take some action. Either that action be a
physician you have a new discharge summary that came from this hospital and it
was one of your patients. That’s the level that we have.
Another is to say here comes the discharge summary with listing all the
medications with this patient and there’s reconciliation with the medications
that I have in my record. I think those are two different things that we want
to get to that future state, but we have to understand where the data lies
today and what are the work flows and processes that are in place today
relative to structuring and codifying data.
What we try to do with the CCR standard is create the ability to have that
highly structured codified data, but also have the ability to have less
structure, less codified data, but not to the point of a large narrative of
saying here’s an entire history, but rather here is a list of medications and
here are individual medications. It may have the entire name, dose/sig,
everything as a string, but it also then can break that up and to have it
individually codified and structured. What we found with the CCRs there are
certain organizations that have that highly structured data. When they have
that they highly structure that and they codify it. With those they don’t still
have the ability to put that information in the standard and use the standard.
There are definitely issues when you talk about okay I have it highly
structured and codified and I get one end that is not. I as a receiver of that
information don’t get as much value a lot of it. I have to figure out and go
into my medical record looking and finding out that I have to add these 10
medications and each of one of those I have to go through the whole process of
clicking the medication at a drop down list so that I can get the RxNorm code
or NDC code. I have to select the right dose and put all the information in so
that my decision support can actually do that.
I think what you want to do is make sure that you understand where the data
is today and provide a very easy path the structure and more codified. One of
the criticisms to the standard and I think rightly so is that we don’t
prescribe what control vocabulary you should do. Any controlled vocabulary that
is in UMLS, 150 or so I think it is now. You can use those in the standard.
What we also said, for example, in the PHR space if we said that for
procedures you had to use the CPT codes, a lot of the person’s health record is
saying so you are going to make me pay $12 a record to AMA to put in a billing
code that I’m not going to send a bill for. We had to make some compromise
during the consensus process.
DR. WARREN: Mike.
DR. FITZMAURICE: There is much rich testimony. It is hard to know where to
start. I guess I would start with Steve Waldren. You use CCR as an example of
the four predictors of success I think it was, five predictors of success for
standards. Then you end the fifth one by saying if we only achieve adoption of
certified standardized health IT products that do not enable us to reach those
goals, we have lost a great opportunity. I tried reading it without certified
and standardized to say if we only achieve adoption of health IT products that
do not enable us to reach those goals, we have lost a great opportunity. I
agree with that. I look upon that as a call for evaluation.
With putting in certified and standardized, it seems to question the value
of certification and standardization. Since you use CCR as an example, I’m
wondering if CCR certified and standardized?
DR. WALDREN: It is standardized, but it is not certified. My point though
for adding both those terms is many times when people talk about technology and
say oh it’s certified. When they say certify they think in the back of their
mind, oh it’s interoperable. It has all the security things that I need. I
don’t want to pick on those, but I just want to make sure that I was explicit
in kind of call the point and saying if we don’t look at
certification/standardization to achieve these outcomes then we are not going
to get those outcomes. I wasn’t saying that CCHIT isn’t providing these values
and HITSP isn’t providing these values, but I am also saying just because it is
certified and it uses the HITSP specifications I don’t think that proves that
we are going to measure these outcomes and that there’s actually been an
DR. FITZMAURICE: Thank you for clearing that up. Alan Mertz, again there’s a
lot of rich testimony. I noticed as I go through it, you dislike mandatory or
speedy or forced implementation of standards. That seems to be the message that
I get from your presentation. Use an example of UCUM that the HITSP
specification or lab results reporting. Our lab results are uniform and their
measures that no standard is needed or using that as an example of don’t force
it on us but it might be nice as a national target. Then I’m thinking of LOINC,
E-links and other standards that these standards were national reference
standards without a mandated use by this date. Will they be okay as a target or
is it the standard you really don’t like?
MR. MERTZ: I’m going to ask -– this is Ken Glassman from Quest
MR. GLASSMAN: One of the problems with the UCUM is that it has not been
adapted by any of the commercial laboratories, public health laboratories. I
reached out to the associations for both and none of them had a single
laboratory that had actually implemented it. Implementing standards that have
not been appropriately piloted in the target industry doesn’t seem appropriate.
I don’t think the lab industry has said not to use UCUM. We are really pushing
that. You need to test it, validate it, and make sure that the receiving
clinician understands what that represents. In some cases UCUM uses some
characters that have to be escaped because it uses HL7 delimiters as a code and
one of the other issues and concerns is if you are going to escape it out in an
ID field, how do you know to put it back in? ID field is typically -– you
don’t expect to do those kinds of features and there are some other issues with
DR. FITZMAURICE: So you have no problem with national offering standards but
you wouldn’t like the expense of changing legacy systems until there has been
enough time to assimilate it and get a body of people who have adopted it?
MR. GLASSMAN: I think first the IT can’t push standards on the medical
people. The medical people have to review it and validate that it is a useful
standard and then once they have adopted it then I think it makes sense to push
DR. FITZMAURICE: Question for Rosemary Kennedy. There was a lot. I’m going
to go back over your slides. There was an awful lot in there. As I went through
them I wondered what is the relationship between ANI and ANA and I often
lamented the lack of research showing that the variation in nursing practices
implements variation of patient outcome that nursing practices do make a
difference. If we knew that then we can all promote best nursing practices.
Part of this could be lack of uniform coding of patient assessments, lack of
agreement on the terminology to be used, the expensive detailed coding, and the
inability of decision makers to get at the nursing notes to get the information
that they want. Do we have a consistent set of nursing codes? Are they freely
available or priced reasonably and is the ANI liaison with the ANA or is it a
niche organization that says here’s a bunch of problems we want to solve
separately and we are happy to have anybody join who wants to?
DR. KENNEDY: No, the ANI is not out solving problems separately. It is
collaborating with all the other nursing organizations so we are moving forward
as a unified single entity in terms of integrating technology into nursing
practice and improving patient care. There are many studies out there and we
have made significant progress in using structure terminology within nursing
documentation. For the admission assessment, the shift assessment and you
referenced the nursing notes whereby our practice is represented in the codes
and thousands and thousands of sites across the nation that are live using the
structure terminology. As a profession we have moved beyond that and we are
looking to quantify the impact that our practice has on patient care delivery.
We are actually taking those codes. They are driving decisions that are being
made at the point of care. If you imagine a template, a screen that has
measurement instruments, has codes on it, has decision support integrated
within it goes to a data warehouse and then we can measure the impact and the
value that nursing care has on outcomes. There are studies out there that have
been published measuring the impact on outcomes on our care. We can use that as
a feedback mechanism to adjust the documentation.
DR. FITZMAURICE: My youngest daughter just finished nursing school and she
is in Greenwood, science-based practices. It is good to hear that there is a
growing body of nursing practice out there that can use these codes to become
more science based.
DR. KENNEDY: And I think based on evidence so the evidence is behind the
codes and the templates and the care delivery. If we can enter the information
the codes once, reuse them for multiple purposes to measure the impact on
outcomes, and that’s the goal. It’s not so much about the terminology. I
brought up the processes of care delivery because that’s critical as well in
addition to the codes and the structure data coming up with consistent
processes of care delivery. There are many testimonies out there, structured
codes, terminologies being used measuring the impact on care delivery.
DR. SUAREZ: I have one question to all of you. I won’t go one by one that
might simply the process for me. Throughout the testimony and it was a very
rich session there seems to be a common thread I think and that was the
questions around some of the process that have been created towards achieving
what we hope to be the adoption of standards, by users. For the moment the
process says basically that we have a body that establishes some priorities and
we have a body harmonizes the standards and we have a body that certifies or
creates certification criteria. My question is really in many of the comments I
read some issues with that process and some questions around that process and
whether it is truly achieving the goal of simplifying the option of standards
and enhancing the interest in users of those standards. If you were to identify
one or two ways in which that set of processes that we have like I said at the
moment one entity or one body that is setting up our priorities, one body that
developing the harmonization of standards and making recommendations about the
interpretable standards, one body responsible for incorporating those into
certification criteria for certifying products. What would be your
recommendation to improve that so that at the end of the process the users can
actually be able to adopt a more practical easy way those same standards?
MR. MERTZ: I can start with an overall answer. I would like to defer to Ken.
One thing that struck me -– there are really actually three things. One is
the broad stakeholder involvement at every step of the way. Two is this be done
incrementally and look at things to make sure they work along the way that they
have been tested in the real world, work in the real world, they have been
adopted in the real world. The stakeholder part is so important. I hope I am
not offending anybody in this room. I was so angry when the AHIC was
established and I remember talking to Dr. Braylor and I think there were 17
members of the AHIC and there was some great people on there. Our industry
about 60 percent of the medical record is lab results. That’s what we are
talking about, an electronic health record. There wasn’t a single member of
this committee that is supposed to move us down this direction from the
industry that has 60 percent of the volume. We had someone from the Department
of Commerce on there. We had no one from the industry that has 60 percent of
the medical record. This isn’t going to work. I like Rod’s comments from the
AHA and from the physicians. We don’t have as much money or time or resources
to be everywhere. That to me is the first step is to have the stakeholders and
to work with us so that we can be part of –
MR. GLASSMAN: I guess a couple of things. First is a very explicit goal
setting of what we are trying to accomplish with HIT. It is measurable. It will
allow us to take our current processes and marry those up to policies and
payment reform for an example of a policy such that that we can actually
complete the entire circle. For example, let’s take the very first use case of
replacing the clipboard. I think the demographic information not a lot of
disagreement of what that should be and adding a couple of things like primary
spoken language not a huge thing. Then creating the set of standards and doing
that. You also now look at going back to the physician. If I collect those what
can I actually do with those that adds value? Does it change the way that I get
paid as a physician? Has it been demonstrated to increase the quality of care
for that particularly patient? Make sure that we go all the way from the
standpoint of how does it impact at the point of care all the way through? Does
it actually evaluate that we actually made a difference? I would love to see
AHRQ spend some of these monies for comparative effectiveness to take a look at
our HIT standards and HIT products and services and processes and say there’s a
full electronic health record make more sense from a comparative effectiveness
than a registry that does E-prescribing that I dictate all my notes that are in
electronic format, but I can make portable. We don’t know the answer to that.
I think when we get to the process there is definitely some process issues
that I think we would love to see different. I think some of the stakeholder
representations are at issue. I think being very practical in saying let’s push
something all the way through from do we currently in the market place have a
lot of practices and hospitals replacing the clipboard today that very first
use case and really make sure that we drive it all the way through. I don’t
think we have to tear up everything that we have done. I think again going back
to Carol Diamond’s point it is a great opportunity to say sit back and say okay
what are the things that are working that we need to keep doing and drive
forward and accelerate. One of the other things we need to say you know it’s
One final thing too. Make sure that it is outcomes base. For example, in
HITSP one of my biggest frustrations are we have deadlines to meet on the
calendar not deadlines to meet on actual standards harmonization and the
quality of product coming out of HITSP. I think a lot of people have done a lot
of great work, but we are running down a hill and a lot of times we are
tripping and falling and rolling part of the way down.
MS. KENNEDY: I would just like to add something to that and reiterate. I
think the stakeholders including a broader range of stakeholders is necessary
and physicians, nurses, CIOs, all those folks that are running organizations, I
think value propositions that take the broader perspective and look at policy,
clinical effectiveness, and quality are going back to what you said in terms of
start with the outcome and back in. Then methods and tools for quick cycle turn
around time to get feedback, provide the feedback to the specification or the
document is critical in finding low-cost ways of doing that and then testing it
through innovative creative business models pulling together various entities
to have quick testing periods, pilot short pilots if you will. Then looking at
how it is impacting the outcomes that were defined up front.
DR. WALDREN: Rod, did you want to make a comment?
MR. PIECHOWSKI: I would just say briefly that I think we are definitely
further than we were and given what an organization like HITSP was against and
what they were tasked with I think they made amazing progress, but their focus
was on technical standards. I think that given where we are at right now with
the amount of money that is out there for the stimulus for information
technology, the fact that a lot of the questions people have about how this is
actually going to be accomplished will be left to the rule making process. This
gives us a perfect opportunity to reassess and to adjust the system that you
have heard described here today. Perhaps we need another organization like
HITSP, but something that doesn’t look at technical standards. It starts
looking at the broader interrelations of all the different organizations that
are involved in this massive undertaking.
DR. WARREN: Every one of you and even this morning’s panel kept talking
about stakeholder involvement. Everyone has come up with it’s hard to get the
stakeholders there. Either they don’t have the finances to show up or there is
too few of them and they have to spread so thin and so we get fatigued and yet
we also have examples where if the stakeholders aren’t there especially with
dealing use cases, critical elements go missing and they are not found out
until the end. I think UCUM is a great example of that. I know on medication
administration no one thought or remembered that the nurse had to verify an
order so that step was missing out of that.
From the panel any suggestions on how we can get representation to the table
to start looking at standards development and harmonization?
MS. KENNEDY: A couple to mind in terms of getting some of the key
stakeholders. I know that ANI reaches out to key stakeholders and through
simple tools, use of the Internet, pictures. We send information out and
inseminate it. I just did that about 2 weeks ago to about 25 people. It was
defined what we were trying to accomplish. They knew the goals and the
objectives and they could provide input. I think that that is one vehicle for
trying to get stakeholder input. I don’t know they necessary need to physically
be at the table, but if various entities could identify who some of those key
stakeholders are and reach out at the local and regional level to get feedback
from people. I am finding that they are very eager and willing and want to give
the feedback. If it is presented to them in a way that is in accordance with
their language in some tool or method that they could look at because certainly
a nurse knows that she needs to sign off on an order and knows the processes
and the steps if there are ways to kind of provide some of that feedback
electronically to get the stakeholder feedback, I think that that is worth
MR. GLASSMAN: One of the things that HL7 does is they actually reach out.
They aren’t ahead of the ballot process when they do not see enough people
signing up for the ballot to say we do not have a large enough cross section
and they message that fairly regularly throughout the ballot period. I think
that’s part of it. People sometimes don’t see the first and second emails as
they go out. If you are not diligent, it’s not going to happen.
DR. WALDREN: I think one thing is almost a forcing function as Walter talked
earlier about. Liking more this notion of that it’s not voluntary standards
that have adopted it. It’s involuntary. You have to adopt these. I think one of
the things also I think if I am correct in the fact is that there is now
requirements for certain stakeholder groups. I think the first is making those
as requirements to really align and say we have to find ways to really engage
those. I do think the technologies are out there to start to engage into it and
engage those clinicians. I think many of our physicians in the family medicine
space don’t have very big margins. If they are out of the office for half a day
or even a week, that is almost too much for them to be able to maintain their
practice so finding ways to offset those costs for them to participate.
I think the other thing too is be very focused and targeted on what really
matters to them today, getting lab results in, not having to pay an extra
$5,000 to $25,000 to my EMR vendor to get a lab interface to Quest and to Lab
Corp. Not to have United say you have been using Quest and you have interface
but now you have to use Lab Corp because it’s cheaper for us. For me now I have
to pay an extra $5,000. Books in those things that are really paying for them
are going to be easy for them to be involved. Again, focus on kind of the
forcing function. Looking at these virtual technologies and make it very
practical to them and hit their paying point.
DR. KENNEDY: I just wanted to add one other comment to that. I think that
there is a culture shift that needs to occur and maybe it has started already.
In terms of some of the key stakeholders who were clinical folks that it’s kind
of this mindset it is a professional responsibility as a nurse or as a
physician or as a lab technician to look at the date and the information that
is needed to provide care and maybe if somehow we can change that culture and
that mindset, it may increase or hold of stakeholders that would be willing to
DR. WARREN: Rod, did you have anything that you wanted to add?
MR. PIECHOWSKI: I have a very difficult time hearing the conversation over
the past 5 minutes stated in and out, but I think I understand where we are
going with this. I have one comment. Someone mentioned helping to offset the
expense of doing this and I believe in the alternate technologies and the
virtual technologies. Somehow we just need to be portrayed as a national
project that all healthcare stakeholders can participate in. My thought is that
if everyone feels that they can assign, recommend, or volunteer experts from
their organizations to serve as in part of a pool of experts on a particular
topic. If this is organized in the right way and coordinated, then there will
be people available in that pool when answers are needed for that and that way
if people do have time constraints and are unable to participate, there are
still others that have been made available and I would think that that would
also give stakeholders a sense that they are contributing and have a voice in
MR. REYNOLDS: First, I appreciate everybody’s comments. Having sat at this
table for 5 years now and watched all the standards be implemented and the
creation of these things. I like where the last bit of conversation is headed,
but I would love to see you go further if there is any way possible. In other
words, talking about how to get more people at the table really putting
together proposals about how to get more people together because if you think
about it up until now, it has been hard to keep the smaller providers to
understand even just the basic HIPAAs. It has really been driven by the vendors
and at times we have all struggled around this table to get all the vendors to
step up. Right now we are about to hit an exponential speed bump, which has a
lot of money, and as Rob mentioned earlier this morning with another core set
of standards being done by the end of the year, there’s money out there for
education. There’s money out there for other things that are being discussed. I
know in North Carolina we understand the agricultural model of getting
information out since we have a lot of farm country in doing that.
I know this committee would really love to hear some more pragmatic ideas
from you to help do that that we can help recommend to make sure that you are
listened to, you are heard, you are accessed, you are given the opportunity,
even if there is money to do that philosophically in the end. The point is you
guys really all of you represent whether it’s the labs or the doctors or the
nurses or the hospitals are really the adopters of this and you are the ones
that are really going to be the end game as well as what you take care of the
patients that you see. Anything further not to be right now, but anything
further that you can come up with, we are going to be holding this as a series
of hearings, anything further that you can come with that would help us frame
something to talk about how to help and how to help you I think is key. In
somebody’s testimony you mentioned how busy providers are going to be and
there’s a series of things, 5010, ICD10, D standards, personal health records,
E-prescribing. All these things are going to happen at once, which
exponentially speeds it up. I think that would be very helpful.
The other thing each of you gave us something I think is something we can
definitely use like the CLIA discussion and some of the others. I think we saw
some clearly pragmatic discussions and some of the conversations about HITSP
and so on I think we can use. That would be the one plea that I would have
because I think we really need to figure how to help you get engaged and I know
that’s difficult. It’s difficult for all of us. We have day jobs too. We
completely understand that.
DR. WALDREN: One thing that I hadn’t thought about until you started talking
about and making the link for me between the health extension office piece
that’s out there as part of the bill. Not only are they going out to the
practice and helping them do the implementation process, but what about putting
in a very deliberate and explicit feedback mechanism that they would actually
as part of that engagement work with those clinicians to get that feedback.
Fifty minutes of our hour is going to be about helping you get this technology
in place. In the last 10 minutes let’s talk about the future of what your needs
are. I want to talk to you about a use case that it coming out of an AHIC or a
set of standards that are coming out of HITSP. Can you comment on how this
interacts with you in your practice? Maybe a way to force that function as
MS. KENNEDY: In addition to that I think if there is a way to close the gap
between those working on developing the standards and those that are actually
implementing technology and providing patient care. If we could move this
standard, the value proposition not the standard, but the value proposition
that takes the standard and goes beyond because it has the whole value
proposition wrapped around it. Prioritize around pain points that most of these
facilities have and somehow get in front of it and infuse it into the
implementations that are currently underway in a way that’s either grant
monies, small pilot. I think it is one way to help make it happen. As you said
these small facilities, they don’t have knowledge engineers and terminology
engineers and yet they are trying to move forward to achieve the goals that the
nation has set forth and if there is a way to move it closer to early testing
and piloting with some funds and different collaborative models, the university
and a provider, and a standards expert around something that they can get their
arms too that is incremental. It’s small. They can see the value. The standards
developers can see the value and can evolve from there.
MR. GLASSMAN: One of the problems with getting engaged with HITSP and other
organizations that are developing standards if you have not been carried along
and grown up through the process is that it is very intimidating to go to these
places particularly because they are already talking about where they are and
what they are doing. They don’t bring you along as you go and therefore not
giving people an education to get them submerged into the process. Basically
intimidates them to the point where they don’t return. One of the things you
might want to consider doing is maybe regional education opportunities to
educate them, explain to them what all these acronyms mean. A lot of people say
HITSP and they wonder what that means. What CCHIT. There are so many ways to
pronounce CCHIT. Everybody is thinking that you are talking about different
things. If you did an educational process in a regional setting, you might be
able to get those regional stakeholders to say okay I’m not going to be as
intimidated by the process because you are going to help me understand what the
process is about.
MR. REYNOLDS: That is a great point. All of us go into any of these new
things and are quiet for the first X amount of time until you catch up enough
to where when you move your mouth, you don’t feel stupid. I think that anything
that again you can do maybe outside that so that we understand the needs and so
the industry understands the needs out there then people can translate them
into the committees, but the biggest problem is understanding what -– a
number of you mentioned work flow. If you spend any time in a small doctor’s
office, that’s a big deal when you start messing with that. They don’t have
knowledge engineers who go out and say okay today we are going to do this and
oh by the way Thursday afternoon we have a new system and here’s the next deal
for you. Again thinking outside the box so we are not cramming them into a
process where they got no chance of success. Leaving them in their own world,
but drawing as much as we can so that we make sure that when we build these
standards the standards align to what they do in real life which has not
necessarily gotten to the table as much as it could have.
MS. GREENBERG: Judy, can I comment on this particular point because I think
Walter had to step out. I think your point is very good and I know that in the
last year particularly HITSP has really tried to do more of that and as I
pointed out to Walter because I think he actually co-chairs the communication
and outreach group. They started holding webinars. They have these introductory
presentations that they provide to a person when they first come in to a group.
I’m sure Walter would appreciate it and I think we would appreciate whether any
of that has really reached people and whether that has been helpful. I still
think that doesn’t address the amount of time that is required to really be on
top of that work and the highly technical nature of it and all the things that
you have said are spot on.
I know compared to in the first few years and the last year HITSP has been
trying to do that and I think it would be helpful assuming that HITSP is going
to continue or something like it to know how that could be more effective also.
DR. WARREN: Rob, you had a question?
DR. KOLODNER: First of all. Thank you. I was sorry that I missed the initial
presentation. I will listen to it later and look at your notes. I think that
the issue of finding ways of having broader stakeholder involvement in the
process as we move forward. I don’t know how many of you were able to be here
this morning, but a number of the things that you are talking about as we go
forward and we have a network in which to be able to exchange this and give the
opportunity to work through some of the issues that you are talking about. I
think that we need to make sure that we monitor the involvement and we make
sure that there is that broad stakeholder involvement as we go.
One of the things that we will also want is given that there are now
resources, there are now incentives, we can begin to make business cases for
some of the things that one of our obligations to government is to look across
to stakeholders. Steve mentioned that for the provider what is the advantage of
that clipboard. Well, we all know is patients that it has. It may not be for
the provider, but it certainly is for the stakeholder and to the degree that
most of us don’t fill out the thing completely every time because we can’t
remember all the things. There is a certain variation that the provider may not
know or data that is missing.
Steve, you mentioned about the lab charges. I am wondering if that is one of
those areas that we need to understand better. I don’t know if that is outside
the scope of this hearing, but it would seem that standardizing that so that
there’s a standard link would drive that down or at least has the potential. It
doesn’t mean that that’s what’s actually going to happen, but at least if there
is really not a new connection but there’s one connection to exchange that that
might take care of some of those charges that are being passed onto the
DR. WALDREN: I think definitely. We have done surveys on our membership in
the rate of inoperability with lab. It has been more significant in the last
few years than it had in the past. Those that have electronic health records 62
percent have some type of electronic interchange with their lab today. I think
having those standards and the standard profile for that interoperability will
decrease that, but there are two things. One if for example in the CCHIT
certification it is the physician EHR. It’s not the lab system either at a
national lab or a hospital lab. Even if I have the system and it is certified
and it has all the standards if the lab companies and we are not engaged and
part of the process, what is the lever to them to participate interoperability
and has to have both entities willing to participate. I think from a
physician’s side they really want this data and there are standards adopt those
you have to make sure that the lab companies involved. They are also adopting
those as well.
DR. KOLODNER: So if I can turn to Alan and Ken. The issue of what can do?
How can we provide incentives? How can we help you move along knowing that you
have challenges with a lot of the consolidations –?
MR. GLASSMAN: The real problem is standards that are in place but have not
been adopted by the laboratories. That is one of the things that we are having
a little bit of problem with the HITSP standard. That is why the E-links
standard is more preferable right now because that is a stretch but it is not
an impossible stretch as the HITSP is. HITSP reengineers a lot of things that
you do in the laboratory if you are talking about the IS01, which is the
laboratory results standard. A lot of the technology that’s in there has not
been adopted by the laboratory or the EHR vendors either. We need to work
together to make sure that whatever we do is incremental and ordered by both
organizations. I understand we want the perfect thing out there in the future
so that it’s so interoperable that you don’t have to think about it, but you
can’t make that leap in one step. It really needs to be incremental. The
advantage to actually putting something out and then piloting it and seeing how
well you did and getting feedback from the clinicians who are actually using
the data. That’s really very valuable and that’s the reason that when you saw
E-links go out there shortly after that even though we thought we did a pretty
decent job in developing that, we had to come back with a 1.1 and 2.0 and 2.1
as we started learning things and we were actually able to go from link
encoding of about 85 percent to I think it’s almost 98 percent now.
There are a lot of things you have to learn as you are putting things in
place because there are so many moving parts. The pathologists at the
laboratories are talking to the physicians in the field, but then you have all
the IT people in between trying to figure out how do you model and how do you
manage all the different things and so all these moving parts have to keep
being adjusted as the clinicians say this is not working for me. They go back
to the pathologists at the lab and say okay we need the model data this way.
What you see is an incremental process that went on as people pilot things and
discover the best way to deliver the data. Speculating about how that data
should be delivered is much different than how it really gets delivered.
MR. MERTZ: There is this perception – I think you are getting the idea
the magnitudes of complexity there is with labs as most of you know are often
compared to prescription ordering which is the drug, the dose, the name. It’s
not very complicated or even to the financial industry. It is complicated. The
volume of data if you looked at it. I looked at it. When you get into pathology
reports and stuff like that, it is tremendously complicated. Even worse for
compared to financial services industries, which that’s not very complicated.
You get into some of these pathology reports and so forth. It is very complex.
DR. WARREN: We definitely will probably be talking to you guys again about
some more examples of stakeholders and with that I’m going to be ruthless and
cut it off. I’m going to reduce the break down to about 10 minutes so that we
can get back on so people try to be back in here about 3:32 or 3:33.
DR. WARREN: I’m going to have Doug Bell present first. Doug has a plane and
since we are getting later I would like for Doug to go first so he doesn’t get
nervous. It seems how we had him almost at the end. We will have Doug Bell
first then Gil Kuperman then Linda Fischetti and then Bill Stead will be last.
He will be phoning in. Any time you are ready Doug.
DR. BELL: Thank you so much, Mr. Chairman, Ms. Co-chairman. It is really an
honor to be invited to testify here on standards in health IT. Primarily I want
to talk about a study that I am currently leading that is sponsored by the CMS
Office for E-Health Standards and Services with the pilot test RxNorm and the
NCPDP SCRIPT 10.5 standard. I am going to walk through a few slides that talk
about how we got to the stage with these particular standards being the key
ones that we are pilot testing now and then we will go on to talk a little bit
about the study that we are doing and a few results. I am open to your
feedback. I will try to get through this relatively quickly. Stop me if I am
going too fast.
I probably don’t have to this committee about the promise and the challenges
of E-prescribing, but the promise obviously is to deliver information of the
point of care that enables more informed decisions just as people were talking
about this morning and more cost-effective medications.
E-prescribing is also of interest because it maybe a model for understanding
how other kinds of health IT get implemented and understanding how the effects
of the information really drive the value just like people were talking about
we need value cases not just use cases. I see E-prescribing as a domain where
we have a better chance connect specific information to specific clinical value
that gets derived from it. We will look at some examples of that. That is what
we are trying to do in the study that we are currently running.
The challenges though in E-prescribing are also around adoption and use.
Probably you know despite advances in the infrastructure less than 10 percent
of prescriptions are going electronically today and it’s not just adoption. It
is actually use even after someone adopts this technology. We found a study in
Massachusetts also found that many prescribers revert to hand writing even
after they have adopted the E-prescribing system. In Massachusetts they looked
in 2006 and the average percentage of prescriptions being written
electronically was 26 percent among those who had adopted E-prescribing.
In our study, which is still in submission, it was 23 percent in New Jersey.
We also did a survey asking prescribers who had adopted why they reverted to
hand written and there were a lot of reasons, but one of the leading ones was
systems integration issues, issues with transmission to the pharmacies and
issues with the technology just not working as well as it should.
Just to dive into that a little bit more because it helps to set up why we
are looking at RxNorm right now. We look specifically at formulating benefits
in medication history. Formulary benefit standard of course I probably don’t
have to tell this community, but this is for transmitting coverage information.
It has been adopted as a final standard for E-prescribing and that is good
because it helps drive the industry, but we found a number of problems in the
use of the standard that tended to undercut its value.
There were several problems but just one of them to mention is the use of
NDC codes to identify drugs in the formulary benefit standard. One of the
things we did in 2006 was a technical expert panel where we had people in the
industry who are actually using the information. We had these panel interviews.
Here is just one quote from our qualitative analysis of that, “Because of
the potential differences in NDC number, items in the formulary files may not
result in a match on the vendor side. When the physician chooses a drug, he or
she may get erroneous formulary messages or not message at all.”
Then we saw evidence that that was in fact undercutting the value of the
data both in prescriber surveys and in interviews that we did. One prescriber
when we asked about formulary information said, “It’s funny, because it
really hasn’t changed much of my prescribing habits, because I notice that,
like nasal steroid sprays I look under Nasonex, they all have the same yellow
face.” Again there were multiple reasons but one reason was difficulties
with NDC code matching.
Also in the prescriber survey that we did e-prescribers and
non-e-prescribers did not perceive any difference in how much time they spent
or how many phone calls they get over coverage issues. There was no evidence
that formulary benefit was helping these doctors.
We found similar results with medication history and just in the interest of
time I will say again NDC code is used as the drug identifier in medication
history and that created major problems for actually using medication history
data to integrate externally derived prescriptions or prescription data that
the prescribing system vendors had downloaded through the medication history
standard with the data that they knew because they had generated the
Again we found that in a prescriber survey there is no difference between
e-prescribers and non-e-prescribers on a question that we asked about whether
they could identify medications that other doctors have written. Again the
value of medication history was undercut by poor drug identifiers.
RxNorm is hopefully ready to come to the rescue, but we will see. Again I
probably don’t have to explain to this committee what RxNorm, but I will just
say because I will talk about it more. The lynchpin concept in RxNorm is the
semantic clinical drug, which is the generic drug concept to the way a
prescriber, would think about it. It basically is a triple of the ingredients,
the strengths of those ingredients, and the form. For each one of those triples
there is one and only one unique identifier now and forever.
An example is the Azithromycin 250 milligram oral tablet. That is one
clinical drug concept or SCD. That has the RxNorm concept unique identifier
308460. That number will never be reassigned or reused for anything else. If
the NLM finds evidence that this drug is no longer around, this number will be
retired. If it comes back they will bring it back. That is the process that
RxNorm tries to do. There are some other nice features that I won’t go into of
RxNorm. We evaluated it just in a laboratory setting in 2006 just looking at
10,000 new prescriptions and 10,000 renewals. We found that it had 99 percent
coverage for non-device prescriptions. Devices are outside the scope of RxNorm.
We actually had three independent matches though and we did find that there
were some inconsistencies in the mappings of RxNorm so actually a bigger
problem than noncoverage was disagreements. They were matched via their NDC
codes to an RxNorm concept identifier, semantic clinical drug concepts. The
different matching attempts did not always agree. This is bad obviously. It
means there is a mistake somewhere for the same prescription did not get to the
same semantic clinical drug.
About half of the disagreements were due to problems with RxNorm and the
other half were do to problems in NDC codes that basically different drug
database vendors had different sets of NDC codes. To make a long story short
the NLM did implement some fixes that actually extended RxNorm as well to cover
drug device packages. They looked at some of the problems like the unrecognized
synonymy that we found and found ways to resolve those problems. That is where
RxNorm stood going into this year.
The other thing we are looking at is structured and codified Sig, which was
I guess the next most important E-prescribing standard that needs testing that
has not been adopted yet. I don’t have a slide right here explaining the basic
structure of the structured and codified Sig, but it represents the information
that gives the instructions to patients. Today it’s a string. It’s a free text
string in the standard. The structured and codified Sig would break that
information down into separate fields. There are actually 13 segments. Each
segment contains multiple fields so there are more than a hundred individual
fields. These are some of the major segments I just listed on the slide here.
To walk you through what we did to evaluate it back in 2006, it was very new.
It actually only came out half way through the year.
We just wanted to see if it could be used. It’s a very complicated standard
and we wanted to see if it could be used consistently. We took three
independent reviewers who each were experts who had been involved in developing
a standard and we just gave them 42 Sig strings which we selected to try to
exercise the different segments and then we looked at how much do they agree
when they are mapping the exactly the same prescription.
First of all there is something in the Sig called a repeat. There were
actually only 15 out of the 42 that we could really compare because they didn’t
use the repeat and the repeats made it even more complicated essentially. I
won’t get into that.
Then if you just look at the 15, we looked at within each segment the
information that these people put into the segment three in all. Each one of
these rows adds up to 15 basically. These are just for these 15 relatively
simple prescriptions. The agreement was actually very poor. In the dose segment
actually for three prescriptions all three reviewers did the same thing. Then
the other 12, 2 out of 3. There was at least one person who didn’t represent it
in the same way. That was the best just about. I guess interval was another
one. There was very little agreement in the usage of the structured and
Then for the others that had a repeat there was really poor agreement
because the repeats are things like there would be take a particular dose for 3
days then take another dose for 2 days. Each one of those thens becomes another
line essentially or another set of fields. People didn’t even consistently
choose the same number of iterations for those and there was virtually no
We fed that information back and NCPDP working group did make some revisions
both in the fields, the field names, and in the guidance about how the fields
should be used. Although it still has about the same number of fields that it
did. That information was acted on. Now we are evaluating the structured and
codified Sig because it’s also potentially important for standardizing this
piece of information that is transmitted in E-prescribing.
Our current project as I said is sponsored by OESS has the following goals.
We are sort of in a lab testing environment which we are determining whether
RxNorm, the Sig Format, and SCRIPT 10.5. I put this in here. Actually the Sig
Format is now part of SCRIPT 10.5. It was actually formerly incorporated as
SCRIPT 10.4. The other thing that we are testing that’s really new in SCRIPT
10.5 is the use of FMT codes in particularly qualifier fields. Those are
actually the two significant changes that are in SCRIPT 10.5 that weren’t
testing for ambulatory environment in any way.
We are looking at whether vocabularies and code sets can communicate
information unequivocally and how the standards interoperate with existing
Then very importantly is moving toward a live pilot test of these standards
assuming that they work at all and it’s looking like they will as I will
discuss in a moment. To look at how these actually work in the real world and
especially looking at workflow and to look at what difference these standards
make in terms of workflow, phone calls, pharmacy productivity, and things like
that. This was our charge.
We pulled together partner organizations on both ends and in the middle of
E-prescribing transaction world. On the E-prescribing vendor side we have
DrFirst and Allscripts. On the pharmacy side we have QS/1 which is a company
that does systems for independent pharmacies. We have Medco representing mail
order. We had Long’s Drug representing chain pharmacies but they merged with
CVS after we put the proposal in and so they were not able to participate and
we are actually still seeking a chain drug replacement.
Others involved are SureScripts-RxHub of course which is doing a lot of work
for the middle of the transactions, Point of Care Partners which is providing
expertise, and other industry experts, for example, Laura Toper, who is the
lead on the NCPDP working group for structured and codified Sig.
What are we doing? We are pilot testing these standards, but the challenges
that each of these standards has used inside another standard. RxNorm is used
inside SCRIPT and potentially in these four SCRIPT transactions: new
prescription, refill request, medication history and formulary benefit. We want
to evaluate all four of those. We are evaluating that, actually I didn’t put on
the slide, but in the context of SCRIPT 8.1 at least to begin with. Assuming it
works we will carry it into SCRIPT 10.5. Then for SCRIPT 10.5 the standards
that we are really evaluating inside SCRIPT 10.5 are the Sig and also the FMT
codes as I mentioned. The focus really is on new prescription and refill
requests. The Sig doesn’t really apply. It could potentially apply to
medication history but there’s not really a place for it right now and it does
not apply to formulary benefit.
Each nesting of RxNorm, Sig, FMT inside of another standard is a use case
essentially. It is what we are calling a use case anyway. For each one of these
use cases we want to look at historical data to see can these prescriptions be
encoded and then we want to project the effects of actually making this
additional information available. We want to project the effects on patient
care. That’s the lab part. Then in order to do live testing we have to actually
work with the partners to develop and test software modules.
The two important things to be aware of is we not only have to add
information into these standards, but then we have to use the information on
the receiving end as well in order for the standard to make any difference. To
really evaluate these standards we have to get the partners to do something
with the information and we will talk more about that momentarily.
To do a live pilot test we have to actually deploy these modules to
participating physician offices and pharmacies and put it into real world use.
We will monitor it carefully for exceptions but then to really look at the
impact we are doing site visits at these sites with interviews and some work
process observations before and after the package of standards that is put in
place and then we are also analyzing secondary data from logs and claim related
Finally we are planning on doing Delphi expert panel to look at the results
and help CMS interpret the actual readiness for adoption of each of these
standards. That’s the whole project on one slide.
Now to go into the clinical detail which is really essential for
understanding what we are doing I have a slide on each one of these use cases
that we were talking about.
First I have RxNorm embedded in a new prescription transaction. What does
that do? First of all the prescribing system basically just has to look up the
RxNorm concept unique identifier and add it into the transaction. There is
already a field in 8.1 where it can go. Then it’s the pharmacy side that is
something with the information. The things that we are planning to look at are
using the SCD, the semantic clinical drug concept to suggest fill options that
are actually in stock.
Today they try to match the NDC codes and the SCD theoretically ought to
provide it actually better suggestions or more complete suggestions of what
they have in stock in the pharmacy so that could help pharmacy productivity and
we will look at whether the fill time and accuracy is actually better after
they adopt this.
The other thing that they can do in the pharmacy is after the pharmacist has
actually selected a medication to fill this prescription the medication ought
to map to the same semantic clinical drug concept even if it’s a generic
interchange for brand name medication. The semantic clinical drug ought to be
the same for the medication that is taken off the shelf and the prescription it
The pharmacies that are participating are going to compare the incoming SCD
with the SCD for what the pharmacist pulls off the shelf and if it’s not the
same pop up an alert. It asks the pharmacist is this a possible dispensing
error. Then will look at the accuracy of those. Actually even in a lab we are
going to project the accuracy of those and then we will interview pharmacists
and look at secondary data to look at the actual mismatches and whether they
might indicate a reduction in errors.
MR. BLAIR: Doug, this is Jeff. I feel very awkward doing this but we are
considerably behind time. I want to make sure that we have time to hear from
all of the testifiers. Your presentation is excellent, but maybe you could kind
of net out the major points that we need to understand.
DR. BELL: Actually I don’t have to show you the rest of these. The idea is
that basically we are looking at a detailed clinical level. That is exactly
what happens because of each of these.
The one other thing that I will say is that the structured and codified Sig.
We have made a lot of progress on that. We are developing an algorithm that
would actually parse strings and it does look like the fields. I’ll just show
this slide. This shows how a particular string breaks down into the 13
individual segments. It takes one tablet three times daily after meals as
needed for three days. We had to a lot of work with the companies to get a
consensus about some of the semantics still in some of these fields. We think
that we have something that works. We will see is the bottom line. We are
starting to explore that. I’ll stop there.
DR. WARREN: Are you through Doug?
DR. BELL: I was going to talk about SNOMED, too. Eventually that is going to
be a big challenge.
DR. KUPERMAN: Thank you very much for inviting me here today to speak about
perspective on standards and interoperability. My name is Gil Kuperman. I am
currently the Director for Interoperability Informatics at New
York-Presbyterian Hospital. It is actually a recently created role that
recognizes the importance of interoperability both in creating information
coherence within our large organization with our immediate business partners
and also with regional and national interoperability activities. That is one
role that I have. I’m also the Board Chair of NYCLIX, the New York Clinical
Information Exchange, which is one of the RHIOs in New York City. Also relevant
to this discussion is I served as the Co-chair of the core content working
group on the NHIN Trial Implementation Project with Jeff Blair over the last
year and a half or so. Those are kind of the activities that inform my
perspectives on standards in interoperability. Also I spent a fair amount of
time developing clinical decision support systems earlier in my career. That is
also part of it.
Both the clinical decisions support work and the terminology that are
implicit in that and the opportunity to work on the NHIN project refine those
core content specifications both in terms of the structures as well as the
terminologies and the associated rules that go along with that and learning the
extent to which organizations can participate in those activities. Those have
really brought the lessons of my comments today.
I’m going skip this slide which is just a long list of standards which is
the frame of reference that I’m using.
With that context of the questions that I set out to answer were the ones I
received which is what have been recent achievements and what’s the reason for
those achievements? What challenges remain? What should be a model going
forward and who should be involved?
I think there have been substantial recent achievements in standards and
interoperability. Some of the HL7 activities around EHR standards, clinical
document architecture, continuity of care document, HITSP activities around the
interoperability specifications, some of the constructs of the patients,
summary document being one of them, the work of integrating the Healthcare
Enterprise, creating profiles for Cross Enterprise Document Sharing, Patient
Identifier Referencing, Patient Demographic Query, the work of CCHIT and its
movement into the Health Information Exchange Certification space, the NHIN
Trial Implementation Project.
I think is a very important recent achievement especially to the extent that
it has leveraged the activities of existing regional health information
organizations. Some of NIST’s activities to testing in the demonstration of
events where important contributions. The HITSP education activities, which
have been mentioned, and the HITSP processes are important achievements and
then some of the demonstration events around Interoperability Showcases. I
think there have been a lot of important recent achievements.
Some of the reasons for that is that some of these activities are built on
literally decades of iterative process refinement, the processes used in HL7,
HITSP, and CCHIT, IHE build on years and decades of refined processes. The
culture and style of these standard development organizations and related
organizations are important. They are open to all. It’s a place where world
experts can come together. Groups can come together without losing their own
identity. There is some staff support, but it’s mostly passionate volunteers.
Many of the processes work well not perfect by any means.
Early comments to the contrary there is broad input that is generally
sought. Plans are articulated so vendors can prepare and there is good
interaction among these groups. These are some of the reasons for those
In terms of challenges that exist. One of the challenges I think is a
challenge born of success really of some of these standards development
organizations, where some of the processes that work well at a smaller scale
creak a little bit when you move to a larger scale. Just how do you organize
these things is one thing balancing activities between volunteer efforts and
paid staff efforts is another thing. You don’t want to have your domain experts
doing administrative work, but that is sometimes what ends up happening.
Various things around the growth of these standards development organizations
one set of challenges.
Aligning US and international activities is another set of challenges.
Clearly a lot of the activities are national but there is an international
context and vendors especially take the international context seriously.
There are some general process issues. In general it works well I think, but
there are some parochial interest that introduce themselves. Some of the
decision-making processes are cumbersome. How to make best use of the talent
and managing the politics are I think just some of the things that have come up
Technically from a standards point of view when there are complex domains,
do you allow for multiple domains or multiple standards or do you try to
harmonize on a single one? How to address overlaps? How do you address gaps?
Things like that. These are some of the challenges that currently exist in
In some of the other aspects around enforcement I will say a bit more about
that later. Enforcement or testing there is still a lot to be learned here
about the best ways. My perspective and others I have spoken to think that the
CCHIT model is generally good, but it’s a heavy model. How to lighten that or
come up with other innovative ways is I think a challenge.
There are some specific domains where there are challenges. Personal
identifier standards were mentioned earlier. Doug just spoke about standards
for orderable medications, which is definitely needed. In certain circumstances
there are subsets of what are considered standards, for example, SNOMED. SNOMED
is great, but for a specific use case you might need a subset of that. How do
you get that subset? There are some more data types that standards are needed
On the technical side there are converters that are needed either between
competing standards or terminologies that are in common use having some kind of
translators between those things and adoptive standards would be very helpful.
There are some technical challenges about doing some implementations in the
absence of good services oriented architecture. I think this goes along with
some of the enforcement comments I made earlier, but having robust testing
approaches is I think still an open question. When you add the end plus first
organization to an exchange, how to assure that they are interoperable with
Per to some of the comments earlier around lab systems, the feeder systems
would be helpful if those were certified and also there are issues around data
quality. You are moving data, but what is that data? What does it mean? Does it
mean what you think it is and things like that?
Then there are challenges around adoption where the standards are still not
compelling for the vendors and the value cases, however you want to construe
that. Where does the value come from? Those are still not well enough
articulated. These are some of the challenges.
In terms of what should be a model the questions that present out said
selection, adoption, implementation. I think the model needs to be a little bit
broader than that. I think that the work that the standards development
organization, the standards themselves is the starting point if the goal is to
get value. The standards is the starting point, but you need more than that.
What else do you need?
There’s a category of activities that I have called value added activities.
These are implementation guides, implementation profiles, quick start guides,
how to guides, examples, and things like that. The standards themselves can be
very cumbersome. It’s not immediately evident. What to do with them? There are
a lot of these out there. HL7 does some of these. Development organizations
themselves do some of them. Some other organizations like IHE do some of these
things. These things are very important. These toolkits that really let
implementers begin to use these standards, a whole category of activities
Profile enforcement or testing. I talked a little bit about that earlier.
Basically need to assure that an implementation conforms to a certain profile.
CCHIT is one example. It has done this starting with the EHR-S specification.
They provided criteria, test cases, and then it also proctors the tests. NIST
as well does some of this. That is a key part of the model.
Selection of standards for particular use cases. This is really a critical
step. This is something that HITSP has done in terms of the use cases. Because
really to get value to achieve a certain goal of one kind or another you often
need a suite of standards as well as dealing with gaps that the standards don’t
address. This selection for particular use cases is key.
Then you have implementation, which is expensive, but it’s where the rubber
hits the road. A lot of time an effort is spent there. The provider
organization the client maybe the ultimate beneficiary but the vendor has to
Then there is another step called feedback and really getting feedback from
the implementers to the standards development organization and making sure this
happens often enough and fast enough. There needs to more of it and it needs to
Then the last step I put here is instigation, making this happen, priming
the pump, and turning the crank as it were. One example here is ONC in the case
of NHIN Trial Implementation creating a set of circumstances where some of this
actually gets exercised in one way or another.
To turn a thousand words into a picture this is kind of more or less
standards development, selection, implementation, value add, profile
enforcement, feedback and then instigation. I might put that forward as a
candidate to model.
Just one comment. I just wanted to put this in here somewhere. The nation
and regions are struggling with inter-institutional interoperability. I wanted
to make the comment that organizations especially large organizations such as
where I work, New York Presbyterian Hospital, still grappling with
intra-institutional interoperability. Myriad departmental systems with hundreds
of interfaces, which are expensive and incomplete. I think the
inter-institutional obviously is important for a lot of things, but just to
keep in mind that the intra-institutional interoperability is still a
In terms of who the players might be, some of them come from the model
standard development organization, enforces, people who create the value added
activities, people who do the selection, people who instigate. I put here
deciders or some kind of a governance model to help advance areas where some of
the decision making either gets stuck or spins. I think we need to be careful
how much of that is applied judiciously, but probably some of that needs to be
present — users in terms of provider organizations and vendors. The academic
informatics community to tackle some of the difficult challenges that are out
there, research challenges, understanding what works and what doesn’t work,
what are opportunities I think is an important piece of the puzzle. The
international community and then funders to grease some of this, to facilitate
the advancement of some of this.
This is my last slide. Just some recommendations. In terms of support for
standards development process I might suggest that the current processes may
not be perfect but they are not bad, so a suggestion would be to build on
existing processes not to create new ones. Have increased administrative
support for the organizations. Create methods to facilitate feedback to the
standard development organization. Try to find value cases or increase the
value proposition for standards adoption. Better integration between standard
development organizations and the informatics community. I think the US should
take a leadership role in standards development but be mindful of the
international context and again as I said before clearer governance.
In terms of implementation and testing support I think this is a key part of
the whole process: toolkits, quick start guides, converters, support groups,
educational offerings, demonstration, test labs. Innovative and maybe lighter
enforcement models. Certain tools to facilitate compliance with certain
requirements. I listed a couple there, pseudonymization and secure document
exchange as examples.
The medication standards again that Doug talked about. From my perspective I
think this is the most egregious gap that we have right now. Medications are so
critical to so many things and it does not have these kind of standards well
completely is a shame and missed opportunities. I suggest that.
Ancillary systems again as the earlier discussion talked about and support
for research into standards. Those are my comments and look forward to
questions later. Thank you.
MS. FISCHETTI: Good afternoon Dr. Warren, Mr. Blair, and committee. Thank
you so much for the invitation to participate today. I’m Linda Fischetti, the
Veterans Health Administration’s Chief Health Informatics Officer, but I’m not
wearing that hat today. Today I am actually representing an office within Dr.
Kolodner’s office, which is the Federal Health Architecture. The person who
should be here giving this presentation is Mr. Vish Sankaran. You will hear me
refer to him throughout the presentation, but he is out of town speaking at a
different community today.
My role within the Federal Health Architecture which is why I am here is
that I am one of the two nonhealth and human services supporters of this and
contributors of this and it’s the value of the cross federal collaboration that
we have been able to pull together within the Federal Health Architecture that
is really an amazing incident that has taken place.
I have been working in health IT within the federal sector for quite a while
and I did not know that there were so many other people who were doing the
exact same thing. Within the Federal Health Architecture we are able to bring
in people from the Bureau of Prisons. People are putting together occupational
health records at NASA. You know places that you really wouldn’t even think and
meet all of your other fellow colleagues and strengthen your community through
In terms of working with standards I have the honor of representing the
Secretary of Veterans Affairs for this period of time until the representation
changes at the National eHealth Collaborative. I serve at the request of Dr.
Kolodner’s office as a federal liaison on the board of HITSP and then I’m
elected by the membership at large into a board position at HL7. So as somebody
who spends quite a bit of time on standards I’m very pleased that you guys have
taken this on as your task and I look forward to the output of your work.
In terms of the Federal Health Architecture the message that we wanted to
bring today is that we are collaborating across a federal space. We do hear
that there needs to be a leadership role within the federal environment and we
are ready to lead. Now we say that of course recognizing that we can’t get
there without the intelligence and smarts of the private sector community as
well. This is not only a federal activity but we are trying to do as much as we
We have over 200 federal employees who participate in SDOs. We also
participate in the harmonization activities of HITSP and within our own
environments we will be using these standards not only as a bridge technology
for interoperability with the communities around us and the communities that we
serve, but also bridging our own systems from their legacy state into the
The questions asked are the same ones that the previous panel member put up.
One thing that you will notice is that we very clearly make the distinction
between development, selection and implementation. We were also very pleased to
see that this committee did that distinction at the time that you asked the
question. At this point in time we are seeing that people are really talking
about harmonization of standards, prioritization of standards, and have sort of
missed some of those elemental steps that the SBOs do the development. HITSP
does the selection and harmonization and then implementation of course is NHIN.
To deal with development the question was what are some of the recent
achievements. We wanted to highlight the HL7 Electronic Health Record System
and Personal Health Record Functional Profile as the previous speaker did as
well because this is something that has really been able to engage the
clinicians to understand the language that is being said within the standards
community and are able to bring that together. We also have a quite of number
of federal participants who worked on that and across federal environment and
were able to bring that about.
Additionally we want to talk about the success that we along with our
private sector colleagues have had with role-based access control.
I’m going to tell you a little story. Back in 2001 at the time that VA was
looking at the development of our future systems in a SOA environment. We knew
that we were going to need role-based access control. We pulled together a
bunch of folks and asked them what are you guys doing. It turns out that
everybody was disparate and different in terms of how do you allow a physician
who has a right to know to see this amount, but then how do you allow an
occupational physician this information and then how do you allow a clerk to be
able to see this information within your system? Based on a role what sort of
access are you going to get into the system?
It was in 2003 that we started an informal work group of other people that
we found were also having this issue. Then it was in 2004 that we took forward
to HL7 and presented it as a project that we asked HL7 to adopt within the
understanding that this would become part of the family of HL7 work products
and that we would continue in-kind support to not only generate it but also to
As you can see here it was in February 2008 that we then were able to
achieve an ANSI-approved standard. We are now revisiting, reversioning,
reballoting that in September 2009. Based on this initial work now we have
additional work that is able to come out.
I really want you to hear those numbers, 2001, 2003. 2004 was when we went
to HL7 and then 2008 and 2009. This is before. This is how we have previously
participated in creating standards with a huge amount of in-kind support from
our office. We know that the future will change and that this type of timeline
will no longer be acceptable. As we do speak with the leadership within HL7,
they recognize this as well and they are very interested in looking at their
process and figuring out a way and I’m sure that you invited Dr. Jaffey at a
later date to be able to change this process by which standards are created.
In terms of selection I’m very pleased that I have the honor of employing an
incredibly talented group who has been very successful in the standards
organization as I have previously mentioned, but then those same folks have to
go over and participate in the selection and harmonization activities at HITSP.
It has been a wonderful activity what has taken place in HITSP. It really
allows for the standards to roll down into a consumable format for the
implementers, but yet it has taken a huge amount of work.
You have already heard this this morning, but for me to be able to find
people who can participate in SDOs and the harmonization they have to be people
with the technical know how. They have to be people who can be effective
socially within a voluntary consensus organization. They need to be people who
have the ability to travel a huge amount of time. This is the only team that
gets to me their weekly report a calendar of their travel.
Almost all of these employees are on tele-work agreements and work from who
knows where because they are on planes it seems like all the time. This is a
very unique skill set. It is one of the highest price points to get on contract
and we need more of these people because while I have the luxury of having
quite a number of people working in standards and harmonization, the issue is
that we need more and we need more very quickly as we move forward.
Most importantly the reason that we support these folks is that we are able
to meet the needs of the federal sector as we engage in a collaborative group.
One of the advantages of FHA in fact is that we are able to take all of the
subject matter experts who are on the plane all the time have them meet with
their federal counterparts and strengthen not only their understanding, but
also their effectiveness within the community as they are able to use their
peers to extend their influence and reach in what they are working on.
Then of course the final part of your question implementation and that’s the
Nationwide Health Information Network. We will continue to see all of our
activities within the Federal Health Architecture and through the federal
standards community becoming more aligned with NHIN. This is really where our
use cases are coming from and our priorities that we select and choose to work
on will very much be the extensions and gaps and what’s needed to make NHIN
The second question. Describe the processes that enable these achievements.
First of all we do want to, if I hadn’t made it clear already, again emphasize
that FHA provides that forum and resources for the inter-agency communications
and collaboration related to standards activities. We are able to do that both
with our SDO as well as our HITASP activities. The federal agencies actively
participate in the organizational processes. This is the processes of HITSP,
the processes of HIE, the processes of NeHC, the processes of the SDOs, all of
the SDOs that we participate. That really drives when we are doing ballot
review, when we’re writing ballots, when we’re posting information as
Then what Vish, within the Federal Health Architecture, is really trying to
get his arms around because up to this point as with all standards development,
it has been sort of a hobbyist activity is that he is working with all of us to
bring us together to work through this by-step process where we identify the
business priorities, we architect the solutions, we plan investments. He
recognizes that if we want to change that story of taking 4 years to develop a
standard that we are going to have to create some investment that in fact will
be able to accelerate that work whether that’s deployment of in-kind resources
from the departments or whether it’s direct funding. We don’t know at this
point in time. Implement solutions which we will actually be working toward
creating those standards and then measure the progress, the standards that we
have actually put funds to support. Do they meet our need? Is that what we were
looking for? It really changes our relationship with this managed development
Then your last question. What challenges remain for Health IT standard
development selection and implementation. In the absence of tooling to enable
programmers to operationalize developed standards we know that everyday we are
getting variations across standards as people are trying to interpret what they
see. The developers can’t consume directly in an electronic manner the
standards that are being created. We need this tooling to bridge that space
between the standards and the developer.
Our original slide actually went two slides long. It was very small font and
we looked at it again and we decided what is the number one thing that we want
you to hear in terms of challenges. We know that this probably isn’t
immediately available, but we would like a cross SDO so not an inside an SDO
with one inside every SDO, but across SDO reference information model,
reference terminology model, security trust framework, information exchange
framework, and common methodology for developing and implementing standards. We
realize that we may need to start with these very simple data elements, such as
patient. How do you identify a patient? Then slowly build into some of the more
To talk directly to Federal Health Architecture activities and how we see
our role in the national activities. This is a little bit of a busy slide. What
you will see are things that you are very familiar with across the middle of
the slide and that is the Standards Development Organizations who then interact
with the selection and harmonization activity of HITSP who of course is driven
by priority setting authority such as NeHC. Who then have their products
consumed by NHIN and CCHIT? Then we overlay on here how does the Federal Health
IT Standards Organization Engagement Plan contribute to this? This is primarily
communication, collaboration, coordination that must take place across all of
these different activities so that hopefully you are not dealing with a lot of
conflicting federal messages that are coming out at the same time. The goal
would be to try to get the federal voice coordinated prior to going out into
these different organizations.
Who are the players and who should be involved? The answer is everyone. If
you don’t recognize the ecosystem that is out there for standards creators, the
harmonizers, the enforcers/profilers and then very importantly the user of the
standard which is why we put so much energy into standards development as the
user perspective then you are going to miss someone and you are not going to be
quite as successful.
FHA deals with this collaboration through the creation of one committee, the
Federal Health IT Standards Organization Participation committee. We provide
in-kind resources for standards work. We coordinate the federal input. We
represent federal needs within the standards development and then identify
priority areas of interest related to standards participation.
Three primary goals for 2009 is priorities, communication collaboration, and
then providing a quick resource of position papers that we can use within the
federal sector to share with our colleagues so that we are all aligned and
moving in the correct direction.
Then of course the other priority that I spoke to has to do with how do we
figure out moving resources specifically to SDOs and the harmonizers in a way
that we can get our needs met much quicker.
Specific recommendations. These are very simple. They are also unabashedly
federal because we are a federal office providing recommendations to a federal
advisory committee. We would like to see continued support towards development
recognition implementation of common models, frameworks, and methodologies
where appropriate. Support the implementation and maintenance of a Federal
standards registry. Support the implementation and maintenance of a national
testing facility and associated equipment. I think you have heard that a lot
today and then support continued interagency collaboration and coordination
related to standards activities. That is the end of the FHA report and we look
forward to questions.
DR. WARREN: Bill, are you on the phone?
DR. STEAD: Yes, I am.
DR. WARREN: Okay you are on.
DR. STEAD: This is Bill Stead. I am the Associate Vice-Chancellor for Health
Affairs and Director of the Informatics Center Professor of Medicine and
Biomedical Informatics at Vanderbilt where I serve as the Chief Information
Officer of the Medical Center and the Chief Information Architect for the
In addition to these responsibilities, I have had three national roles that
are related to this. I served on the Commission on Systemic Interoperability
2004-2005 and then I recently chaired the National Research Council Committee
that just released the report Computational Technology for Effective
Healthcare. I also in January chaired the National Library of Medicine Board of
Regents working group on health data standards.
What I am going to try to do is summarize the relevant observation from the
NRC Committee site visits and then conclude with a synthesis of recommendations
from these various perspectives.
The Institute of Medicine’s vision for healthcare and wellness calls for a
system that is safe, effective, patient-centered, timely, efficient and
equitable. As the committee looked at this vision, we basically identified six
information intensive aspects that would be necessary to allow it to happen.
Comprehensive data on patients’ conditions and treatments and outcomes would
be the foundation. Then you would need cognitive support for healthcare
professionals and patients to help integrate patient-specific data and
evidence-based practice guidelines and research results.
Then you would need the tools to manage a portfolio of patients and to
highlight problems as they arise. You would need rapid integration of new
instrumentation, biological knowledge, treatment modalities, et cetera, into
the learning healthcare system. You would need to be able to accommodate a
growing heterogeneity of locales for provision of care and finally be able to
empower patients and their families in effective management of healthcare
decisions and their implementation. When we set the target for improving
healthcare that is what we are expecting from our information systems.
We then visited eight health systems, sites that are leaders in the use of
healthcare information technology to improve quality because we wanted to
assess the gap between the best of what is deployed today and what would be
needed to achieve that vision. The sites represented a broad spectrum:
government, for profit, not for profit, academic, community, commercial
systems, and homegrown systems. While we saw a lot of successes, the
information systems that we saw, even in aggregate, fall far short of what
would be needed to achieve the IOM’s vision for healthcare. I think frankly the
expectation of Congress for dramatic reduction in healthcare expenses.
The problematic aspects of what we saw and these are places that have a lot
of what Gil referred to as intra-institutions healthcare IT. We still saw
fragmented patient records. We saw clinical user interfaces that mimic paper
without human factors and safety design. We saw poorly integrated biomedical
devices. Systems that were often used to document what had been done after the
fact for regulatory and legal uses. Rarely did we see support for
evidence-based medicine and computer-based advice. There was little integration
of clinical research activities into clinical care. We found centralization to
be the predominant method of standardization while almost every example we saw
of innovation was local close to the action.
That led to three sort of central conclusions. The first was that the
current efforts aimed at nationwide deployment of this technology will not be
sufficient to achieve the vision for better healthcare and it frankly might
even set back the cause. Success will require much greater emphasis on
cognitive support about which we mean assistance for thinking about and solving
problems and in the near term to deal with this we feel that it is very
important to embrace measurable healthcare quality improvement as the driving
rationale for the IT adoption efforts.
One of the epiphanies during the process of the site visits was that
healthcare uses information technology mainly for automation and transaction
processing, supporting work that can be done over and over with little
variation. Healthcare makes much less use of information technology for other
purposes and other industries.
For example, simple connectivity, the social networking approaches to
linking people to each other and systems. Decision support, the simple
approaches to making choices clear, and data mining for discovering
relationships among data. While automation works well for small well defined
systems, the other approaches are much more effective for complex systems with
diverse data sources. We believe healthcare needs to achieve much more balance
among the four uses of information technology.
With that background I make the following recommendations for consideration
in the standards component of our healthcare IT initiative. The first really is
that we should redefine the objectives of the standards initiatives to reflect
these challenges. By that I mean I think we are right to embrace interoperable
health information as the goal, but we need to define interoperable data as
data that can be assembled and interpreted in the light of current knowledge
and then reinterpreted as knowledge evolves.
Reinterpretation requires access to an archive of raw signal: voice, image,
text, biometrics, et cetera not just a relatively lofty set of standard
concepts. Second we need to ensure the separability of data from applications
so that other applications can use them and third limit the use of standard
data, by which I mean data that can have only one interpretation to situations
where the meaning is explicit and stable over time, such as with drug
So with that a driving frame I then believe we got to take a portfolio
approach to enabling interoperable health information. The elements of that
portfolio would include standard practices which would basically give the
techniques for cognitive support of health professionals and patients, for
support of human factors, for adaptability to support iterative process
improvement, and for effective use of quality.
We also need terminology frameworks, use a knowledge-based approach to
terminology management, leveraging synonymy and annotation of relationships
among concepts and across terminologies, such as have been done in the Unified
Medical Language System to allow us to compute interpretation.
We need standard product identifiers and vocabulary. This means we need to
execute on the Commission on Systemic Interoperability recommendation to work
with the manufacturers of drugs, devices and test kits to achieve standard
identifiers in labels, packaging, and data outputs of devices and test kits.
Then downstream participants in the information supply chain if you will could
use this information within their local system much as the retail industry
leverages product bar codes. The key idea is to apply the standard at the point
of manufacture of the information instead of trying to apply it at the
inter-connections among systems.
We need to continue the development of the data interchange standards which
I think is one of the places we have made huge progress with and include in
that standards for authentication of an individual of their record and for
role-based authorization, et cetera.
We need metrics. We actually need to couple the development of standards
with the development of metrics to assess if the technology built to, or using,
the standard achieves the desired result when combined with people and process
in real health systems.
The final piece of the portfolio is to invest in the tools and services that
will facilitate the adoption and effective use of the standards by vendors and
If we take that portfolio approach we then need to drive to value. In
essence I think there are three management approaches we should take to that.
First would be to bring together standard developers and users to define and
test how information models, clinical data elements and value sets can work
together to achieve health improvement in the near term.
We then need to develop test beds to demonstrate high value use scenarios
and then finally create and maintain a roadmap of what needs to be done in each
lane of the portfolio, and when specific steps need to be completed to achieve
measurable interoperable health information goals.
Let me take an example of how the portfolio and drive to value approach
might work in the near term using end-to-end medication knowledge management as
that example. The goal would be to reduce the time and cost of deploying and
updating medication decision support content by linking the information
published by drug developers, the FDA, the drug knowledge-based vendors, and
later PubMed and ClinTrials.Gov.
The first step would be to extend RxNorm into a medication terminology
framework of ingredients, dose forms, strength, and classes such as the latter
coming from NDF-RT and then to link the NDCs and the products to RxNorm during
the FDA approval process. To take advantage of NDC link included in the
electronic labels, to link the commercial drug knowledge bases by including
their linking codes in RxNorm directly or indirectly through the NDC.
With that type of framework you could then work with test bed sites to use
concept matching algorithms and the information in those links to create
automatic linking from the site’s formulary to ingredient, dose form, or
strength or as many as are possible. You would then be able to link the site’s
preferred drug knowledge base to the formulary.
You then should be able to develop machine assisted site-specific filters of
alert rules for drug-allergy, drug-drug, et cetera, by user role such as
clinician or pharmacists. You should then be able to measure interoperability
with the elapsed time from issuance of an urgent update to its deployment in
the operational systems at the site transitioning the entire if you will
knowledge supply chain.
You should then be able to measure the healthcare improvement by performance
on the LeapFrog scenarios, the percent of alerts overridden by role, and the
percent of adverse drug events following an override, et cetera. That would be
a near-term view.
In terms of the long term as recommended in the NRC report, I believe we
need to support research into computable knowledge structures and models that
are needed to make sense of available patient data including computable
guidelines and approaches for comparing, assessing, updating, and integrating
them into a library for a given patient and to support systems that can infer
clinical conditions from raw data, such as inferring that the patient is
feeling more pain from the report of an upwards adjustment of the IV drip of a
pain management drug.
Because the clinical interpretation of data depends on the current state of
knowledge about medicine and about physiology and how people respond to
treatments and so, computable structures are important because they connect
medical knowledge to patient data in machine-readable and machine-executable
form. Thus, they can provide the abstractions for the healthcare provider and
for the clinician to help them understand what is going on with a given
patient. That’s an overview of my thoughts.
DR. WARREN: Thank you. I think we are all sitting here with faces that are
looking at the challenges ahead of us. With that I will open up to questions.
DR. SUAREZ: This has been a great session and thank you for your testimony.
I think it is very rich in recommendations and in perspective on what is going
on now and how it can be improved to ensure that we reach the goals that we are
I have a question and this is for Linda I think. In your presentation I
found incredibly useful the concept and the importance I guess of the
development of components across SDOs. I think that by itself should be a major
priority as I think we have been trying to harmonize standards from a variety
of standard development bodies that have multiple or different or sometimes
don’t have a reference model or a data model. My question is who do you think
should be actually developing those cross-reference models?
MS. FISCHETTI: I think that it should be a collaborative effort of the SDOs.
There has been a wonderful activity recently the SDO summit, which is now SCO,
a couple of my staff sit with them. I don’t know if you have them coming as one
of your panels. You do? You might want to pose the question to them and see if
they could have dialogue on it prior to them coming here, but I think it’s
really the leadership of the SDOs that should work collaboratively to embrace
that or reject that concept and then if they embrace it move forward.
DR. WARREN: I would just like to mention when the SDO summit formed, they
did contact our standards panel and so we are on their mailing list and things
like that. We are aware of the work that is going on and that has been very
fruitful to see the agendas and some of the products that are coming out of
MS. FISCHETTI: And Vish Sankaran from the Federal Health Architecture. He is
a permanent member of their group.
MR. REYNOLDS: Again thanks to everybody. It was kind of interesting. We
started off with Don Detmer this morning for some of you that didn’t hear and
then we ended up with Bill and in between were the people that are doing it. I
think it’s a great bookend actually to where we look at how we are trying to
drive it and we know with the bill that just came out people are going to want
to drive it faster.
We had a couple of good bookends reminders that we need to do this
purposefully and we need to do this structure. I think since we have this panel
together is a good dichotomy of those two. I would love for the three that are
here and Bill you can comment afterwards. I would like the three that are here
to take what Bill was saying and if you heard anything that Don said this
morning about this whole thing and communication.
It’s not just building the standards is more of making sure that it is
communicated so that people can use it in the right ways and draw some of that
together and see what you think. Each of you had obviously very clear and
precise things that you are doing and you know the space and then we heard
those two discussions about how we can think out of the box. Tell me what the
box looks like to you now that you sat and listened to that.
MS. FISCHETTI: What I would like to do is to defer to any of my panel
members who might have been here this morning.
MR. REYNOLDS: You can just stay with Bill. That’s fine. You can just go with
what Bill said.
MS. FISCHETTI: I’m writing notes. I actually have a pharmacist on my staff
who is working with GS1 to deal with some of that supply chain issues of how do
we label small doses and medications in such a way that it’s acceptable to
provider. That’s a role that I support. Manufacturers and distributors all the
way from manufacturing through distribution to the bedside and when you hand it
to the patient.
I’m sitting here taking notes that the gentleman on my staff, Mr. Tucker,
who works in GS1 needs to make sure that what he is doing is consistent with
some of the national medication terminology activities and so that we really
can take the national activities and drive it through all of our standards and
environments that we are working within.
DR. KUPERMAN: I’m not sure I fully digested what Don had to say this
morning, but it was very provocative. I guess I think what Don said this
morning, what Carol Diamond said and Bill Stead said are all the same thing
that we need to be driving to a purpose. I hate to say I’m old enough to, but
I’m old enough to having been doing informatics now for 20 years there is a
vision of what can be accomplished with robust informatics tools. We have put a
lot of tools into place and we have a lot of technologies that are available to
us that weren’t available to us 20 years ago and some not even 10 years ago. We
are not done and we need to leverage what we have to its fullest extent. We
have to be very thoughtful about what we can accomplish with that, what we
can’t accomplish with that. We have to keep goals in mind.
The current drive towards a quality improvement and cost reduction is a
great framing set of principles. That’s what we should be trying to work
towards. We should be very thoughtful about what we can accomplish and what
else we need to accomplish those goals and not just, I can’t remember who
exactly said it, but it may have been Carol, but not just to take what we have
today and make a whole lot more of it. That is probably insufficient. Standards
is a big part of that. It is not the only thing. That is kind of the outcomes
The communication piece of that is sometimes when I talk I say medication is
information intensive. It’s also communications intensive. Physicians talk to
physicians. Primary care physicians talk to specialists. Physicians talk to
nurses. Nurses talk to pharmacists and a lot of what we use “IT” for
is to facilitate that communication. I think that’s one of the points that Don
was trying to drive home with. To think about that explicitly is helpful. How
that aligns to standards I’m not sure I can wrap my hand around. Clearly
communication is one of the things that we do with computers.
Any solution how we integrate communication with information management I
think is something else we need to think about and also wrapped into that is
work flow and education of a staff, training them to do professional training
kinds of things and also in case it’s not tough enough there are multiple ways
to solve any solution. Some things we grapple with today whether it’s
anti-coagulation or pressure ulcers or medication reconciliations.
You can take clinical decision support and order sets and configuration and
you can mix and match those things to achieve the same end. You want to allow
that. You don’t want to prescribe how we should provide the best healthcare.
That balance between facilitating certain components but then allowing
healthcare organizations to mix and match them to their own particular problems
in the way that they like best is all part of it. Just some random thoughts.
DR. BELL: I was reflecting especially on what Dr. Stead said about the
potential for making a large investment and making things worse like somehow
building or getting silo systems adopted that wouldn’t actually interoperate or
where the information was not well. I want to reread his concepts again as they
were so information packed.
On the other hand you don’t want the perfect to be enemy of the good. I
think that we are in a really interesting time and I think it’s going to be
very important to monitor what is happening and to try to continue doing the
building the explicit understandings of how the specific information affects
clinical outcomes and iteratively improving that understanding. If we have to
roll out electronic medical records to -– sorry I don’t remember the exact
revision of the goal you said it was doubled -– yes, by 2014. We are not
going to have a chance to explicitly model everything so we have to prioritize
and monitor I would think.
Going out into the future the slide I had on SNOMED was that we are relying
on SNOMED a lot as our concept. I think it is the officially blessed concept
certainly for most of the fields in the structured and codified Sig is the
concepts that we are supposed to use. SNOMED has good coverage and a lot of
concepts, but when you look at then the links between those and how close it is
This is actually from another project we are doing where we are trying to
model lower urinary tract symptoms. I apologize for not having my remarks
–- you would want lower urinary tract symptoms in a particular guideline
included increase frequency of urination, urgency, painful urination. Lower
urinary tract symptoms is in SNOMED, but it’s not where you would expect to
find it in the hierarchy and the links to the things that are lower urinary
tract symptoms are not there.
In fact it is kind of a mess at least to my eye. Somebody arranged these in
a particular way that they did for some particular reason probably. When we get
down to the level of representing the concepts I just think we need to do a lot
more work on the resources that we have.
DR. WARREN: Bill, did you have any comment that you wanted to make to
DR. STEAD: I did much better at hearing the presentations than I have heard
the questions under the discussion. I obviously couldn’t see people so I don’t
know how what pieces of what I said people connected with or not.
I wouldn’t want what I’m saying to be interpreted as thinking that there
weren’t substitutive things we could do that would make a major difference in
the short term. My take on this is that if in fact we could pull together a
number of efforts that are quite different from one another to achieve a goal,
we could move extremely quickly.
I think what I described in the medication example if we wanted to, could be
done in 6 months with relatively little dollar support and with predominantly
— take COAA sort of first 100 days convening approach by the administration.
The thing about it that would be different is that we would actually be
managing each of the pieces to get what we need to take that step. I believe
right now we have very good efforts, but each of them in essence moves along
under its own steam and according to its own priorities. We don’t have a way to
bring them together in a way that will let us test which parts of them are
adequate and which parts are not.
I think that the most important thing we can do would really be to create
these test beds that in fact would let us drive to result and would in that
process tell us what to emphasize or de-emphasize in the various standards we
are developing. What the role of data fusion and visualization technologies are
relative to automation technology. What the role of maps collaboration and
social networking technology are to automation technologies et cetera. I think
in essence what we are going to discover is we need to make those very
different approaches work together and I believe that the majority of the focus
of our standards efforts today have been in doing the things necessary to
automate transaction processing if you will across the healthcare continuum.
That is an important thing to do but it is probably amongst the different
ways you can use the information technology. It is the most expensive. It is
the most time consuming and it is the least flexible. If we can begin to
actually set these different approaches side by side and drive them to targeted
improvement goals I think we can make a difference in the very short term. I
think it would be a different way of looking at the problem.
DR. KOLODNER: Thank you for excellent presentations. I think you outlined a
way of thinking and evaluating a complex standard as is being developed in a
feedback group that certainly in some of the areas that would be an important
adjunct to the way that some of the standards are developed now.
Gil and Linda, you both described your own and also Linda the heavy
involvements of the feds. We heard today some concerns that their stakeholders
who aren’t as involved and also concern that there may be dominance from some
of the stakeholder groups. For both of you how would you recommend that we get
a broader involvement and do you think that the involvement of 200 feds or of
Gil, your colleagues, has a balance there or is it really something where you
feel like you are there, but only you can nudge things towards the provider’s
MS. FISCHETTI: In terms of having 200 feds in the standards development
communities. We are not all in the same communities. We are distributed. We are
also not all in the same ballot pools and to tell you the truth we are not at
the point where we let’s say in VA know exactly what CDC want us to do on each
and every ballot. We are getting to that point.
That’s the predominance of the issues of balance within a voluntary
consensus organization. That’s of lesser importance to us. What we are really
focusing our coordination and collaboration activities on within the federal
space is the creation and development, doing the grunt work that has to take
place just to get something to the point that it can go out to ballot. When
things do go out to ballot, it is very important to do things not just in the
US realm to make sure that you do get that international opinion, that you do
get the opinion of all the stakeholders because that’s the final environment
that you are going to be implementing and using these standards with. We truly
do not want a predominance of government in especially that ballot pool. That
needs to broad and consensus based.
An example that I’ve seen is the professional society for physicians that
work in emergency rooms. What is that? Whatever it is. They actually have been
funding physicians to come together HL7 to work on electronic health records
system functional model profiles specifically for emergency room care. It’s not
just physicians. They have also done an outreach to the nurses, emergency room
professional association who has provided them with contacts.
I don’t know if they have actually supplied funds for the nurses to
participate in face-to-face meetings, but nurses have participated in the
creation of that work. I think the professionals that the associations are a
wonderful place to contact and to make sure that you get that cross and
disciplinary representation in the standards communities.
DR. KUPERMAN: I’m not sure if I have any magic answers here. The work that
Jeff and I did we were creating consensus across 20 groups that were brought to
the table for a particular purpose and even within that we had to just
negotiate some unique needs for certain groups. I think there is an art to
managing it once you have it. Creating it, I don’t know. There might be a set
of principles that are established that standards development organizations
should be sure to consider the following professional organizations in their
Given that the activities are so distributed, I think it is hard to really
assure it, but maybe a statement of principles of some sort and then make sure
that certain key bodies try to enforce that so whether it’s HL7 or CCHIT or
something else. That might solve 80 percent of the problem. My theory around
building consensus even within my own organization is it’s a rule there’s
always someone you didn’t ask. Some of that I think is part of the situation.
DR. KOLODNER: If I could ask Bill one question because Bill you certainly
summarized the report which laid out a very ambitious agenda and yet I hear you
saying that the issue isn’t that this is going to take tremendous monetary
resources but that we need to organize the efforts differently. Is this
something that you see that will require multiple federal agencies suddenly
drinking Kool Aid and simultaneously acting things? Is this something where
private sector is going to drive? When the report was put out there, how did
the group see this coming about? What would be the things that would really
enable this process to trigger off and move forward?
DR. STEAD: I am trying to couple the findings of the report which in essence
said that the kind of things we have been focusing on are at best one quarter
of the problem not four quarters of the problem or not even three quarters of
the problem. Then trying to extrapolate from that to how could we apply those
thoughts to the standards effort.
I think that if we could take the example of medication terminology standard
that largely exists. What we would have to do is use the force of national
persuasion to overcome the reluctance of the drug developers, for example, to
explicitly match their products which by definition they want to have be unique
to a link it to a standard terminology that in fact shows the relationships to
other products. We would have to do the same thing with the drug knowledge
I do think we have a major opportunity to convene and to try to show people
how they can thrive in something that supports the common good. Of all the
standards that is the simplest one. If we cannot in fact nail that, Lord knows
we cannot nail description of physical findings. I do think there needs to be a
convening aspect to this that I think frankly may well take not simply agencies
although they need to play. I think it is a national convening effort that
probably will take the President or whatever. Since it is a convening effort, I
think that is doable.
Then what you would in essence do is put natural language, concept-matching
algorithms, which are quite simple next to that knowledge map or knowledge base
to let you show that you can in fact make it simple to reflect those linkages
in installed commercial systems in hospitals and clinics. That would all focus
on that one standard. I think that is a much simpler task than solving what I
think of as the standard label, standard product identifier, and tagging
information at the point a manufacturer with its information content. I don’t
know what words connect to people. That is a much simpler task because we get
it right at the start and we patch it through the system.
The systems actually don’t have to be able to use it. They can just pass it
through. The next system that hits it that can use it will be able to use it
and eventually people can turn on the lights in the various systems. That is a
much simpler, much less to agree about than if I actually have to work out the
interchange standards that allow disparate systems to recognize what each other
is saying. The latter takes having all the people that build all these things
to actually agree on something. That is the process we have used. It is an
important process for where it is necessary. In that then we fairly got to have
all of the federal agencies required whatever we are trying to agree with
people use because that is a huge forcing function as you know.
I think there are some things we could do if we just focused in on some
targets that could actually be done very quickly and relatively non-invasively
to the various systems and so forth and the interchange standard et cetera that
we have been developing incrementally with quite a lot of success for years. I
am just arguing for some parallel paths here. I may not be making it clear how
different that idea is of standardizing and making sure that something standard
that can just be passed through whether or not the system recognizes it much as
a grocery bar code can be today.
DR. WARREN: Mike, a quick question?
DR. FITZMAURICE: First of all as Linda was going through the – by the
way before I get to the questions whoever put this together, it’s Jeff, it’s
Judy, have done an amazing job. I know you have expectations coming in and each
panel of presentations is just as good as the other one and the ideas that have
come out of this last panel has kind of blown your mind away. It gives us a
framework to think. I just get gushy when I think of how much brainpower there
is behind here and there are people thinking about how to put all this
Linda, when she was describing what Federal Health Architecture has done and
FHITSOP has done, one of the things through these meetings that we have had
with the other federal departments and federal agencies about what we are doing
and how we might do it better — VA and AHRQ have begun working together to
make the data elements in the HITSP use cases more visible, freely available
and more accurately reflect what’s in the chosen standards. A lot of this is
the expertise that VA already has in vocabulary. We have also had discussions
with NLM about how they then might be linked to NLM’s valuable resource. Just
because of this coordination aspect of what ONC has been doing under FHA, we
are talking more and more about each other.
I had a quick question for Doug Bell. In your SNOMED discussion you based a
question about harmonizing standards and about making choices. I am thinking
how do you choose between one code set versus another? Do you use different
standards for different purposes? Do you use the largest installed base
currently – the largest potential future use? The most comprehensive code set?
The one-stop shopping? The lower price or lowest cost for use?
Where it comes in particularly is on root of administration for drugs which
is why I’m asking this of you, Doug. Using federal medication terminology,
which could work, or SNOMED which could work on root of administration for the
short term and the long term. We are facing this problem in HITSP. It has been
a matter of discussion in FHITSOP as to what the best way to go is. Keep two.
Do one. Do the other. Do you have any thoughts on that? You did a nice analysis
of the compatibility but you didn’t get into discipline. I am just wondering
what are your thoughts on this?
DR. BELL: We have the luxury in this study of having it dictated to us by
NCPDP. NCPDP in the committee that came up with insertion codified Sig said
apparently after a lot of debate which is not documented, this is just what I
have heard from people that it shall be SNOMED for all fields in the Sig except
for dose form, which will be FMT. I guess because there it was that it matches
to the FDA dose forms essentially. It is derived from the FDA dose forms, but
everything else they said would be SNOMED including route.
We are constrained. How should that decision be made is really your
question. Ideally we should have some empirical processes of looking at the
adequacy of the code sets which we don’t have yet I don’t think. The FMT that
is being used I should say. We say FMT, but what is actually specified is NCI
thesaurus. That is what is supposed to be for the dose forms. That’s where it
is housed. They have put a lot more energy into the relationships I think into
getting relationships and ISA hierarchies inside their ontology. It’s more of a
real ontology than just a terminology, NCI thesaurus.
When you look at some of those frankly is like a clinician. You still
wonder. They actually might have gone too far. I’m just talking off the top of
my head here. There are a lot of ISA relationships. There are a lot of concepts
that they constructed to kind of group things together in multiple ways. I
think it needs to be evaluated.
DR. FITZMAURICE: Gee, thanks, Doug. I was hoping for the answer. Thank you
DR. WARREN: Walter is going to be the last one asking. Is it for Doug?
DR. SUAREZ: My question is -– and it has been a great day because we
have been talking about standards. We have been talking about standards and a
lot of things. Initially we have probably paid attention to standards related
to information exchanges and related to interoperability. Those were the three
things. Standards related to information exchanges and interpretability. I
think we have opened the back to other concepts around standards. I wanted to
go back to that — standards for health information exchanges and applicable to
or related to interpretability.
My question is about the three levels that I think I recognize with respect
to interpretability issues, which is interpretability within an organization.
You have a large system or even a small system with two hospitals. There are
interpretability issues. I think you mentioned a few of them. There is an
internal organizational interpretability. There is an inter-organizational
interpretability within an HIE, a defined HIE. There are exchanges between and
within that structure.
Then there is inter-HIE interoperability issues and exchanges between two
HIEs. In many respect would be focusing quite a bit on the inter-HIE initially
and looking at how the messaging across HIEs are interoperable. We have focused
– this is probably on the on the intra-HIE and much less on the
My question is are we missing some of that opportunity of going inside an
HIE or focusing too much on the inter-HIE? Should we begin to go more deeper
into the HIEs and even farther try to set some of the inter-organizational
interpretability issues and add into the agenda?
MR. BLAIR: We never limited the questions.
DR. SUAREZ: I’m not saying the questions. I’m saying actually the scope of
the work that is being done nationally. I’m not saying we limited the
questions. I’m just asking about should the agenda, our national agenda on
standards particularly Health Information Exchange Standards. Do these go
beyond inter-HIE and look into intra-HIEs and even go farther into the
MS. FISCHETTI: From a federal point of view we of course have our internal
departmental issues and then of course we are, any one of us, CDC, VA, DoD
military health service, Indian Health Service, we are the size or larger than
a lot of the HIEs out there. HIE to HIE communication. We are in all of these
HIEs so that’s not really a federal issue at all.
What we mean is to make sure all of the HIEs can communicate to the NHIN.
When the NHIN’s standard come together, you are using NHIN as the secure
connection between the HIEs will meet you there. What you do off of the NHIN,
we are not going to get involved with. Then we have the same issue that
everyone else does. We have legacy systems with disparate data that is not yet
influenced by this and we all need to get away from translating edge systems
because you lose integrity of the information.
This is critical information. As the NHIN standards and some of the stuff
that I mentioned earlier in terms of terminology models, information models
become standardized for the nation then we can begin to drill that into any
time we have a new system upgrade or we are acquiring or building a new system
which will get rid of that translation at the edge which over time will
incrementally move us all towards a better understanding of information as it
DR. WARREN: I think with that Walter if you could do it in about 30 seconds.
DR. KUPERMAN: That question is very relevant to New York State because New
York State is proposing an architecture where the HIEs will be able to
communicate with the central architecture, but the vision is that the
individual members of the HIE at some point in time to be determined, they
would have the protocols directly. Theoretically the vendors, the clinical
system’s vendors would implement those protocols and rather than going through
basically could render the HIEs obsolete. It is part of the vision and the
architecture. Now whether that’s a 3-year or 10-year or 20-year architecture,
we don’t know. At least in New York State it is explicitly been raised.
MR. BLAIR: Let me wrap this up. Let me remind you again that this is the
first of 4 days of NCVHS hearings. After that will follow our pulling the
information together to see what observations we could make that will be of
value. If there are clear recommendations, we will make those. You may have
noticed that NCVHS doesn’t always recommend.
The thing I want to reflect upon and of course I have to do it from my
perspective. In a lot of ways the testimony we heard today we tried to have
mainstream presenters and folks that represented as much diversity initially as
If you heard diversity in different viewpoints and perspectives, that was
very intentional. There are some of our testifiers that are at the leading edge
and this last panel represents a lot of the folks at the leading edge that are
looking at clinical decisions support and the tremendous benefits and values we
could harvest for the country if we just take the next step or two. There are
others that are focused on standards for their constituencies whether it’s for
ambulatory practice or laboratory practice or nursing practices. They have a
different perspective. There are others that are feeling a sense of urgency in
terms of moving faster. There are others that are afraid that the momentum will
be so strong they will be swept away with their constituencies not fully
protected, respected and recognized.
This is normal. This is typical of NCVHS hearings for over many years. When
we go to our second and third and fourth day of testimony, there will be more
of this diversity.
I’m going to ask that we all reserve our judgments, learn as much as we can,
respect all of the diverse opinions, and store it away, start looking for the
connections, but reserve your judgment because we will be hearing from all
–- I can’t say all of the SDOs. As many SDOs as possible in a day plus
written testimony to get the others that we can’t get in.
We are going to be hearing from a lot of users. Some of the users are
represented by vendors. Some of the users are going to be represented by
profession associations. Others will try to have represented in other ways.
Some of the other players in the health IT community.
I would like to thank the terrific group of testifiers that we had today. I
want to thank Judy Warren, my co-chair. I want to thank the rest of the
standard subcommittee because -– thank you, Michael, for the compliment
that Judy and I pulled this together, but we listened to the rest of the
standard subcommittee and many people from beyond the subcommittee that made
many helpful suggestions. Those are my closing comments. Judy, do you have
DR. WARREN: The only comment I have is you all have been patient as we have
gone over just about time thing and so it is my pleasure to adjourn the
(Whereupon, the meeting adjourned at 5:34 p.m.)