[This Transcript is Unedited]
Department of Health and Human Services
National Committee on Vital and Health Statistics
Full Committee Meeting
February 10, 2011
St. Regis Hotel
923 16th Street, N.W.
Washington, D.C. 20006
CASET Associates, Ltd.
Fairfax, Virginia 22030
(703) 266 8402
Table of Contents
- Call to Order, Review Agenda
- Quality Letter – ACTION
- Standards Letter – ACTION
- Overview, IOM Report on Learning Health System
- Subcommittee Report Outs
- IOM Report on Learning Health System
- Bridging SNOMED CT and International Classifications
P R O C E E D I N G S
DR. CARR: Let me call the meeting to order. First, let’s go around with the
(Introduction of committee members)
DR. CARR: Is that everyone? Okay, great. A little bit of housekeeping here.
It’s come to our attention that there are some scheduling conflicts of folks
who have to get out earlier than our original departure time of 3:00 p.m.
Therefore, I make the following recommendations that we get some food and work
What I want to do, we may move quickly through the letters and then move
then to the discussion. What I’d like to do, we have an IOM discussion of
learning health system, but I actually have put together a PowerPoint building
on our retreat in December, November, to just make sure what are all of the
things. We heard yesterday about the HHS strategies, we know we have our
legislative mandates, we’ve got PCAST.
So, I tried to put that together to say here are the things that we’re
accountable for and what are we doing, and then, how do they fit in. And, I
think we can work the IOM into that. So, I don’t know if someone could give a
call over to Chuck Friedman because he wanted to be here, and my guess is that
we’ll, we may be able to get to this sooner.
PARTICIPANT: Around when would we do that? At noon?
DR. CARR: Yes, or earlier. I mean if we go through the quality letters,
pretty straightforward, standards.
PARTICIPANT: I can e-mail him.
DR. CARR: Yes, just say we’re going to be moving it up. So then, we would
have that discussion and be finished up by 1:00 for the SNOMED discussion. And
then, we could adjourn.
MS. GREENBERG: Okay. So, with your slides, we would have the subcommittee
DR. CARR: We’ll have the report. We’ll use that as a kind of working
opportunity. So, let’s get right to it. Just to review, we’re going to vote on
two letters today, and our voting means we have to have at least ten people
vote and for to pass.
PARTICIPANT: A quorum is ten.
DR. CARR: A quorum is ten, and do we have ten? I think we do. (Counting)
Ten. No one may leave the room. All right, so it’s very important. And when we
vote, the voting is by, passing is by simple majority. All right. Begin
Quality, take it away.
DR. MIDDLETON: We have taken the input from yesterday’s discussions and
deliberations and drafted a revised quality letter which is, copies are at your
places. What I would like to do for the next few minutes is review our
discussion and the changes that we made to yesterday’s version so you know how
we reacted to all the input and what edits we’ve made.
So, what I plan to do then is just focus on the discussion and the edits and
not read the rest of the letter again, leave that with you all as having
already been reviewed and read. But, you can see here, this is the track
changes version, how we addressed the very first issues that came up. The focus
of this quality letter, as you recall, is on a roadmap for development of a
quality measure infrastructure and improved measures for the consumer focus
point of view, the value focus point of view, and then system efficiency and
In observation one, we suggested we should focus new quality development on
a patient-centric point of view with development of measures that matter for
patients from the patients’ perspective and the like, as you recall. There was
some discussion on whether or not this should be focused on the consumers’
need. I think, Walter, you might have suggested it or consumers’ perspective.
After some discussion, we left it as focusing on the consumers’ perspective.
We’d like to keep it perspective as it is more inclusive than need, we thought,
and it reflects orientation with respect to value assessment. So, we think that
the original language, in this case, is probably okay.
I don’t know if I should go through it all and take questions at the end or
DR. CARR: No, why don’t we go through point by point.
DR. MIDDLETON: Okay.
DR. CARR: Any questions on that? Okay.
DR. MIDDLETON: There are none. I’ll go ahead. In the first observation
bucket, we also tweaked the language to try to reflect the discussion again
which was very helpful. And, you can see the revised text here: The testimony
is focused on the consumer’s desire for information that is relevant to their
personal health needs and their preferences in a way that is easy to access and
understand. Obviously, the preferences and a value orientation was central to
our thinking about this bullet, so calling that out very early in the
observations was very useful. That was a good comment.
Since patient engagement is crucial to their participation in managing their
health, we recommend that the paradigm be shifted to focus on the consumer’s
perspective. This is the orienting sentence, if you will, that justifies the
rest of the observations and then the recommendation to focus on the patient’s
perspective in a more direct way.
Currently, most quality measures are directed toward providers and are not
as useful to consumers. In addition, they generally do not encompass population
health which is central in the emerging models of care delivery, etcetera. So,
that wraps up the edits and revisions to the observation paragraph, and that
leads to the first recommendation: prioritize and fund development of measures
that are meaningful to consumers. We tried to simplify this statement to be
much more direct and punchy, if you will, to focus on this new paradigm and
orientation for measure development.
Then, the justification –
DR. CARR: Blackford, we’ve got a couple questions.
DR. MIDDLETON: Oh, sure. Please. Mark?
DR. HORNBROOK: Blackford, just a question. In my experience, there’s
something about HIPAA that is not very friendly to certain cultures, that is,
that privacy is an individual phenomenon. There are some cultures where privacy
is a family phenomenon, and there are certain cultural beliefs about who has
the right to deal with health care information. And, western-based privacy
notions are disrespectful of that. Do we care about that?
DR. MIDDLETON: We absolutely do care about that. I think that’s the right
DR. HORNBROOK: Well, I mean, it’s the question is, should we begin to think
about broadening the perspective from an individual consumer to the notion of
multiculturalism and the fact that some cultures have very different ideas
about what privacy is all about and what data is all about.
DR. MIDDLETON: Yes. You’re actually touching on two important points. We do,
of course, think the end user or the consumer’s perspective on values which may
subsume, frankly, cultural orientation to value, etcetera I think is important.
And, it’s not mentioned explicitly, but the notion of individual patient
preference and values is. So hopefully, that might be enough to open the door
to that line of thinking.
On the second point about the privacy orientation and Leslie’s comment
yesterday, as well, we thought that that was very apt and apropos. We thought,
though, that all of our letters actually have strong recognition and respect
for privacy under them and to annotate each or each point within the letters
with a bullet on privacy, you know, seemed to be not the right way to go.
DR. CARR: Well, I would add also that our observation was, we were talking
about the creation of the measures and that the privacy would be more in the
application or the derivation or the use of the measures. So, I think that’s
where we left it.
DR. STEINWACHS: Just a small point. I sometimes find it useful to talk about
patient-reported outcomes. And, you list a series of outcomes there. And, I
don’t know whether it makes sense to you, but I always worry a little bit that
people report for patients, and so we, omissions, others say I know what your
functional level is, I know this, that, and the other. And so, you know, just a
friendly amendment, and you can refuse it, but you could have like the first
one, patient-reported experience or something like that.
DR. MIDDLETON: It’s a good thought, Don, and what we thought, we did bring
that up. It did come up. And, we have the phrase at the end of that list of
type of different data, patient-generated data. I’d be happy, if it’s a
group-wide consensus, to amend that to be patient-reported data or
patient-reported outcomes. You know, either way. But, the patient-reported data
concept is there.
DR. STEINWACHS: Okay. So, I see the patient-generated.
DR. CARR: Leslie?
DR. FRANCIS: I guess the reason I’m a little uncomfortable with what Justine
and you just said about the resolution of the privacy thing is. Actually, I
think that one of the really difficult points is that privacy tends to get
rolled up as a barrier and not thought of as a much more nuanced, new set of
So, the example of the comment that Mark made, the reason I was bringing it
up isn’t to roll up the barrier point, but to say that if there are new ways of
measuring quality, sure this is about the measurement, but once you’ve done the
measurement, you’ve got a certain way of using the information that we may want
to be, as we try to build new measures, be thinking about whether there are new
and different kinds of issues and ways of thinking about those issues that
should be developed along with the measures. That’s all. I mean, not, and
that’s why I’d asked how radically different these measures are because this is
really meaning to get information from patients and meaning to get very
different kinds of information. That’s all.
DR. WARREN: You know, we talked a lot about this, that this was all about
just developing a measure. And, when I hear you talk, Leslie, it almost sounds
like there needs to be another recommendation under observation 1. And, the
recommendation would be that as measures are developed, that privacy issues on
their collection be considered.
DR. CARR: Let me just put another thing. I think that when we have new
measures, we need to be thinking about the standards for that, we need to be
thinking about the privacy, we need to be thinking about the population. So, I
guess it’s more of lumping and splitting, as Paul would say, if he were here,
that I mean if we’re going to annotate about that we would always want to be
aware of, you know we would be cognizant of privacy issues, I think we should
be cognizant of standards issues. And, my inclination is to not, to say that
that’s implicit in this, and that the letter is directing toward this new focus
of the patient. Matt?
MR. QUINN: I’d just say, I mean, the real point to this is, the real point
of these recommendations are an orientation. And, there are particular issues
that are going to be associated with the current measurement paradigm as well
as, as we expand to new ones. There are a whole host of issues that we could
get into around risk stratification or care coordination, all of these, and
including, and all of these should be considered when we’re developing and
putting forth the actual development process of measures.
I think that we, our main message here is to focus on we need to orient from
what we’re doing today this new paradigm shift. And, to get into some of those
details would potentially distract from that main message.
DR. CARR: Raj?
DR. CHANDERRAJ: I gather with the standardization of this measure, does it
include potential abuse of a patient using this measure to, over quantifies
disability or dysfunction.
DR. CARR: Right, and so what we came back to yesterday is we have a gap in
the information from the patient about the patient. And, we believe that that
gap needs to be addressed. And, all of the comments around the table will be
part of that implementation and rollout but would make this a very much longer
letter when, in fact, it’s just a directional, first pass one of many letters
DR. WARREN: Yes, another way to look at it from trying to remember our
conversation from yesterday afternoon is, what we’re trying to do here is to
identify the thing that needs to be done, not all the processes, procedures
DR. CARR: The what, not the how.
DR. WARREN: And, if we start doing that, then this letter will never get
published, and it will be 8,000 pages. But, we’ll have to do more hearings and
stuff because people will always come up with, well, you didn’t think about my
standard or terminology or high risk.
DR. CARR: But, I think this raises an important thing in terms of how we
think about our letters. And, I think I was struggling a bit with the standards
of it, but it’s about the what and not the how. But that we may want to have
sort of standard language that goes in all of our letters that underscores the
relevance of how the implementation unfolds with regard to standards and
privacy. But, I think for now, Blackford, you want to continue?
DR. MIDDLETON: We actually spent most of our time yesterday in the afternoon
on this very observation and recommendation because it is, it’s very deep, and
there’s lots of attendant issues which is, I think as Justine’s pointing out
though, it’s really on the implementation side. Once we get directionality, so
let me proceed.
On the observation 2, focus on health care value, we tried to modify the
early paragraph, sort of the justification paragraph, to reflect some of the
issues that we heard yesterday. So, there was a consistent theme in testimony
across all stakeholder groups of the challenges in measurement of health care
value, quality, premium, and cost.
Consumers want information and transparency about cost and quality to inform
their health care decision-making. Costs include not only the cost of the
procedure or visit, but also include cost of workdays lost, travel, childcare,
and caregiver time. One of the, it was sort of reflect discussion that occurred
about the other related costs that are actually impactful to the patient beyond
just the costs of care itself, opportunity costs, etcetera that are going to be
extremely meaningful, perhaps more meaningful to the patient than just this
idea of getting the cost of care right. So, it’s really both points we tried to
make sure we were reflecting.
There was no change to recommendation 2, fund research and development of
improved assessments of a value health care based on measures and information
about cost and quality that are relevant to all health care stakeholders. A
little simplification in the following paragraph. Given the delineation of
costs above, we took out some delineation of costs below here.
Now, observation 3 was the focus on accountability and care coordination. We
tried to modify the opening paragraph here again to reflect the rich discussion
about this observation. New measures and information are needed to hold the
health care system accountable for the things that are important to consumers.
Patients are often transferred between providers without complete information.
Patients themselves often do not have complete understanding of the treatment
plans or follow-up instructions. This results in inadequate follow-up,
excessive utilization, and sub-optimal outcomes.
In a way, this justification or clinical motivation was missing from the
introductory paragraph on this observation. We thought adding these details
actually framed the observation more usefully than it had been before based
upon the discussion.
DR. FRANCIS: I would just want to point out that standards do get mentioned
a couple of times. You don’t mention how to do it, but you do mention that it
needs to be standardized.
DR. CARR: Standardized and standards aren’t the same.
DR. FRANCIS: But, it’s going to sound like there needs to be a common set of
ways of doing it. And you do mention, I love it, a reliable and transparent
method must exist to track patients. I love that, but that’s also going to
sound like, somebody’s watching me. Right? And, that’s the only, that could
scare somebody. Right?
DR. CARR: Are you recommending respecting patient privacy?
DR. FRANCIS: Yes. Or something like that. That’s the only – track is a weird
DR. CARR: Mark?
DR. HORNBROOK: It gets back to a very basic point that there are some
patients who want to be their own navigators and want that privacy, want that
control, want to be respected for being the person in control. But, there are
other patients who want their runner as their agent. Very different purpose of
DR. FRANCIS: I would even put, leave just the word respectful. That would do
it because I’m not trying to say there should be a single set of way of doing
it but that those issues just need to be thought about. That’s all.
DR. CARR: Respectful of patient preferences?
PARTICIPANT: And privacy.
PARTICIPANT: Patient preferences or patient privacy?
DR. CARR: Well, that’s what I’m asking because what we think of as privacy,
I mean, it’s ultimately the patient preferences which might be privacy, might
be, you know, preferences I think are more encompassing.
MS. MILAM: Justine, you’re saying to put in preferences instead of privacy
and confidentiality which would be inclusive of privacy but other things too.
DR. CARR: Yes, right.
DR. MIDDLETON: We’re quibbling a little bit up here about the word track.
Let me just see if the first amendment suffices, Leslie, actually to keep the
ever-present notion of patient privacy and confidentiality in mind I said,
respecting patient privacy and confidentiality across delivery settings and
PARTICIPANT: That’s quicker. Good.
DR. MIDDLETON: Thank you. How about track? Is that okay?
DR. CARR: Hearing no objections.
DR. MIDDLETON: Hearing no objections, Madame Chairwoman, thank you very
much. So, the recommendation here was just to simplify it a little bit
grammatically. Fund research and development to provide information and
measures that enhance the ability to assess accountability and care
coordination in health care. We simplified the last, we struck the last
And then, observation 4 is the one about the system lifecycle of quality
measurement and etcetera. The original title was a mouthful, so now it’s
Observation 4: Coordinate Development and Assessment of Quality Measures. A
little plain vanilla by my view, but maybe it’s good enough.
This is the one that, the lack of measure coordination reduces comparability
of quality reports and increases the measurement burden on providers.
Coordinated measurement activities would allow providers to collect health care
data once to support multiple uses, a common NCVHS theme including care
delivery, quality improvement, public reporting, paper performance, maintenance
of certification, and other activities. We kind of got the laundry list of
stakeholders in there.
Recommendation 4, convene quality measurement stakeholders to formulate a
strategy to coordinate development of quality measures using standard metadata,
i.e. names, definitions, and representations, that improve the efficiency of
data re-use. Okay. The emphasis paragraph is unchanged, and then the closing
paragraphs are the standard NCVHS boilerplate. And, there were no further
changes. So, if there’s no further commentary, I’d like to submit this letter
for the Full Committee’s review and consideration as a letter on behalf of our
Chairperson to the Secretary.
DR. CARR: Okay, so Blackford made a motion on his own letter.
DR. MIDDLETON: I make a motion to submit —
DR. CARR: And, it was seconded by Mark. All members of the Committee in
favor? And, all opposed to the letter? And, abstaining? One abstention. Okay.
DR. MIDDLETON: Thank you very much.
DR. CARR: All right. Very good, well done. Judy, you’re up. Can I just ask
that we put the view on like 150 or 200 if we’re going to have it up there
because I can’t see anything?
DR. WARREN: So, I’m going to do the same thing that Blackford did and go
through the edits from where we ended yesterday. We added this sentence into
the very beginning that NCVHS has responsibility for recommending standards for
HIPAA transactions. We felt this was needed to give a setup for the rest of the
letter since that was one of the things that we did. And then, we go on to
specify what our task was through ACA. Any questions on that? You like it? You
want it to stay in there?
Okay, moving right along. Now let me, I’d like to come back to this one
because I have a fix for this. This just came up this morning in our
subcommittee, so I’ll come back to that comment. We had a request to provide a
header, put a header there called The Organization of the Letter, so that the
reader could really see what they were doing. So, we’re really doing three
different things in the letter. So, I added the header, and I clarified
overarching themes to include, concerning the selection of operating rule
authoring entities and the implementations of standards and operating rules.
DR. CARR: So, I would just say, I think another way to do that, and not to
perseverate upon this, but we can just say, the problem, the approach, and then
the observations speak for themselves. I think it’s smart to, as you have
above, to say here’s the problem, and then to understand the specific just have
a header, the approach, and then the observations. So, we have overarching and
DR. WARREN: So, you want the title to be, The Approach?
DR. CARR: I would put a problem, on line 54, I would have something that
says, the problem, it’s what we’re trying to address. And on line 64, I would
say, the approach, we convened yada, yada, yada.
DR. WARREN: I don’t know where you are.
DR. CARR: Oh. Okay.
DR. WARREN: I want to, 54, what I’d like to save to the end.
DR. CARR: Okay. It’s too much wordsmithing, so let me just leave it to you.
MS. GREENBERG: I would recommend against that header.
DR. CARR: Okay. Against what?
MS. GREENBERG: The current header.
DR. CARR: Organization of the letter, yes. I think that that’s not helpful.
I think telling someone here’s the problem, here’s the approach, here’s the
observations is okay. But, organization of the letter, I wouldn’t use either.
DR. WARREN: So, you want me to delete this?
DR. SUAREZ: Yesterday, there was a paragraph based on the testimony.
DR. CARR: I would call that the approach, well oh no, those are the
observations. Yes, delete that.
DR. WARREN: Okay. So, we’re back where we were yesterday.
DR. SUAREZ: I think what the interest was was to try to highlight the
organization of the letter. We thought putting the title would make it, but I
think it’s already there. The organization of the letter is structured there,
so I don’t really know the need for adding that title there.
DR. WARREN: So then, the next piece comes down in our recommendations. So,
we did not change the first four recommendations. We did change recommendation
1.5 in that it was suggested by the Committee yesterday that we take this last
sentence and make it a footnote. And so, I’ve done that. So, there was no
changing in wording of that.
And then this morning, we heard in our subcommittee concerns from the
banking industry that making this a HIPAA transaction then puts them in the
position of possibly becoming a clearinghouse under HIPAA. So, we felt that we
needed to add a recommendation here that we consider the implications of
whether, as a result of the adoption of the health care EFT standard, banks
would become health care clearinghouses as defined by HIPAA.
And, the reason we need to do that is that we really need to look at what
clearinghouses are, how they function, as well as what happens when banks come
into this. And currently, banks are already functioning in this role without
the HIPAA designation. So, any questions on that?
DR. CARR: My question is, Marjorie, where did she go? Consider, do we often
tell the Secretary to consider something? I mean, is that an actionable —
MS. DOO: Yes. And then, you would make a decision to what would happen after
that consideration with you’re telling us to do it, we’ll do it. And in the
regulation, you would see commentary to that. We would address it specifically.
DR. CARR: Okay.
DR. SUAREZ: To give you a quick example, on health plan id, we didn’t
recommend a specific level of granularity. We said, you know, consider the
DR. FRANCIS: I’m just not clear on what that says. Does it say to the
Secretary, you should consider whether banks are clearinghouses —
DR. WARREN: Health care clearinghouses.
DR. FRANCIS: Right. Or does it say, we think they are, and you should write
a rule that deals with their —
DR. WARREN: We think it’s ambiguous, and it needs to be carefully
DR. FRANCIS: So, maybe it would be better to say what you just said because
it’s, currently –
DR. WARREN: Well remember, it’s a recommendation.
DR. FRANCIS: Well, what I was, consider the issue of whether banks are
clearinghouses and what the implications of that might be indicates that it’s
MS. GREENBERG: I don’t think this belongs as under 1.6. You’ve got these
recommendations adopt, adopt, or you know, define. I think this would be more
of a comment after those that says, this needs to be considered in light of
these recommendations, whatever, needs to consider whether as a result of the
standard, banks would become health care clearinghouses.
Now, I don’t know whether you, were you talking about possibly having a
hearing on this? Is the Committee planning to provide any information? No. This
is just, so I just don’t think this belongs as a 1.6. I think it could be a
comment afterwards, or it could be in the observations too.
DR. SUAREZ: I think if you could maybe explain what this is because, you
know, hearing the comment from Leslie, it might not be clear the way it is. So,
the situation is this. If we adopt the standard the way it is recommended,
there is a possibility that in some cases, health plans will submit the
transaction to the bank in a non-standard format, and the bank will convert it
into the standard format. That is the definition of a health care
So, the implication of the recommended standard is that in some instances,
banks would become health care clearinghouses of the health plan. Banks are
already health care clearinghouses in many other instances. This is creating
another instance where banks would become health care clearinghouses. That’s
the point that we’re trying to, and so the concern, I guess, with the request
is really, has the Department considered the implications of the fact that by
virtue of adopting these standards, banks would become, in some instances,
health care clearinghouses.
MS. GREENBERG: Does that mean maybe they shouldn’t adopt these standards?
DR. SUAREZ: No, no, no, no. That’s a normal —
MS. GREENBERG: What do you do with that? I guess that’s where I’m a little,
I mean, what do you do with those implications? I mean, does it speak against
adopting the standard? Or, does it mean that then you’ll need some other
standards or you’ll need some other policies? I guess I wouldn’t know what to
do with that.
PARTICIPANT: Judy? Do you want to speak?
MS. MILAM: Yes. It’s my recollection, probably five or six years ago, the
Department addressed this issue at that time, not based on this new standard,
but at that time people were raising issues as to whether banks should come
under HIPAA. And, I recall the industry felt very strongly that the banks are
separately regulated by a whole different structure under the FTC with
Gramm-Leach Bliley that there were adequate protections in place, and they were
not pulled under HIPAA. So, I think the Department has already spoken that far
in the past, and we might want to think about that as we craft this language.
DR. SUAREZ: But again, you have to answer that with banks are already, in
many instances, health care clearinghouses. So they, it’s not that anybody is
forcing them to be clearinghouses. They can choose to be clearinghouses, health
care clearinghouses, and provide the services that health care clearinghouses
provide to payers and providers. So, in addition to that, they are subject to
all the regulations. So, the implication here is really banks would become
health care clearinghouses, and so, it’s just highlighting that point is what
we’re trying to do.
DR. CARR: Okay. I have Raj and then Bill.
DR. CHANDERRAJ: I think banks, like Walter says, are already clearinghouses.
The issue is, I think, they have, although they come under FTC regulations,
Jan’s point that they’re already doing this and it’s only facilitating them to
do it in a standardized format. That’s all we’re saying, I think. That’s what
the implication would be.
DR. SCANLON: I think the way, from this morning’s discussion, some banks are
clearinghouses, but they’re clearinghouses voluntarily. They’ve decided they
want to be a clearinghouse. The issue here is that the creation of the standard
will mean that a bank that is not currently a clearinghouse but is processing
electronic funds transfers is going to become a clearinghouse because the
information coming to the bank is in a non-standard form, the bank converts it
to the standard form, and by that act, becomes a clearinghouse.
What this is actually asking is, does that matter at all in terms of a bank
that is only a clearinghouse to that extent? That would be the issue because
it’s not a clearinghouse doing other functions. You know, there are other
clearinghouse functions that this bank that has been doing these things in the
past is not going to necessarily take on, it particularly will not take them on
involuntarily. But, what we’re trying to deal with is the involuntary nature of
doing business that you’ve been doing in the past. That’s all it is.
MS. GREENBERG: And, I heard that from our representatives that they would
prefer that that not happen. But I mean, that is an unintentional consequence.
But, is there any way to avoid it?
DR. SCANLON: But, I think the issue for asking for there to be consideration
is it an unintended consequence that has significant consequences? I mean, is
it, it’s an unintended result technically. But, is there material consequence
of this? Is it going to lead to something that imposes an obligation on banks?
The point was raised this morning that there’s a need sort of for some kind of
clarification so that it’s clear about what this really may entail.
So it was, I think more, the discussion this morning is more understand the
consequences and don’t let it fall through the cracks, okay, as opposed to this
is necessarily something that’s going to create this incentive for banks to
say, I’m not going to process electronic funds transfers from health plans
because that would be very disruptive. I mean, it doesn’t seem like we’re
talking anything like that. What we’re talking about is clarification as
opposed to anything else.
DR. CARR: Walter, but I was going to say, do you have a recommendation for
an edit to the letter based on that?
DR. SCANLON: Well, I was comfortable with that earlier this morning. I mean,
and I think Marjorie’s issue about whether it should be a recommendation or a
comment, again, I’m not sure of the magnitude of the difference there.
DR. CARR: Walter?
MS. GREENBERG: Well, it’s not a question of whether. It sounds like as a
result of adoption of the standard, banks would meet the definition of a health
DR. SUAREZ: No, no. Let me clarify one point because I think that, today
banks provide the service of converting a transaction, you know, an EFT sent by
the health plan today. So, that’s what’s happening today. This was not a HIPAA
MS. GREENBERG: So, that didn’t make them a clearinghouse.
DR. SUAREZ: By virtue of making it a HIPAA transaction, entities that are
required to comply with this, which are health plans, have two options. One,
either they conduct the transaction that is now under HIPAA via a
clearinghouse, via an entity submitting the transaction in a non-standard way
to an entity that converts into a standard so that’s a clearinghouse, or they
themselves generate the transaction now. So today, it’s outside of the
regulation, and so they do all sorts of, well, they do different ways to do it.
But under this regulation, what this adopting this standard, it doesn’t
matter the standards, I guess, adopting the standard and establishing this as a
HIPAA transaction, it will require that entities that conduct the transaction
have to make a decision about how they do it. If they continue to do it the way
they were doing it, then yes. You know, the entity that received the
transaction and is receiving it in a non-standard format will have to be a
clearinghouse. If they decide to do it differently and they themselves generate
the actual standard and send it to the bank, then the bank is not a
DR. SCANLON: All right. I know where this is going to help would be to
insert, some banks would become, and to indicate the involuntary nature of it,
de facto clearinghouses. Okay?
DR. CARR: Judy, did you get that?
DR. WARREN: Did I get what you wanted, Bill?
DR. CARR: That was a language proposal.
DR. SCANLON: Some banks would become de facto health care clearinghouses.
MS. GREENBERG: Keep the first part the same?
DR. SCANLON: Yes.
MS. GREENBERG: But doesn’t that, some banks will become?
DR. SCANLON: Would become. Hopefully that gets at the notion that we’re not
talking about all banks, that it’s part of the relationship between the plan
and the bank that’s going to determine this. And it’s, this is the involuntary
nature is the de facto part.
MS. GREENBERG: Is the word “whether” really right there then? It’s
some would become.
DR. CARR: Consider the implications that —
MS. GREENBERG: As a result of the adoption of this, some banks –-
DR. CARR: Of the fact that, implications and the fact that.
MS. GREENBERG: It’s the “whether” that I was having problems with.
Just, if you change that to that I think, consider the implications —
DR. CARR: Of the fact that —
MS. GREENBERG: Of the fact that. It’s a little wordy, but —
DR. CARR: Okay, so Bill has said that, as a result, some banks would become
de facto health care clearinghouses as defined by HIPAA. And, I think and
you’re saying that that will never happen?
PARTICIPANT: No, I’m saying the legislation –(Inaudible- off microphone.)
DR. CARR: So, instead of “would”, do we want to say,
“may”? What if we restate it? What if we say, consider the
implication of the fact that as the result of the adoption of the health care
EFT standard, some banks may become de facto health care clearinghouses as
defined by HIPAA.
PARTICIPANT: Yes, I think that’s better.
MS. GREENBERG: And, if you like it there as a 1.6, that’s fine.
DR. CARR: Make it 1.7.
MS. GREENBERG: You need “of” there. Consider the implications of
DR. CARR: Of the fact, yes. Okay. Anymore comments on 1.6? Good job
everybody. Judy, what’s next?
DR. WARREN: I think that is the last one. Let me just scroll through to be
sure. Yes. So, now we need to go back up to the discussion that we had this
morning. So, it’s this particular paragraph. So, one of the things that we were
looking at this whole piece of defining the problem, etcetera. Walter and I
came in, and I haven’t had time to really come in and do this. But, we were
recommending a sentence, something like, and this is very rough.
DR. CARR: While you do that, can I just read what I wrote? It might not be
right, but just to say why this came about. I felt that a stronger statement
was needed as to why all this work is going on. So, rearranging the words in
the language up there I said, according to the US Health Care Efficiency Index,
the usage of the standard electronic remittent advice is only 46 percent today,
a full seven years after the compliance date.
And then the other fact, and maybe it doesn’t follow exactly, but the other
fact in there that I think needs to be identified is, if intermediaries are
excluded, the rate of adoption of electronic funds transfer is only ten
percent. So, I’m just trying to say, we were supposed to do this seven years
ago, and we’re not even halfway there. And, that’s why this is important.
DR. CHANDERRAJ: I think your point is well-taken. It brings out the problem
in a more clear definition.
DR. CARR: So, I may not have been precise in my language, and there’s some
additional thing that Walter pointed out in the adoption. And perhaps, I think
the language is there, it’s a matter of wordsmithing. And, I’m respectful of
the fact that if we feel the Executive Subcommittee can finish the
wordsmithing, to just punch out the language here. I think we’re done with
MS. GREENBERG: It shouldn’t hold up the letter.
DR. CARR: Okay, so would someone like to make a motion that this letter be
moved forward after the Executive Subcommittee completes the wordsmithing
around the problem statement?
MS. GREENBERG: Well, I think it should be, I think the motion needs to be
that this letter be approved, but finalization —
DR. SUAREZ: Yes, I make a motion to approve this letter with the final edits
from the Executive Subcommittee.
PARTICIPANT: I second it.
DR. CARR: Any discussion? Okay. All in favor? Any opposed? Any abstentions?
So, unanimous. Yeah. So, where are we now? We’re at 11:00. How incredibly
efficient of us.
MS. GREENBERG: I would just say we should do this final editing quickly.
DR. CARR: Oh, yes. Since we’ll finish early today, you know, maybe folks, I
have my language, and I’ll be happy to share it. What I would like to do at
this point is to use this time to now talk about what are all the things we’re
doing and how do they align with what we’re supposed to be doing. And you know,
can we just keep building on this framework that we have in the back of the
book? So, I’ll go up here.
DR. CARR: Okay, thank you everybody. So, I tried this morning to put
together what we have been talking about, what are we doing. And, this is sort
of an overview, no pride of ownership, but it’s just directional to help us
keep thinking about what are we supposed to do.
So, we talked about identifying the focus and themes for the next 18 to 24
months. And they should align with our charter, our national priorities that we
heard yesterday from Jim as well as legislative mandates, emerging issues in
health and health care, and then be aligned with resources and timeframes of
Committee members and meetings. And then, I put as a footnote here also
Blackford’s recommendation that we speak to our partner agencies on how we’re
doing, are we doing what we need to do.
So, the first slide is just from our charter, the objective and scope of our
activities. So, the Committee shall assist and advise the Secretary on health
data, statistics, privacy, national health information policy, and the
Department’s strategy to best address those issues and shall also assist and
advise on the implementation of administrative simplification of HIPAA.
And you know I think it’s worth, especially given the hearing on Tuesday, to
remind ourselves of our construct from shaping a health statistics vision for
the 21st century where we talked about the population’s health,
community attributes, context, place, and time. Also, from that document, the
overlapping circles of provider, person, and population. Also, this was just
from our summary last year, but to remind us of all of the work that’s been
done on the various legislative mandates and reports to Congress.
And now, our current initiatives. This is just a summary of the HHS
priorities, and I’ve selected the ones with stars because they seem to align
with the work that we’ve done. And, I thought one way of approaching this is to
take those initiatives and the description and talk about what it is that we’re
doing and that this isn’t the totality of what we’re doing. And perhaps, let me
just take you through the rest of the thinking, and we can come back to this.
The other is emerging data sources, and here we have the Community Data
Initiative, the PCAST report, and then the IOM digital health data from EHR via
meaningful use contributing to rapid learning system. This is just a slide that
we had put together on communities as learning health systems. And again, we’ll
come back to it in a minute. We have the IOM report and what that said and then
our Affordable Care Act assignments and how we advise the Secretary.
So, with that construct, I thought this is a good way for us to talk about
what is the work ahead on the subcommittees and link it back to why are we
doing this, and does it align with what we’re doing? So, let me just stop there
and take any comments about whether this is a fruitful way to proceed or any
other suggestions. If you love it, feel free to say that too. Anybody? Anybody?
DR. CHANDERRAJ: What’s the output? I mean, in terms of reviewing at all
those things in practical ways, what meaningful change has this Committee
DR. CARR: So, we have two documents that are on the website. One is the
50th anniversary, and the other is the 60th anniversary.
So, as a new member that would be great, you know, to kind of review that. I
think we won’t have time today to talk about what the output has been. But, I
was hoping that this would be helpful, especially to you and the other new
members coming on to think about what is the work we do and how we do it. So,
I’m going to hold on that. Yes, Mark?
DR. HORNBROOK: Are you thinking at all about the FDA in this conceptual
framework? I mean, the FDA has the Sentinel Network, 100 million people
conceivably having their medical records looked at continually.
DR. CARR: Right. So, as we talk about the work that we’re doing, that’s a
piece of it. You’re right, I haven’t because there are many others, obviously,
work plans, but that’s an excellent point. As we talk about the acceleration of
process of scientific discovery and improvement of care and then the work of
the IOM in trying to create this learning health system from the digital health
information available including the FDA and all of that kinds of stuff.
DR. HORNBROOK: That stuff is becoming really real now.
DR. CARR: No, I know. That’s why I want to, to me it was very important that
we have this discussion to integrate the threads that we’re developing as
committees. Other suggestions? Mark?
DR. HORNBROOK: The other thing is the Ingenix contract to build the patient
data warehouse for Medicare/Medicaid and any other source that wants to
contribute it, and Ingenix has approached the Mini-Sentinel about using the
same data model so that the two data systems could conceivably, with some
vision, be pooled.
DR. CARR: So, let’s go to the report outs from each of the subcommittees.
And, as we hear what it is and if we can identify what it links to and what
we’re going to do, I think that’s a way to begin to populate this. And then, if
there are other things that might represent work we feel that we ought to be
doing or that, you know, that we are doing that doesn’t fit in this construct,
that might be a way to do it. Yes, Blackford?
DR. MIDDLETON: Do you want me to just launch in?
DR. CARR: Sure.
DR. MIDDLETON: For Quality, we actually spent, as you know, most of the time
talking about the letter. But, I think we did touch upon PCAST the day before,
and everyone’s been reading it and thinking about it. So, the actual action, I
think, you know, that the NCVHS might take upon itself and perhaps Quality
could start, but it’s clearly going to touch Standards, as well as Privacy and
Security, it’s a broad issue. What is the right approach to the development of
standards for metadata?
We’ve talked about standards at every other level and every other way,
etcetera, but you know, this is the new world. And, there’s work underway with
USHIK and other kinds of repositories, but I think we need to inform ourselves
and perhaps develop recommendations on where does that standardization process
fit into the country’s standardization process for data?
DR. CARR: I think that makes a lot of sense. What I want to do today,
though, is that each of you coming out of your subcommittees have a plan of
things that you’re going to do. And, I want to get those on here. So for
example, if we talk about the PCAST report, we’re going to have a letter, you
know, with input from each committee, right?
DR. CARR: So, that’s something, that’s our to-do with emerging data sources
and why we, NCVHS, are involved in it.
DR. MIDDLETON: I think we can go further and say, we wish to hold hearings
whenever it fits into the agenda on standardization and security and privacy
and issues related to metadata.
DR. CARR: Okay. So, that would be potential future role, potential because
I’m trying to differentiate what we said we’d do today, as opposed to what we
DR. MIDDLETON: Well, I think if we attend to the standards of US health care
data or meeting to attend to the standards for US health care metadata, pure
DR. CARR: Right. Okay. So, speaking for Quality, what is the next step? We
had your letter today. Is Quality planning any hearings?
DR. MIDDLETON: No, no hearings are planned. But, I propose we consider joint
hearings between Quality and Standards, at least on the issue of metadata for
health care data.
DR. CARR: Okay.
MS. GREENBERG: There is a way to make that bigger too because I think the
rest of us otherwise are kind of out of this conversation.
DR. WARREN: Yes, because I think the metadata tags, that’s where Privacy is
hoping that solutions will occur. So, we may even take a look at could we have
half day hearings for the Full Committee? Because I think everybody is going to
be involved, and there may be meta tags that have to do with population
DR. CARR: Okay. So, in addition to the letter with input from each
committee, joint hearings on metadata.
DR. MIDDLETON: I mean, it’s the thing that Mark just mentioned which is
going to be so incredibly powerful and important if there are emerging jointly
used information models and the like. You know, it goes beyond RIM. This is
actual practical applications of standards and models that is going to be very
important for getting use of the data right.
MR. QUINN: So much of the PCAST report was about metadata. And, it
prescribed a very futuristic approach to that, and there are dimensions that
are very closely related to quality. I think that this PCAST summary that we
put together or comments that we put together are going to be heavily
influenced by the letter that we just went through. But you know, this is where
we can start talking about the relation to privacy and standards and these
other areas, and talking about a practical path based on, for example, past
hearing that we’ve had.
DR. CARR: So, here’s a question I would like to put out because we heard
that there are going to be hearings that ONC is holding next week or so. And
clearly, ONC has a lot of activity and a letter coming out in April on this,
and they have already assembled a team. Is there, do we know today whether
there’s, are we going to be redundant? Or, how will we utilize our resources so
that we add value and not be duplicative?
DR. WARREN: To me, what is not duplicative is I think we need to keep the
future views. So I mean, that’s talking about we need to take a look at what
are the implications of metadata five to ten years out? And, I would also like
to add the Semantic Web because that’s just now kind of hitting the not a few
people discussing it but a lot more of the industry. It was hinted at in PCAST
but never really identified, and I think with the ontology work and the data
manipulation that comes out with the Semantic Web, that may also influence
DR. CARR: So in contrast to ONC, we’re saying we would, the NCVHS, would
think about implications or hearings on long-term implications. Walter?
DR. SUAREZ: What I was going to suggest, I think this is good, but I think
it’s going to be much better to define a role once we see the recommendations
in the report from ONC to the Secretary because they might, it doesn’t matter
if our letter is redundant with some of the recommendations of ONC’s report. It
doesn’t matter if they are actually opposing views.
I think at the end it really, you know, what is going to be our role. And in
order to define our role, we need to understand what is the role of others.
DR. CARR: Marc?
DR. OVERHAGE: I just want to take Judith’s comments up one level because I
think the other thing that we have the opportunity to comment on, as sort of
stewards of thinking about standards and population data, is what’s the impact
of, I’m trying to choose my word carefully here, radical reengineering of a
trajectory given that we’re trying to move a whole community? There is some,
not as much as any of us would like, momentum. But, you were just highlighting
a bit ago how slow adoption of some of these things are.
It seems to me it might be appropriate for us to comment on how do we manage
that overall momentum. It’s hard to get the country moving, but if you get it
moving, it’s got some momentum, it may keep going. If you keep throwing pebbles
underneath that, it may slow down that momentum.
DR. CARR: Bill, did you want to add anything to that? Because I know when we
spoke yesterday —
DR. SCANLON: No, I’m much in line with what Marc is. I mean, you’ve got to
get this momentum going, and it’s an issue of pressing need. And, there’s going
to be sort of a whole lot of pebbles, we have to be able to run a rough shot
over them in some respects to get this momentum going and to accomplish what is
required. Because it’s really neither one of these things where we can continue
the way we’ve done in the past.
DR. FRANCIS: One of the things that we’ll be in a pretty good position to
know what they’re going to be thinking about because they’re having their
hearing, and they have set the 15th and 16th. And they, I
believe, have questions that they’re specifically looking at at that hearing.
So for example, it seems to me a question that might be very interesting to
us, and I don’t have a clue whether they’re going to be raising it is, how
should we be thinking about meta tagging if what we want is to realize the
dream of increasing the availability of the information that’s in exchanged
clinical information or all those communities we heard about on Tuesday? Now,
that’s an NCVHS question. I don’t have a clue whether ONC, I was guessing it’s
not on the list, but it sure should be on our list.
DR. CARR: Okay, great. So, is there anything else from Quality, Blackford?
DR. MIDDLETON: Yes. I mean, the only other one that did come up, and in
Paul’s absence I feel obliged to mention it, is to return to the idea of
identifiers in some way. I’m going to say that as carefully as I can. And maybe
in the current context, it’s actually more about identity management as opposed
to universal health identifiers because metadata will have needs to understand,
you know, what is the appropriate use for these data via person of type,
etcetera. And, if we don’t have a model for that or an understanding of what is
fair and appropriate there, which is not a valid universal health identifier,
we still may misuse or abuse clinical data.
DR. WARREN: You know, Blackford, as soon as you said identity management, I
think that’s the strategy that we need to take. When I held hearings a couple
of years ago on unique patient identifier, we renamed it matching patient data
to the patient. But, I think identity management is a much more encompassing
higher level concept of which identifier is only a part of that. And, I think
that this Committee could contribute a tremendous amount if we looked at
DR. CARR: Okay, so have I written it correctly? Identity management matching
people to their —
DR. WARREN: No, I would just put identity management. I wouldn’t put
matching, or anything like that.
DR. CARR: Okay.
DR. SUAREZ: It’s important to note also, this is the subject of the Health
IT Quality Committee and the Health IT Standards Committee already. This is in
their radar, identity management.
DR. CARR: Okay. Policy Committee.
MR. QUINN: The general, I mean, for all of these issues, what are we missing
today is that we have the Quality Committee and the Standards Committee, and
these are decisions that need to be made for right now, based on the technology
that exists today to help IT systems that exist today, programs that exist
today, and then we have these reports that are looking out at ordinary things.
Somewhere in between is half the transition from yesterday’s technology being
used today and technology of the future. Is this incremental or some sort of
transition then? I think that that’s really a sweet spot, how do we get from
here to there?
DR. SUAREZ: Yes, but the risk is the “there” might not be the
right there. So, if we try to create a path of going in the direction where the
goal is the wrong one or is maybe not the, you know, and is a 20 year goal down
the road, in today’s world, 20 years in terms of information technology is 20
So, the vision of metadata we have today might totally be changing in five
years with the next iteration of whatever comes out there. So, to me, metadata
is something that’s already happening. It’s already being done.
MR. QUINN: This isn’t a policy on things to do, I think. Somewhere in
between there, sometimes an incremental strategy is going to work. In other
places, it requires free platform or more radical approaches. And, I think that
that’s, when you pack the PCAST report on top of the near-term requirements of
high-tech, that’s where there’s this freaking out.
DR. MIDDLETON: You know, what I would offer, Justine, just on this very
point, and Mark alluded to this already, I mean in a way it’s just so palpable
for me at least, the ONC is barreling down a path with tactical decisions and a
near-term strategy to try to get implementation done in effective use,
etcetera. We really need to be three, five, or more years out.
DR. CARR: Exactly.
DR. MIDDLETON: And, it’s not only strategic, but there are principles which
I think this Committee does come up with which are above the technology or
apart from the technology at a level about security, about privacy, about
quality, about identity, whatever. You know, and the principles will be
consistent if we get them right because the technology will always be changing.
But, we need to be at the strategic level, acknowledge and support and benefit
from ONC at the tactical level, and really just get back ahead a little bit. I
think we’re a little bit behind.
DR. CARR: Exactly right. Okay, I’d like to hear from Populations. Don and
DR. MIDDLETON: Can I come back to quality organizations?
DR. CARR: You will. They have to leave at 12:00 though, or Sallie has to
leave. So yes we can, Blackford, but I want to sort of get the basic things
DR. STEINWACHS: Sallie, why don’t you start, and I’ll join in.
DR. CARR: And I know I’m really stuck on having grids and stuff like that,
but I thought about it in two contexts, a community data initiative and
liberating the data, but really also in just even, you know, the initiative to
help Americans achieve and maintain healthy weight, promote community-wide
policies and interventions at work. I think it ties in with both of those.
MS. MILAM: Don and I agree. Those are the two that jumped out to me,
Justine. Larry and Don and I, at least, I don’t know if you and Larry have
talked. I know I haven’t chatted with them about it. Do you have other
DR. STEINWACHS: No, you know, I think the first one is certainly the most
relevant of looking at the community as a change element, looking at the
community as a place where you make interventions that can reduce disparities
and increase access. And so, that fits under helping Americans achieve health
goals more generally as well as certainly obesity is high on that list.
DR. CARR: Yes, what is the next step then? What are next steps coming out of
your meetings yesterday?
DR. STEINWACHS: Well, you’ll hear more from Leslie about the next hearing.
So, the first hearing was really trying to understand what communities are
doing and some of the barriers in opportunities and needs of communities. And,
we heard that. And then, Security is taking the lead in the next dual hearing,
Privacy, Security, I never get your title right.
MS. MILAM: Well, it’s Privacy, Confidentiality, and Security.
DR. STEINWACHS: Okay well, I can pick any one of the three and then it will
work. I’m in good business. So they’re taking the lead, but it’s a joint
hearing again. And, it will be a one day hearing that begins to look more
closely at the privacy issues.
DR. CARR: And, with the information that we got yesterday, is there going to
be a –-
DR. STEINWACHS: Yes, Susan, we spent about an hour going over an outline
that Susan is developing, we’ll call a report. I think the idea is to have an
executive summary that’s probably about four pages, but the report itself may
be, I don’t know, 20 pages or something like that, but the idea of an executive
summary. And I think out of that, it will be possible then to make a decision,
you know, what should be communicated with the Secretary, but there would be a
product coming out of it.
DR. FRANCIS: And, the interface with Todd Park? We’re going to be presenting
at the June 9th.
DR. CARR: Okay, what are we presenting?
DR. FRANCIS: We’re going to be presenting both of the hot points from the
workshop on Tuesday as well as the, so that was the, what are the new ways
communities are using data? How are communities getting engaged in using data
to improve the overall health of the community? And, that’s not exactly the
same thing as health care in the community. That was one of the things we
DR. CARR: Yes, so it’s really community definitions.
DR. FRANCIS: So, the first step in steering that ship is, what are the
communities doing? And, you know part of this was this was a workshop, so it’s
what are a whole lot of the nifty ideas that are happening? And so, then the
next is going to be Privacy taking the lead looking at what are, I mean,
ultimately what we want is a kind of primer of best practices. So, how do you
think about generating community trust? How do you think about levels of
community engagement? How do you think about everything from having the
community involved in setting the questions to having the community involved in
trying to make sure that the kind of answers that get generated are answers
that are of interest to the community to more what people might think about as
privacy protection kinds of questions.
So, what we’re going to be doing in the next workshop is bringing in people
who’ve been thinking about these community trust issues. And so, product number
one is going to be the report that Susan’s sketching. Product number two is
going to be we’re scheduling this thing for the first week in May on the
Privacy side so that we have both the cool communities and the cool trust
bullet points that can be a presentation on June 9th for Todd.
Then, on the privacy, confidentiality side, the goal is going to be
something that looks like a primer of what are, you know, on the public health
side, I’ll just sort of sketch the territory. There have been three —
DR. CARR: Before we get into the details, I want to just take it back up a
level. So, do we know first of all what Todd Park’s expectations of us are? How
much time? What audience?
DR. FRANCIS: We’re asking him for a half an hour to 45 minutes.
DR. CARR: Okay. And, when we talked about this at the last meeting, we said
that it’s a short timeframe for us to come up with recommendations.
DR. FRANCIS: Right.
DR. CARR: So, we would put concepts out there and feedback at that meeting.
DR. FRANCIS: That’s the goal. And, what are some of the concepts that people
are using these mashups need to be thinking about?
DR. CARR: And again, for the presentation, is it going to be here’s Sonoma
County and how they’re using it and how it got better?
DR. FRANCIS: No. It could be partly that. It’s going to be look at some of
the community initiatives, look at some of the ways that communities have
brought the community in to learn from this data.
DR. CARR: So, it’s data driving improvements in the community. That’s the
DR. FRANCIS: That’s Populations, so I shouldn’t speak for Populations.
DR. CARR: Okay. Don and Sallie? Data driving improvements in the community?
DR. STEINWACHS: I think that is fundamentally the message that says that
data, bringing organizations together and community organization are driving
the opportunities and in reality, they’re driving actual changes and
improvements in the community and policy in the community.
DR. FRANCIS: Coalitions are using data, really.
DR. STEINWACHS: Yes. Community-driven, better said. Thank you.
DR. CARR: It’s community-driven, yes.
DR. STEINWACHS: And while some of us drive our data, but I can see how it
should be driving the community. No, no. So there was, Leslie, I thought there
was a comment about whether or not maybe that presentation at the
9th meeting might involve one or two community representatives. So,
I think that’s a little bit different because then it does talk about, and so I
don’t know whether that’s possible in terms of Todd funding that and so on.
DR. CARR: Yes, I think we need to have a conversation with Todd Park about
what is it that that audience of all the mashup folks and Google folks and
stuff like that want to hear. And, if it’s following on from what was last
year, I think they want to say, show us that liberating this data and making it
user friendly is a valuable thing to do, and give us an example, this community
had the data. And then, what are the hot points? What are the issues? Everybody
would like to do it, but they don’t have the infrastructure, and you know, how
do we think about that.
And then, on to as you do this, here are the privacy concerns. But, I think
that there are a lot, our paper is going to say a lot of things, but I think we
need to have a very crisp, targeted presentation for the June 9th
because if you look at last year’s video, that’s how it was. Time was short,
and you had a small amount of time to make an important point.
DR. FRANCIS: It’s just not written yet, Justine.
DR. CARR: No, but I mean you could write it ten different ways. I’m just
saying that the write-up of what we write may be one thing. What we present at
the meeting may be a truncated segment of that.
DR. STEINWACHS: So, is it clear who is going to have the contact?
DR. CARR: I’m asking you, yes.
DR. FRANCIS: Marjorie was going to talk to Todd.
MS. GREENBERG: Wait a minute. I am happy to communicate with him,
particularly to clarify how much time the Committee has. And I think though, I
need some collaboration here with the Co-Chairs as to in that communication I’d
like, I think we should say, this is what we’re thinking of doing and we’d be
prepared to do, and then let him react to it as opposed to just, what do you
want us to do? So, I think we need a little joint work on raising that.
DR. CARR: Okay, so Marjorie —
MS. GREENBERG: I mean, we talked about that yesterday about then some of the
presentations from the last time in that, you know, we’re not coming in with an
app or something flashy. It’s more of a policy.
DR. CARR: Sort of guiding principles kinds of things.
MS. GREENBERG: But I mean, depending upon how much time he’s giving the
Committee, we could bring in a community or two voices. If it’s very short, we
DR. CARR: So, all right. So, we have that set up. Vickie, did you have a
MS. GREENBERG: So, I’ll work with Sallie and Leslie on this, well the four
DR. MAYS: I just want to make a comment. Again, I’m on catch up here, so I
may be totally kind of off-base. But, one of the topics that you’re saying you
want to do is this trust issue. And, trust really gets around people’s
feelings, it gets around past behaviors, and I don’t know that that’s the most
productive going forward. I could guarantee you with the community that if what
you wanted to do was to really talk about the strategies or ensuring
protection, ensuring the right use of data that it would just be better. But, I
think you just might not get in the place you want to get to by focusing on
this issue of trust.
DR. CARR: Yes, it’s a theme that has really emerged repeatedly. It’s sort of
an attribute that’s underlying, so we’ve been talking about it for a while.
Okay, so I’ve heard from Populations, Privacy —
DR. FRANCIS: Oh, Privacy will be having the next workshop in this series
with the longer term goal, if we’re looking say 18 to 24 months out, would be
to have a best practices primer or public health. And, what I was starting to
say a minute ago is that, you know, there’s sort of three themes out there, all
of which I think we need to get beyond in a certain kind of way. One of them is
privacy is a barrier.
A second one is the way to solve the privacy problem, and frankly I think
PCAST in some interesting ways buys into this, is with de-identification. A
de-identification, we all know, has limits, both how effective it can be and
also, if you listen to the transcript of whatever the workshop was, a lot of
the stuff that’s of interest is not using de-identified data. So, you’ve got
qualitative stuff, you’ve got telling people stories, you’ve got, you know, so
another piece of it is getting beyond the de-identification solution.
And a third piece of it is, out there in the public health world, another
model, you know, we heard the New York model. They get everything which is not
a desirable model either if that’s, in fact, the New York model. But, what we
want to do is think about what are some of the, learning about what are some of
the different strategies that are being used, what are some of the better
strategies, what are some of the resources, all those kinds of questions.
And, Sallie may have more to, Sallie and I were just talking at breakfast,
and she may have more on that as well. But, that’s a whole program for 18 to 24
to 36 months. And, the way it fits into what you’ve got there is, if what we
need are kinds of information that help achieve the, let’s call it, let’s say
the healthy weight goal, this is all about how public health can get and
effectively use the information.
MS. MILAM: Let me just add on a statement or two. I think what we would like
to do is to demystify privacy so that data can flow more fully, more
efficiently to its full extent. And I think, as Leslie and I were talking, the
reason a lot of times it doesn’t is because people don’t know how to apply the
privacy laws, regulations, etcetera that are in place. So, we would like to, we
heard yesterday or at the workshop about how it’s a disconnect, how that lack
of knowledge and expertise is a barrier. And, we think there’s a huge
opportunity for us to really improve data use through privacy.
DR. CARR: Okay great. Perfect. So, we’ve heard from Privacy, we’ve heard
from Populations, and saving the best, before we go onto Standards, Blackford,
I wanted to give you a chance.
DR. MIDDLETON: Oh, so thank you. I guess the only thing that I didn’t
mention was I think the other thing that’s happening with the national policy
agenda and the adoption of HIT is much more increased focus on clinical
decision support and its impact on processes and outcomes of care. And while
our title, NCVHS, doesn’t necessarily imply attention to knowledge per se, I
think we might want to consider looking at knowledge management, knowledge
representation and its relation to decision-support because that’s of such
prime interest for the national policy agenda.
DR. CARR: Okay, and this would be, again, sort of not at the granular how to
implement it into an EHR, but more conceptual.
DR. MIDDLETON: Well, I think the standards issue is very relevant, very
pertinent. I represent knowledge for sharing and effective application and
decision-support. I’m imagining hearings and then some derivation of principles
and recommendations at that level.
DR. WARREN: Right. Because once we get into terminology standards, we’re
into knowledge representation. And, there’s a lot more work that has been done
since the initial NCVHS recommendations on terminologies that we need to
consider. So, this would be a joint —
DR. CARR: Yes. And so, this is evolving though. This didn’t get discussed
yesterday at the —
DR. MIDDLETON: It was discussed when I joined the Committee.
DR. CARR: Yes, okay.
DR. MIDDLETON: I’m just bringing it back.
DR. CARR: So, we’ll get that out there, but we might want to flesh it out a
little bit. So now, on to Standards, and then after we do that, I want to come
back to Learning Health System. Chuck, are you on the line? Chuck Friedman?
DR. FRIEDMAN: Yes, I am.
DR. CARR: Okay great. So, Standards?
DR. SUAREZ: Well, there’s two groups of activities we’re doing. One is the
responsibilities under the Affordable Care Act and HIPAA so that continues,
that’s what drives our priorities. And so, there’s a box there that talks about
health reform and Affordable Care Act. That is one of our priorities, one of
our drivers for priorities, and so, I don’t know if we need to get in the
details of that.
But, the other big group of activities related to standards that I’ve not
necessarily followed the ACA and that are elements that apply to the other
areas are really, you know, the concept of who is looking at establishing and
defining, identifying about a way to define and recommending the standard for
the next generation of elements like metadata. You know, who is really out
there looking at standards for metadata?
DR. WARREN: Metadata, the ontologies —
DR. SUAREZ: The ontologies, the standards for public health transactions and
data collection and exchanges. And so, I think there are a number of efforts
out there working on these areas, but I don’t know that they have been engaged
or included in the agendas of groups like HIT Standards Committee, HIT Policy
Committee, so I think we have an opportunity to look into those. Primarily, I
think the areas in, clearly HR standards, a lot of the terminology, a lot of
the clinical messaging are the purview of the HIT standards, but we need a
So, I think it’s more standards around public health data exchanges and
standards around metadata, standards around other topics like personal health
records. You know, what are the standards that are being used to exchange data
with personal health records? So, I think those are two big, in my mind,
spaces. One is the current responsibility we have under HIPAA and under the new
Affordable Care Act and sort of the new areas of standards that are not
necessarily included in the purview of other advisory groups.
DR. FRIEDMAN: Justine, can I jump in with a comment?
DR. CARR: Yes, please.
DR. FRIEDMAN: So, I think the standards observation for public and
population health is very, very important. I’ve been struck and have been
presenting my views about this for about a year now, and the overall
observation has been that meaningful use is necessary but not sufficient for a
learning health system and just about every aspect you can think of in
standards is one of those aspects.
So, we will get many standards that are useful for population health and the
learning health system out of meaningful use, but there are other kinds of
standards, and I think the classes of them are identifiable even now that we
will not get out of any conception of meaningful use that I have heard. So for
example, if a learning health system is going to require some mechanism for
members of that system to respond to queries that other members of the system
make to be put against their data, we need standards for distributed query.
And, standards for distributed query, to the best of my knowledge, don’t exist
with this kind of context.
There may be examples from other fields that could be drawn from, but I
don’t anything specific to the purpose exists. Maybe some of the work of CDISC
and other organizations is relevant here. But nonetheless, I think there is
work to be done to develop distributed query standards and other standards that
are seemingly absolutely necessary to have a learning health system.
DR. WARREN: Justine, I’d just like to add, I’m glad Chuck brought up CDISC
and some of the others. But, we also need to look at standards from the
research community as we start having this digital infrastructure of being able
to share data from clinical to the research, both ways.
And so, in looking at CDISC and caBIG and then we have some of these
initiatives started already, the CTSA initiative is also beginning to look at
it, but they’re not pulling in the full view of a learning organization where
all of these things, all these entities talk to each other. So, I think that’s
DR. SUAREZ: So, it’s just going to link, you need to add specifically
standards for public health and population health. That’s, you know, the
distributed queries is sort of one element to it, but there’s a large body of
work being done and to be done around standards for public health and
population health, standards in the sense of data standards and data exchange
So, we have population and public health, research, personal health records.
I think those are three areas, public health and population health, clinical
research, and personal health records are three major areas that we’re
currently really not touching from a standards perspective, and we meaning the
DR. CARR: So Chuck, as you think about other next steps coming out of the
learning health system report, standards is one that aligns with the Committee.
What else would you put forward there?
DR. FRIEDMAN: Another would clearly be in the policy domain and policy and
governance. A learning health system is going to need a governance to create
and enforce its system. The learning health system has to function as an
aggregation of parts to accomplish tasks that are not accomplishable by the
individual constituents acting alone, so there will need to be a governance for
that and a set of policies that the governance both creates and enforces. So, I
think advice about what those policies should be, this is obviously enormously
complicated, it brings in critical privacy considerations, would be enormously
helpful to us.
DR. CARR: And what’s the, ONC has a governance committee now to address the
DR. FRIEDMAN: Yes.
DR. CARR: And, what is their focus?
DR. FRIEDMAN: They are focusing on the governance of what used to be called
the NHIN, but is now called the NWHIN or NEW-HIN because of trademark
infringement issues. And, they’re actually working right now on scoping what is
going to be their product. They’re under a statutory requirement to set up a
governance for the NWHIN as stipulated under HITECH. They may very well scope
down what they produce to not address many of the critical governance issues
for a learning health system. They may just feel they have enough to do with a
system of information exchange that primarily supports mobility for health
I don’t see how they can avoid putting in place some principles and elements
that will support a learning health system. But, as I’m tracking what they’re
doing, I don’t think they’re going to take it all the way. And, I’m reasonably
certain at this time that some additional work is going to be needed, and if
we’re going to develop the learning system, the timeline that I’m hoping for
and aiming for, we probably can’t wait for that work to be completed. So, in a
serial sense, they can take it up at some later time.
DR. CARR: Great. Judy, you had a comment?
DR. WARREN: Yes. I was just going back over some of the themes that came out
of this report, and the one thing we haven’t been addressing, and Chuck, I’ll
defer to you on advice for this, is this ultra-large-scale engineering
approach. This is something that the Committee has had a document to read on,
but we’ve never really explored it or discussed it to see what the implications
are in this domain. So, I’m not sure where that goes, Chuck, but —
DR. FRIEDMAN: Oh, okay. Good question. I’m sure you recall visibly every
word I said in my little presentation yesterday. But, one of the things I
mentioned was these two workshops that we held in January, one of which was
called, next generation interoperability for health. And, a report that is
going to come out of that, the ULS perspective, the ultra-large-scale
perspective, is going to be a key element of that report.
So, maybe a very efficient way for the Committee to consider the ULS
perspective would be in the context of this report which will have ULS as a big
piece of it. And, as soon as the report is ready, we can forward it, this is
just a proposal, but we could forward it to the Committee for its
consideration. And, there could be a discussion of the next generation
interoperability report as a whole which will fold into PCAST report, the
semantic web, and the ULS perspective among others.
DR. WARREN: Okay, that’s where I thought it fit was with the meta-tags and
the semantic web work but more as a method to incorporate those so they kind of
fit in both, up there in the PCAST report as well as down in the digital health
DR. CARR: I know this is a kludgy construct, but it was just to get us to
make sure that we are tying out to what we said we are going to do. But, it
needs work obviously. So Chuck, additional ideas coming out of the IOM report?
DR. FRIEDMAN: No, I think the two domains, and they’re large domains that we
just discussed, standards and the domains of policy and governance strike me as
very fertile areas for this Committee to explore.
DR. CARR: Great. And Mark, did he just leave? Okay. Any other ideas? So
then, taking this back to the reality, what we have are immediate issues are
we’re getting the PCAST report response together, we’re getting the community
data initiative paper put together, we’re planning for the presentation, for
the next hearing focusing on privacy and then putting that together for the
report, the presentation in June.
And then, Standards has got lots to do. You’ve got your letter, your DSMOs,
you’ve got all of the stuff you outlined below. And, Quality doesn’t have any
immediate hearings left, but I think that what we’re finding is the ebb and
flow of the themes allows us to have cross-participation, and I think that this
is working very well.
DR. FRIEDMAN: Justine, are you saying that the Committee would be open to
more specific suggestions from us in this learning health system domain for
things it could do? Or, are you saying that the dance card is full and to check
back in a year or so?
DR. CARR: The dance card is full. No, I think that, you know, as we have
new, we’ll be having new members joining us in June, and we’ll be up for a full
complement of 18. And, I think by our June meeting, we will actually have a
significant amount of work done, offline outside of the meetings. And, I think
that’s a good chance to, we’ll also have the opportunity of the updates on the
PCAST report and the ULS report, so I think we can revisit other pieces at that
DR. WARREN: Matt just reminded me that one of the standards up there that
we’ve not addressed in the past and we probably need to consider whether it
goes on the dance card or not are the human-computer interaction standards that
are now beginning to come out.
DR. CARR: Okay.
DR. WARREN: HCI.
DR. SUAREZ: The other thing that I can, Chuck, that you might want to
consider, something like this, I think it would be very helpful to hear about
an identified set of priority areas coming out of the learning health system
because we are creating our own set. But, there might be some that, Chuck, that
you might have sort of identified as it would be great to, as next steps on
this learning health system process, work on these three or four topics. And
so, if you have those and you can provide those to us, that would be incredibly
helpful to incorporate into the agenda.
DR. FRIEDMAN: Right, right. So Justine, we can take this discussion forward
any way you want, but in a general sense, should we be thinking about bringing
you something more specific for consideration at the June meeting?
DR. CARR: Yes, I think that would be great. I mean, just at the moment, do
you already have a couple of topics off the top of your head? Directions that
you’re thinking of?
DR. FRIEDMAN: If pressed, I could expand those two general areas. For
example, under standards, the distributed query topic as a more specific one
quickly bubbles up. But, I think this is a nice timeline given some other
things that we’re doing to move the plan for the learning system along. Let us
ruminate on that, and with your, if invited, we can bring those more specific
ideas forward at the June meeting. I think that’s very good timing.
DR. CARR: That’s great. Leslie?
DR. FRANCIS: I also wanted to say that you know what our current, I think
you were listening in and hearing what some of the plans are, for example, the
populations, privacy jointship about, I’ll call it best practices in public
health and all the kinds of new initiatives. I don’t know if you heard that
part of the discussion, but if there’s anything that’s going on on your end of
things that we might, that’s relevant to that, to what’s going to be happening
between now and June that we might miss, the more input we have, the happier.
DR. FRIEDMAN: Okay. That’s terrific. I appreciate having the chance to have
this discussion with you. I think these are a great set of ideas. I have to
jump off onto another call. I’m sorry I wasn’t able to join you at first today.
DR. CARR: Okay, no, thanks Chuck. That was great. Thanks. Bye.
DR. FRIEDMAN: Bye.
DR. CARR: Okay, well, we’ve been very efficient today. Is there anything,
are we missing anything? Was there anything else you wanted to bring forward?
Mark, I was calling on you in your absence just in terms of the FDA and the
other things that you brought forward. But, I think that they do actually fall
under this concept of governance of the learning health system as well as
standards. Those are the two things that Chuck was talking about.
DR. HORNBROOK: What I was raising was the issue of, if you’re a data holder
in the sentinel, there’s no IRB. IRB is totally excluded because this is public
health surveillance. And of course, it’s perfectly appropriate and legal, but
it seems like it would be appropriate for the Committee to reinforce that this
is a true public health important event, important activity, that there isn’t
something fishy going on, there isn’t something untoward going on, that it
really is important for people to understand their data are being used for
their own protection even though there are plenty of people who’d say, I don’t
want my data to be used for people looking for adverse events.
So, this is a true example of where privacy is trumped by public health
because we don’t ask people, is it okay to look into your charts? The FDA has
mandate to go in and look at your chart for detecting adverse events. So, the
Sentinel Network is based on first, electronic surveillance of claims and EMR
data and second, a drill-down into manual abstraction of charts for
verification of safety events. And, there’s a whole methodology for taking a
signal that’s moving from a random association to a cause and effect
association to a literal hazard association that has something you can do about
DR. CARR: Okay. So, thank you. Marjorie?
MS. GREENBERG: Yes, I was taking my notes here too. And, I think we have
about the same to-do list, at least short-term. But, in regards to the June
meeting, as Justine said, we do hope to have potentially six or more new
members. Our plan would be to invite all retiring members as well as to have
the new members here. We hate to not have an opportunity to do that transition.
I think there’s value to it, and we always look for a chance to celebrate. So,
just to keep that in mind, keep that on your calendar whether you’re a new or
retiring member or not sure. But the, none of us are never quite sure.
The other thing is, it’s not too early at all for I want to thank Jeannine
for this lovely venue for this meeting. We do have a few more options in
February than we do in June, so we need to start looking for space already for
the June meeting. And, I haven’t heard of any subcommittee that was looking to
tag onto the June meeting, but so speak now if, you know, we’re talking about
the Standards and Privacy, Populations, a week that’s either maybe the last
week in April or the first week in May it sounds like.
And, you weren’t here I think when they were saying, well, they would want
to meet at the same, I mean, sequentially, as overlap and then obviously it’s a
standalone meeting. But, then it does save travel costs. So, I assume neither
of you would be meeting in conjunction with the June meeting. We also will have
just participated to some degree in the June 9th. What about
Quality? Are you not intending to have any kind of hearing or workshop in June?
DR. MIDDLETON: Between now and June or in June now?
MS. GREENBERG: No, I mean, so we’re really talking about two days. Actually,
on the other hand, I think what we would probably try to do is have a half day
orientation for the new members which we would invite the Executive
Subcommittee, at least one co-chair of each, you know, at your will, to be
here. We’ll have to orient them through the mails and with a teleconference or
webinar or something because we can’t wait until then, assuming that we have a
timely naming here of members.
But, I think that would be a good use of, we probably do need three days of,
or at least an afternoon, and then two days. So, I just wanted to clarify that
for our planning going forward. Anything else from Debbie?
DR. MIDDLETON: Are those dates already known?
MS. GREENBERG: Yes, it’s June 15th and 16th. So, I
guess it would be the 14th, probably the afternoon of the
14th that we would be doing that orientation.
DR. HORNBROOK: Do you want all of us to try to come for the three days or
just the specific people for the first day?
MS. GREENBERG: I think at most the Executive Subcommittee probably.
DR. CARR: Yes, and I would even, I mean one co-chair.
MS. GREENBERG: Not even the whole, I mean obviously Justine and maybe one
co-chair from each subcommittee. If one’s there, each subcommittee needs to be
represented because I know there will be a scramble for membership to put your
best face forward. You’ll probably be contacting them as soon as you get their
But anyway by then, of course, we will certainly have Raj and Vickie as full
members also. So, and it is possible that it could, well we just have to see
how this moves along as to whether any of the new members, certainly the two of
you should think about what subcommittees you want to serve on and then whether
you, it’s three of them are actually going to be meeting and going to be having
a meeting prior to the June meeting. You know, if you’re going to be on one of
those, then we’ll pull you as well. So yes, I think only a reduced Executive
Subcommittee unless we come up with something else.
I mean, there are so many long-term goals and ideas and visions, etcetera
that are being, that have just been discussed as well as all the short-term
things. But, you know, plans that are maybe 18 months to 36 months as we said
that after we get the new members on and have sort of settled in, we might
think in the next year of having a Full Committee retreat.
DR. CARR: It seems like only yesterday, we just had one.
MS. GREENBERG: Yes, well, just to explore some of these things in a more
robust way maybe around the IOM report also. It’s just something to think
DR. CARR: So, are there outside speakers that we want to invite into the
MS. GREENBERG: For June?
DR. CARR: Yes. Topics that we’ve been talking about here.
DR. SUAREZ: Just quickly, you mentioned June 9th. It’s really
MS. GREENBERG: Well, June 9th, no. June 15th and
16th with the June 14th being the orientation of the new
members, but then June 9th is this IOM meeting that we’ve committed
at least a few members participating.
DR. SUAREZ: Oh, I’m sorry. Okay. That’s Todd Park’s meeting.
MS. GREENBERG: Todd Park’s meeting, yes.
DR. SUAREZ: Okay. Sorry.
MS. GREENBERG: I’m sorry if I was confusing you on that.
DR. MIDDLETON: We should get the new national coordinator to come visit.
MS. GREENBERG: Do we know who that is?
PARTICIPANT: Does anyone know?
PARTICIPANT: Somebody knows.
MS. GREENBERG: Well, it’s something to think about. I should mention
actually that we will have two new liaisons, I mean two new people serving as
liaisons though not agencies that we have worked with as I understand it.
Brenda has informed me that, do you want to tell us who —
DR. EDWARDS: It’s my understanding that NIH has identified the NIH liaison
is Dr. Kaplan. He, I believe, is the new head of OBSSR, the Office of
Behavioral and Social Sciences Research. I always get them mixed up, but thank
you, and you see I don’t quite have all those acronyms. And I haven’t met him,
but I will be trying to set up a time to see if he would like to have me
handoff the materials from the last couple meetings I’ve attended on behalf of
DR. MAYS: I was just going to say, he is actually a psychologist by
training. His area that he’s working in is comparative effectiveness research.
He has a big webinar coming up I think in the next couple of weeks where he’s
going to be talking about CER. He just left us. He was the Chair of the
Department that I was in at UCLA in Health Services.
MS. GREENBERG: Okay. So, I mean I really think that some time obviously will
be well spent at the June meeting getting to know the new members, new
liaisons. The other person is, as you all recall, Steve Steindel, many of you
do. Steve Steindel was our liaison from CDC. And sense he retired, we haven’t
had a replacement.
But, Dr. Seth Foldy, who is now the Director of the Public Health
Informatics and Technology, sort of the office that is the successor to the
National Center for Public Health Informatics at CDC. He will be coming as the
CDC Atlanta liaison. He was sorry he couldn’t join us at this meeting.
DR. FRANCIS: He would make a lot of sense to have be involved in the joint
POPS privacy stuff at new uses of data. We should get a chance to talk to him.
MS. GREENBERG: Well, I did tell him briefly about the meeting we were having
on the 8th. He had to be out on the West Coast, but certainly feel
free to engage him.
DR. HORNBROOK: Bob Kaplan is the Bob Kaplan of Quality Well-Being scale.
He’s the health status measurement expert. It would be wonderful to have him
here to be thinking about patient reported outcomes and linking into EMR data.
It’s just amazing support for our work.
MS. GREENBERG: Is this the one and same?
DR. EDWARDS: Yes, I’m sure it is. He’s new to NIH, so Vickie knows him. I’m
just now getting to know him. I know he’s coming.
DR. HORNBROOK: He’s well-known to many of us.
MS. GREENBERG: So, I think both of these liaisons, it sounds like, can
really bring some good strength to our, yes, what did you want to say?
MS. JACKSON: A quick follow-up. Vickie and others may remember, we had those
short bio sketches maybe three years ago, three or four years ago, was an
attempt for the members to know each other. And, when we got this large cadre
of people coming in, it might be nice, we can unearth the old ones, we could
update, and then we’ll be requesting for them to submit theirs as well.
So, at the June meeting, there will be a lot of give and take in
communication to keep that synergy going because, as Marjorie was saying, so
much has been happening in a full unified sense for the Committee in its new
kind of pattern versus everyone kind of going off on their own. So, that had
been a way of working through the Committee. Everyone would just kind of select
a subcommittee and stay there in a silo, and now, you’re not only dealing with
all of that amazing amount of content and expertise, but really sharing and
doing a coalesce.
So, from June and on, there will be a whole lot of working and sharing going
on. So, I’ll communicate and pull out your old bio sketches, and then we’ll be
requesting for the new ones as well and use all of that to help that synergy to
hit the ground running in June.
DR. SUAREZ: So, do we know the status of the appointments of the new
MS. GREENBERG: I think it is under review by the Office of the Secretary,
recommendations under review.
DR. SUAREZ: Okay.
MS. GREENBERG: I mean, these appointments are made by the Secretary to
develop, you know, I think we got nearly 50 nominations. And so, we’ve worked
DR. SUAREZ: So, usually with new members, you know, we the co-Chairs of
subcommittees get a chance to do a little pitch to —
MS. GREENBERG: Well, they haven’t been released yet, so we can’t give you
their names. But, that’s what you would do possibly on the teleconference that
we would have with them, but certainly on the afternoon of June
PARTICIPANT: They’re all quality, Walter.
MS. GREENBERG: Speaking of learning from within, Jorge Ferrer, who was our
liaison from the VA, wasn’t able to be here this week. But, he is going to be
making a presentation or doing a demo at the HIMSS. It may, first of all, if
any of you are going to be at HIMSS, I’m sure some of you will, I would
encourage you to go there. And, he’s using this, in part, to follow up on the
discussion at the last meeting about trying to incorporate ICF into some data
exchange around functioning in the VA and other things as well. So, it may be
that he might report to us in June. My sense is there’s going to be more than
enough going on, but if there are high priority external presentations that you
want, let us know.
DR. FRANCIS: It sounds to me like we have a large population of new people
who meet our standards for high quality. (Laughter)
MS. GREENBERG: Very well done. And, we’re keeping them very private.
DR. CARR: Okay, so gosh, what do we do now?
MS. GREENBERG: We have lunch. Do come back at 1:00.
DR. CARR: So, can you just give a little headline about the presentation at
MS. GREENBERG: Yes. I’m a little concerned because the person who really
asked for, I mean I think you’ll all be interested, but who really asked for
this presentation is Paul. And, I realize now he’s not here. So, I don’t know
if anyone else talked with him, but our understanding is, and it’s in the
minutes, the draft minutes, that we sent out to you, but that he, that this
obviously as people are beginning to, and it relates to this whole semantic web
thing I think, and apologies and all of that, but terminologies obviously.
As people are really beginning to implement more granular terminologies into
electronic health records and at the same time were transitioning from ICD-9
code sets to ICD-10 code sets. How are these going to work together? And, what
are the, you know, we talked about the mappings between ICD-9 and ICD-10 code
sets, but what type of mappings are being planned between the terminologies and
the classifications? And, not just ICD, but ICF, ICPC, you know, several,
although ICD is obviously the top priority.
Now, that’s going to be done and is being done through a joint collaborative
activity of the World Health Organization or other proprietors of
classifications but certainly for ICD, the World Health Organization and the
International Health Terminology Standard Development organization. So, I
believe Vivian Auld, but it could be Betsy Humphreys because Vivian, there was
a possibility she was going to be called —
DR. WARREN: It’s going to be Vivian that gives the presentation.
MS. GREENBERG: It is? Okay. She thought she might be called for jury duty.
DR. WARREN: Oh, that’s right.
MS. GREENBERG: Yes. So, she will be here, one of the two of them or both. It
might be that Betsy will be on the phone too. I’ve also asked Donna Pickett to
be on the phone because she’s got the lead on ICD-10 CM. And, they will –-
DR. WARREN: Be prepared that I might have a possible conflict of interest
with the presentation because I sit on IHTSDOs Quality Assurance Committee and
am very active in the nursing workgroup which has done quite a few products as
MS. GREENBERG: Okay.
DR. HORNBROOK: I think that’s sufficient.
MS. GREENBERG: What?
DR. HORNBROOK: Having Judy disclose it is sufficient. She doesn’t need to
leave the room.
MS. GREENBERG: No. She certainly doesn’t need to leave the room, but with
that conflict, if there’s discussion about recommendations or whatever, then
she would have to recuse herself. Is there anybody else who would have a
conflict in that regard? No.
So, that’s what we’re going to hear about, and I think one of the issues is
whether there are sufficient resources to do this type of work particularly in
the US because what WHO and IHTSDO will work on is the SNOMED-CT to ICD-10 map.
We will have to take, go from there to do the 10-CM. They clearly won’t be
doing anything with 10-PCS. It’s not an international classification. So,
there’s definitely work that has to be done in the US. And so, we’re going to
hear about how that’s coming and what might need to be done in order to make
sure it happens.
DR. CARR: We’re going to start back at 1:00, so we need time for lunch.
DR. WARREN: We need to end pretty quick.
DR. CARR: Yes, so it’s time for lunch.
MS. GREENBERG: Yes, but since we thought we could get lunch in 15 minutes,
I’m sure we can get it in 45.
(Whereupon, a luncheon recess was taken.)
A F T E R N O O N S E S S I O N
DR. CARR: Welcome to our afternoon session of NCVHS, and so welcome to Betsy
Humphreys. Thank you so much for coming today to speak with us about bridging
SNOMED and international classifications. So, let’s turn the floor over to you.
MS. HUMPHREYS: Okay. It’s a pleasure to be here, and I am speaking on behalf
of myself and my colleague Vivian Auld who would have been here perhaps
speaking to you if she hadn’t been called up for jury duty this morning. I am
going to, I think, go through this pretty rapidly and leave time for questions.
And, if people want to interrupt me as I go along, it’s fine with me.
In case there’s anyone here who wonders what a nice place like the National
Library of Medicine got into a business like this or how we did. In fact, the
National Committee on Vital and Health Statistics over the last decade or more
had a fair amount to do with that. But, we are currently distributing both the
HIPAA classifications and code sets and the terminologies required for EHR
certification within our unified medical alignment toward a system
We are the designated HHS coordinating body for clinical terminologies. And,
the NCVHS is responsible for that, having recommended that to Secretary
Thompson, and he took the recommendation some years ago. We are the, a
principal funder I should say of LOINC, and we are not the only one because the
Regenstrief Institute obviously is a major supporter of it as well. We are the
developer of RxNorm clinical vocabulary, and we paved the US membership in the
International Health Terminology Standards Development Organization which is
the reason why we have US high use of SNOMED. So, we do all of those things.
And, that’s just the background for why the National Library of Medicine is
involved in this.
And, I’m going to go first directly into this notion of, or the topic at
hand which is bridging between SNOMED CT and the classifications. And as this
group knows all too well, when we say, we’re bridging between SNOMED CT and
ICD, we have to qualify that so we all know which version we’re talking about.
So there are, in fact, efforts that are bridging from SNOMED CT to three
different versions of ICD, so I’m going to go through them in order.
There is distributed with SNOMED CT a crosswalk or mapping between SNOMED CT
and ICD-9 CM. This was the one that was initially developed by the College of
American Pathologists and has been maintained by the IHTSDO as part of the
intellectual property that they acquired from the College in 2007. And of
course, as I think many of you know, in addition to this mapping which is more
of a conceptual mapping although it has some billing features attached to it or
some features that would point out that there are multiple codes that a SNOMED
CT, for example, concept might map to depending on what else was extant in the
record and had to be the deciding factors between two potential disease
We all know that many people have mapped SNOMED CT to ICD-9 CM within their
individual systems, either parts of it that matter to them or big chunks of it,
certainly for example, organizations like Kaiser Permanente have mappings that
they’ve created themselves. And, there are others who have done that as well.
MR. LAND: Betsy, is this complete? Is the mapping complete?
MS. HUMPHREYS: This particular mapping, which is a conceptual mapping, is
relatively complete. It is not, however, fully done out at the level of if your
purpose is to generate bill. And, we have posted graph mappings which are from
a subset our 6000 SNOMED CT concepts to ICD-9 CM where the mapping was, in a
way, more complete because it basically supplied in some cases the information
about the rule you would have to apply to decide whether A, B, or C was, in
fact, the correct classification number for that concept.
We have found very little use of this thing. We have posted it and asked
people to use it and provide feedback. I can tell you, we have received very
little, and it’s been there for a couple of years. And in general, everyone has
said, this mapping is not useful to us because it is not a one-for-one mapping.
And, that of course is when you turn around and say, well, I’m sorry, you know,
you actually do have to combine a problem list entry of X with information
about something else, a comorbidity, an age group or something in order to get
the classification. So, in a system like SNOMED CT, you may very well need to
have a combination of factors, and where are those going to be in the record,
not necessarily in the problem list, I mean, if you have a full patient record
and you have an age or whatever other information on the patient.
So, it’s clear from all of this that in terms of implementations here in the
United States, and this would apply to all of these, everybody is asking for a
one-for-one map. And, it is an interesting concept because I’m not so sure that
we actually want to put those pre-coordinated concepts into the problem list
for the other purposes for which we would want to use electronic health
DR. HORNBROOK: Which, to me, suggests that the requests are coming from
coding clerks, as opposed to people who have more decision-making and
MS. HUMPHREYS: I would not actually accuse the coding clerks of that. I
think it’s more the issue of the vendor has talked to whomever who’s buying
their system. And, whether it’s a practice-based thing or whatever, this
practitioner is saying, fine, you know, I’m going to pick this and then there
should be, you generate this. Or alternatively, I’m going to use ICD-9 CM in my
problem list which I think can also be problematic because of the combination
of factors which means that subsequent things you want to do with the record
are less easy to do because of that.
DR. SUAREZ: Just a comment. I think more and more people are really
realizing that there is no simple and there will never be really a simple
one-to-one crosswalk between the various concepts whether it is ICD-9 to
ICD-10, ICD-9 to SNOMED, ICD-10 to SNOMED. I think what people are looking for
is the intelligence that is going to be needed in order to discriminate between
the various maps, the various one-to-many matches, and create the right one
map, one-to-one match for that particular circumstance. So, I think that is
kind of the next step is really what’s the sort of intelligence and the
additional information about the particular situation, condition.
MS. HUMPHREYS: And obviously, you have the situation where people are
incrementally moving into an electronic health record. And, the problem that
they see with this is that maybe they do not yet have in their record all the
other information they would be required or it’s buried in text or whatever.
So, then we have an effort which is ongoing now to map SNOMED CT to ICD-10.
This is the one as issued by the World Health Organization used in this
country, I gather, for cause of death but not for anything else but used more
broadly in other countries. And, this is a project, a joint project of the
International Health Terminology Standards Development Organization which owns
SNOMED and the WHO. And, I have grayed out the WHO only because they are, it’s
a question of relative effort being put in, and most of the effort is on the
IHTSDO side at the moment.
Jim Campbell, a former member of this group, am I right? Anyway, he is the
Chair of the project group. Some of my colleagues at NLM, and there are a
number of active US participants in this, they are also providing requirements
based on this project for a mapping component of the IHTSDO’s terminology
tooling workbench, and they are trying to get this done as soon as possible.
And, if any of you have people who would like to participate in this,
they’re actually asking people to submit proposals that they might participate
in getting this map finished more quickly. You have to move very fast. I think
they just posted this, but they basically want anyone who is interested to tell
them, submit a proposal by February 22nd. You probably can’t see,
but they’re desirous to have this thing completed this summer which is why
they’re trying to move as quickly as possible. This is of interest because some
of the methodology and the tooling and the ideas that were developed over a
fairly long period of time by a bunch of people really staring at this problem
we believe are also applicable to the 10 CM mapping.
So, just to tell you where you are with mapping, this is a picture of the
data we have on ICD-10 CM in the UMLS Metathesaurus. So, you can see that at
the synonymy level of the 97,000 plus items, concepts in ICD-10 CM, we found in
inserting into the UMLS a 12 percent overlap with SNOMED CT. So the question
is, now will that 12 percent include many of the most heavily seen problems and
conditions in the United States? So, absolutely.
So, I know colleagues who have done preliminary work to look at, for
example, the SNOMED core problem list that NLM generated which was based on
frequently seen problems in a variety of places across the country. And, there
was more like a 70, you know an 80 percent, close to 80 percent overlap with
DR. FITZMAURICE: Betsy, does that include CPT?
MS. HUMPHREYS: No, no, no. No, you see you go down to this, and if we were
talking about PCS, the Procedure Coding System, it has a less than one percent
detected synonymy with any other source, and of course, you know why if you’ve
looked at it. So you know, its structure is very different from anything else.
DR. HORNBROOK: Is there an ICD-9, ICD-10 CM for the US or only the ICD-10
from the WHO?
MS. HUMPHREYS: The data that you are looking at here is for 10 CM.
DR. HORNBROOK: It is for 10 CM.
MS. HUMPHREYS: Yes, clinical modification.
DR. HORNBROOK: The US, clinical modification.
MS. HUMPHREYS: That’s right.
DR. SUAREZ: Well, the third box that says 11,697, that’s 11,697 concepts in
MS. HUMPHREYS: CM that we detected there were synonyms for in SNOMED CT.
Now, are there more synonyms? Undoubtedly, there are. Are there huge numbers of
additional synonyms? Probably not. Now, I mean again, it’s again the question
of granularity and structure. And, let me say that there are certainly things
in ICD-10 CM that are more specific than anything in SNOMED because that same
concept in SNOMED would be formed by a combination of terms.
Then, there are also things that are more specific in SNOMED than anything
in ICD-10 CM. So, it would be really a nice thing if we could say that one of
them was more specific than the other in every case. But, that’s just not the
way it is. Along some dimensions or in some areas, one of them is more
specific, but on the other hand, the other one is because in the case of 10 CM,
there are combinations of things that are not combined.
DR. HORNBROOK: So, you’ve done a metathesaurus analysis. You aren’t looking
just at literal character strings that are the same. You’re looking at
underlying meaning that you deem actually close enough to be the same.
MS. HUMPHREYS: Yes, that’s what we do when we, that’s what the National
Library of Medicine does when it inserts any terminology into the UMLS
Metathesaurus. Our goal is to detect synonymy between the new thing that’s
going in and anything else that’s already in. And, the UMLS has more than 100
different vocabulary synonyms.
So, certainly a first step or a reasonable first step in developing any
mapping between SNOMED CT and 10 CM was to put 10 CM into the Metathesaurus
where SNOMED CT already existed. And, this was done relatively recently. And,
it had to do with changes that were being made to 10 CM, and at the point when
NCHS thought it was reasonable for us to put it in, we put it in.
DR. HORNBROOK: To recapitulate then, you’ve taken ICD-10 CM into the
Metathesaurus, SNOMED into the Metathesaurus and then done the linking in the
MS. HUMPHREYS: That’s right.
DR. HORNBROOK: And went backwards then to link them.
MS. HUMPHREYS: And of course, we do this, there is, you’ll see it in a
minute with some examples, I have a little example that might give you this
idea, but the thing is that there is very, very rich synonymy in the
Metathesaurus. So, it’s highly possible that a concept is in SNOMED and a
concept is in ICD-10 CM. And, the terms between the two of them are not the
But, already in the Metathesaurus is a link because this one is a synonym of
this, and in this source, that was a synonym of that. So, we have such rich
synonymy that the Metathesaurus itself, the existing connections, can very
often help make connections between other things.
DR. HORNBROOK: This includes both single words, but also pairs or triples of
words that are strung together in a grouping.
MS. HUMPHREYS: Yes, this is not at the word. It’s the concept name, so it
depends on, you know, we don’t break apart the concepts in each source. We’re
saying, if this is a multi-word name of a concept, we’re saying, does the
concept that represents like that, does it exist as a single concept in another
vocabulary even if it might have some totally different name in that other
Now, is this perfect? No, of course not. And, it gets better over time and
10 CM was recently added. So, some of the sources that have been in the
Metathesaurus for, you know, going on 20 years. If they’d both been there that
long, we’d probably have, you know, reduced the undetected synonymy really down
low. I don’t think we’re there yet at that level with 10 CM.
DR. HORNBROOK: What about the concepts that are nested so you can have an
anatomic location be upper and then a more narrower or the narrower fits in the
broad, and is that within your Metathesaurus?
MS. HUMPHREYS: Yes.
DR. HORNBROOK: Ability to make that kind of —
MS. HUMPHREYS: There are many other relationships. Synonymy is our primary
first goal when we’re inserting a new source, but there are many other
relationships that are represented in the Metathesaurus. So, that’s based on
the hierarchies in the existing relationships between the sources, and we also
sometimes make other relationships between sources to say these five things are
narrower than this one in some way.
DR. WARREN: So, I just want to make a point, Mark. This is one of the
reasons why both Blackford and I said we needed to get into the knowledge
representation area because the questions we’re asking are all about knowledge
representation and especially when you start talking about strings. SNOMED was
never developed as a term oriented vocabulary. It is a concept oriented
vocabulary. And, when people come new to it, they read the words and interpret
them the way they see the words instead of looking back as to what the meaning
is. Fortunately for UMLS, they look at the semantics of the concept and then
match them that way.
So, it’s a very subtle difference, but it has to do with the accuracy of
using it. One reason why there’s such a difficulty of mapping to ICD-10 is that
a lot of those labels have a lot of history behind them and evolution. And so
now, we’re taking something that really was developed as a set of terms and
trying to map them to concepts.
DR. HORNBROOK: And, I’m trying to deal with the terms that are specific in a
medical specialty like radiology of a big conceptualized space, a big
conceptual space that includes what is the target of their imaging, what’s the
energy source, how much —
DR. WARREN: Okay, what you’re talking about now is a data model –-
DR. HORNBROOK: Right, exactly.
DR. WARREN: And what set of terminology do you need to use to populate the
data model. And so, in SNOMED, it provides you a set of concepts that you can
populate a data model with. And, we have those concepts in different
hierarchies which helps to preserve context and semantics so that when you see
this concept, it comes with all of that machinery behind it, of context so that
it can populate your data model accurately.
DR. HORNBROOK: So, I’m wondering then how this Committee could then help us
bridge into the metadata model language to say, you know, here’s the way
cardiologists’ talk. This is the domains that they’re actually creating so we
cannot just be captured by the way cardiologists’ talk, we can talk about the
domains that are important to cover heart disease.
DR. WARREN: Well, SNOMED doesn’t talk about cardiology or neurology or any
of those. They talk about patient observations. So, one of the assumptions
under SNOMED is it’s a patient-focused terminology. And so, it’s all about
patient description. How you use it as a cardiologist or oncologist is the way
you’re going to populate the data models. Would you say that’s fair, Betsy?
MS. HUMPHREYS: I would say that was fair.
DR. WARREN: So, there’s that middle point of jumping off that we’ve got to
get, there’s a gap there of getting between where people, where cardiologists,
especially cardiology, their American College has done a lot of datasets and
data collection to reflect a lot of their work and so has oncology. Now,
there’s that marriage of what’s in those data models that’s not captured by a
terminology. It’s the same dialogue we have with HL7 as it’s putting together
its messages. Where does terminology fit? What goes into that data model and
what goes into the terminology model?
MS. HUMPHREYS: So, this is just showing you sort of base synonymy as we have
detected it so far, and I know that you’re probably interested in the
international ICF as well. And that situation, you notice there’s a 99.4
percent overlap with the one for children and youth. I assume this is by
design. And this has, it’s a much smaller body of terms obviously, and this has
nearly a 19 percent overlap with SNOMED as we inserted it.
MS. GREENBERG: That’s fascinating, though, because it looks like the overlap
is about the same between SNOMED and ICF as there is between SNOMED and ICD. I
wouldn’t think that.
MS. HUMPHREYS: You’re seeing two SNOMEDs there because it’s Spanish, and
it’s showing its overlap in both the Spanish and the English version of SNOMED
which makes perfect sense. Right?
Okay, so we have a mapping project that is in, what I consider to be, a
nascent phase in the sense that we have a set of participants, we have
requested allocation of somebody from CMS. We assume that will be forthcoming.
I know they have a few other fish to fry. And, the group that has met has
pretty well developed a project plan, has specified the use case, knows what
kind of tooling they’re going to use which will be the mapping component of the
IHTSDO Workbench. We are, however, essentially in a holding pattern pending
getting the input from CMS and also getting funding for this activity. So, I’m
afraid that that’s what I can tell you about it.
So, having told you what I know about mapping that is the current state of
at least what we’re working on with these maps, one of the things that we want
to get this project set up, there’s a little bit of, there is interest in a
number of players who would have developed, who really want this 10 CM mapping,
to come in and work with us on this including people from Kaiser Permanente.
So, we’re hoping that we can get this going, but it is literally true, we need
funding to get it, move it forward.
So, just to quickly run through a few other things that may be of interest
to the group, NLM has recently redone its access to UMLS. Resources have a new
entry point in terms of UMLS terminology services. The first thing that are
available, a series of browsers that are going to be more sophisticated, APIs
to the data as well. And, we now have available a SNOMED CT browser which has
this synonymy feature that we were just discussing, that is essentially all the
synonyms from everything else in the Metathesaurus that map to SNOMED can be
used to retrieve SNOMED in this browser. So, you can go in with other ways of
saying things and still end up where you want to be.
Coming relatively soon but in quite a basic fashion, again funding issues,
will be a US extension to SNOMED that will be covering two things principally.
One is concepts that are needed rapidly for meaningful use types of
applications. And, we want to be able to assign permanent identifiers to them
and get them set up prior to their inclusion which may very well be likely in
the international release of SNOMED CT.
The others would be things that are quite important here because of our
history or legacy or whatever and are less likely to be useful with other
countries because they may or may not have the same idiosyncrasies in their
health system that we do. So, we don’t expect this to be a massive extension.
What we want to do is get most of the stuff into the international release as
soon as possible and just have a relatively small set of items which represent
the peculiarities of health care delivery in the United States.
We will also be having a more sort of robust, interactive request submission
system, again, in its first phase later this summer. And, something that we’re
very interested in, in terms of international distribution workloads and so
forth, is the IHTSDO and NLM’s systems people who are actively participating in
this want to move to the workbench to a true international collaborative
environment where, for example, somebody going in to, say a certified editor in
the United States could go in, and they could actually see that the area that
they were trying to work on was already being worked on in Australia so that
there wouldn’t be people in different countries working on overlapping things
that would then have to be somehow resolved later on when both of them
submitted to the international release. But in fact, they could sort of look
over each other’s shoulders and perhaps avoid subsequent, you know,
time-consuming alignment activities.
This, you probably can’t see it, but this illustrates what I was telling you
before in the SNOMED browser which you put in inflamed eye, and lo and behold,
it comes up with a totally different term. An inflamed eye may have nothing to
do with SNOMED CT, but you still get it because this was connected somewhere
We have produced two subsets, and another one is in development. If I have
misspoken, I’m sure Judy can correct me. But, we have the core problem list,
and we have also a much smaller subset which covers routes of administration,
those that can be used in the NCPDP format. And, we worked it out with the FDA
and with NCPDP so that the set represents what FDA needs and is in SNOMED CT,
so it matches what NCPDP’s standard is for use in routes of administration.
That particular subset at the moment includes some things which are actually
US extension and some things that are in SNOMED CT international release now.
We expect them all to be in the international release. And, a nursing problem
list is being developed as well.
DR. MIDDLETON: Betsy, just on the core problem list subset, is the
expectation reasonable that that would be managed as an ongoing resource?
MS. HUMPHREYS: Yes. Basically, what we’re doing now is we update it every
time SNOMED CT comes out, and we update it also every time we release a new
version of the UMLS. And, we cannot instantaneously incorporate a new version
of any source into the UMLS because we are doing this actual, you know,
semantic connections. And, although we make heavy use of very sophisticated
flexible tools and inferences and so forth, in the end, a human being has got
to vet the matches.
DR. FRANCIS: And, this is probably a really way back sort of question. I
mean, I’m not a medical person at all. What I’m trying to get a handle on as I
hear you talk is I keep hearing people say that SNOMED has enormous advantages
as a semantic language because, and it will be able to, for example, if we’re
trying to figure out how to manage parts of a record differently, that SNOMED
has all kinds of advantages for doing that.
So, starting from there, I’m trying to figure out when you talk about what
we can find in the thesaurus and what we don’t find in the thesaurus, here we
have an example, I’m two slides back I think or one slide back. I was going to
plan to ask this question earlier, but I saw the slide with the ocular
hyperemia. So, there’s an example where you matched. Could you give me some
examples where you’re not able to? Just so I know so I can have something of a
feeling for —
MS. HUMPHREYS: I should have brought some examples of where we did not find
matches between 10 CM and SNOMED CT. And, I’m really sorry I didn’t bring any
with me. I think, though, when people talk about the advantages of SNOMED, in
many cases, they are talking about the advantages of the structure of SNOMED
and the relationships within SNOMED of concepts to each other in definitional
ways. And, the ability to go in and find at varying levels of granularity
groups of patients and the ability to not only look for everybody who has whole
thing, but to look for people who are at higher levels, groups of patients. And
so, I think much of the advantage of SNOMED CT is the connections and the
hierarchical and definitional relationships to other concepts that are in it.
DR. FRANCIS: Within SNOMED.
MS. HUMPHREYS: That’s right.
DR. FRANCIS: That’s why I was just trying to get a feeling for if you
plugged from one to the other what you’ve got and what you don’t have.
MS. HUMPHREYS: Well, I think what I should do is send you some examples, but
I think that in many cases when you say, you can probably say almost anything
in SNOMED CT, but the question is, did they create a pre-coordinated term for
it or not? And that, of course, is the age old question in any vocabulary
because you can say, well, I’m going to classify my patients, and I’m going to
classify the people who have X but also are diabetic or people who are
diabetic, and they have these associated things.
And, there are just different ways of categorizing, and in some cases, they
put all of that into one concept. And in others, they don’t. You know, they
expect you to, in essence, say those things separately, and it’s a deep, dark
DR. HORNBROOK: Are you aware at all whether EMR vendors are tying into these
data sources, invoking them in their software?
MS. HUMPHREYS: Yes. Actually, I am. The fact that there is an increasing use
of SNOMED CT in problem lists, part of this obviously has to do with the
meaningful use and the EHR certification regulations, and there is increasing
use of, I’m going to mention a little bit about RxNorm drug vocabulary and
LOINC and there’s increasing use of those as well.
DR. HORNBROOK: The nursing problem list would be very logical for an
MS. HUMPHREYS: The nursing, by the way, the issue here is it’s a subset
because the nursing problems are in SNOMED.
DR. HORNBROOK: They’re embedded.
MS. HUMPHREYS: But, what was not the same way, all of the frequently seen
problems were in SNOMED, but it became fairly useful for us to use some
statistical frequency of seen problems at a variety of people and say, here.
Here’s the subset that covers, you know, 85 percent or 90 percent of that if
you want to get started. And in essence, this is the same thing. The concepts
are there, but it’s making them, putting them in identified subsets so people
can just pull them out if they want to.
DR. WARREN: And Mark, back to your question about EHRs, that’s probably one
of the questions you need to ask your vendor is, what is their strategy for
mapping to these terminologies because some of the vendors have incredibly good
strategies and some don’t. And then, if anybody wants to know more about the
nursing problem list, I was lead on that. So, I can tell you more about what we
DR. TANG: So Betsy, can I ask you, the core problem list from SNOMED, has
that been mapped to ICD-10 CM, for example?
MS. HUMPHREYS: No. You missed, Paul, that very little has been mapped to
ICD-10 CM. We have a, as I was showing the data, the insertion of 10 CM into
the UMLS, we discovered relatively little direct synonymy. And, I’m not
claiming that we necessarily discovered it at all because we just recently did
that, but it’s very small.
And the, as I was just saying, we have a project group together, a plan and
approach about how we’re going to go do it. We will do frequently occurring
ones first, but we’re sort of at a wait state at the moment because of the
funding for the mapping project.
So, a couple of things that I think should be of interest to you, you may
have noticed these messages when they happened, Kaiser Permanente has donated
their CMT or is donating it in pieces over a period of time for it to be
incorporated as appropriate into the international release of the US extension
or the US extension of SNOMED CT. They have done mappings, and their mappings
will also be made available. And, they have not done much mapping to ICD-10 CM,
although they have expressed willingness to be part of this US project to come
up with a more definitive mapping.
The second thing is one that I’m very excited about, and that is that the
model, the business model for SNOMED CT is, countries have a member, and if you
are a member you’re paying your membership fee, and as a result of your
membership fee, everyone in the country can use it. And then, if you are not in
a member country, you can get an affiliate license. If you are in one of more
than 40 low income countries as defined by the World Bank, then your licenses
are free. If you are not in that group, then you’re paying an annual affiliate
license, not outrageous in my opinion, but you’re paying a license fee.
This has caused some problems, more from the point of view of people wanting
to include SNOMED CT names and codes in resources that are internationally
available. So, the IHTSDO has recently come up with a policy which I am very
enthusiastic about. And that is, you can apply and describe a case where you,
as a license holder, would like to create a resource or make something
available that will have the names and the codes, not the full relationships
and so forth. And, you can get approval to do this.
So, the National Library of Medicine was sort of the first one to take
advantage of this policy because we want to have SNOMED CT names and codes be
used as the names of diseases in some international genetic information
databases that are going out. We thought it would be very unfortunate if they
had to standardize on a different name for a disease that was in SNOMED. So,
And, it will also permit, it permits the case where, for example, HL7 could
say, here are the 40 allowed values for a particular part of a message. And,
they could be SNOMED CT names and codes and be distributed with HL7s. So, I
think there’s going to be a lot of advantages to this.
The second case, that was just approved, was that people who are developing
measures, quality measure specifications here in the United States, sort of
they were developing them anyway, but there’s a lot of impetus to this because
of the meaningful use criteria that in fact those measure specifications will
be able to be published, really available with the SNOMED CT names and codes in
them. So, I think this is a good thing. It should obviously be useful for some
public health purposes as well.
So on the, just to very quickly give you a couple of updates, this week’s
release of RxNorm includes greatly improved NDC coverage in two directions. The
obsolete NDC codes are segregated, so if you don’t want them, you don’t have
them. They’re there, but you can easily exclude them. And First DataBank, to
whom we are very, very grateful, has wanted us to include their NDCs in RxNorm
for the first time. And this, of course, is a very heavily used source. And so,
this is very good.
And, just to tell you that recently, a CMS-sponsored study that was done
actually a little while ago but has seen the light of day in publication now,
verified that a fairly large sample of ambulatory care prescriptions, the
coverage of the drugs in RxNorm was essentially 100 percent. So, I think that
this is good.
There are all kinds of new features on the RxNav browser and APIs if you are
interested in getting drug information. There are also frequently, some very
good work on frequently occurring tests and test results in LOINC to give
people going forward.
And the newborn screening work, both the HL7 messages, draft messages and
the Guide to the Newborn Screening Test LOINC standard and SNOMED CT for the
conditions, that’s coming along very nicely. I’m very excited about the fact
that the state of Kentucky is really applying the messaging and moving forward
with a real test of it for statewide exchange of newborn screening data.
DR. CARR: Vickie?
DR. MAYS: Can you say a little bit about the mental health diagnosis like in
terms of what’s happening with DSM and ICD-9 and ICD-10 because we’re going to
move to a new DSM-V? And, there’s a real push now in primary care to actually
do the, you know, to do diagnosis within that office. So, is SNOMED at all
involved in any of that?
MS. HUMPHREYS: Well first off, Marjorie, do you want to talk about the
relationship of DSM to ICD-10 CM?
MS. GREENBERG: Let me just ask, is Donna Pickett on the phone? She was going
to call in. But, I know that we have worked now for many years, 9 CM actually
worked very closely with the American Psychiatric Association, aligning 10 CM
with DSM-IV. And, I know they’re working on DSM-V, but it’s still being tested.
It hasn’t been finalized.
DR. MAYS: The release should be in two years.
MS. GREENBERG: Two years? So, there is a mental health, meanwhile there’s
also a Mental Health Topic Advisory Group for ICD-11. So, it’s sort of like the
budget as Jim mentioned. We’re working on 9 CM, we’re working on 10 CM, we’re
working on SNOMED, and we’re working on ICD-11. And so, the question of once
DSM-V is finalized, I mean, I think they may actually, as we all may know or
may not know, there is a partial but a pretty substantial freeze on updates to
ICD-9 CM starting with October 2011.
So, those will be the last updates actually, the last ones to 9 CM unless a
new disease, I hope it doesn’t happen, is identified or 9 CM Volume 3 unless a
new technology, surgical technology, is identified. And so, and at that time,
ICD-10 CM will also be at the same freeze until October 2014.
In the meantime, the Maintenance and Coordination Committee will be meeting,
and so they’ll be taking on more recommendations and proposals that can be put,
first of all to determine whether they meet that threshold of not being covered
by the freeze because of their novelty, new aspects and also accumulating them
for the 2014 update of 10 CM. Of course, at that point, nobody will be using 9
CM. So, I think the APA may be bringing a few things that they’ve identified
and aren’t going to change, you know, to this process. But, it’s a little
premature for all of DSM-V because it still hasn’t been finalized.
There, of course, will be an effort to align by all parties DSM-V with
ICD-10 CM, ICD-11. I mean, we have quite a good alignment with 10 CM, but then
we haven’t yet implemented 10 CM and DSM is moving to V. I mean it’s difficult
because, I’m sure Betsy could speak more eloquently than I could, about all
these different terminologies being at different points.
MS. HUMPHREYS: Well I think that the issue, of course, with DSM is that the
real substance of DSM is the clinical definition of what the disorder is —
MS. GREENBERG: That’s true, and we’ve never —
MS. HUMPHREYS: — rather than the name of the disorder which changes less
frequently than sometimes the clinical definition or the determination that
something either is or isn’t considered a disorder.
MS. GREENBERG: But certain things, I think, are being at least recommended
though it hasn’t been finalized for moving or being in different places or
being defined differently which would have implications if you wanted to align
them. So, there are some things that could be significant changes. But, the
definition part, now that is being built into ICD-11, so there will be more
coordination needed there.
DR. CARR: Paul and then, Mike.
DR. TANG: Thanks, Betsy, for your update. It was encouraging to hear that
you said that there’s some uptick in the use of SNOMED CT for problems and
presumably some related to meaningful use. So the next question, of course is,
what’s your recommendation for either what vendors or providers should do in
making that transition awaiting the mapping from SNOMED to ICD-10 CM?
MS. HUMPHREYS: What my recommendation would be?
DR. TANG: Yes. Well I mean, because everybody is essentially asking
everybody, including us, what to do if we’re going to take advantage of the
richness and the advantages of SNOMED, and they follow our recommendation to
use it. And the problem is, how are they going to get it translated into their
billing system and etcetera? Because there’s obviously, there’s private
companies that are filling the niche which, in my mind, add a free text
vocabulary aside, and you probably know who I’m talking about. But, what
alternatives should we offer them?
MS. HUMPHREYS: Well, I guess my feeling about all of this is that my own
view is that it would make a great deal of sense for there to be an official
mapping. And so, the issue is whether, I believe that that would be, at the
macro scale, the most cost-effective thing to do at this moment. However, it’s
sort of like everything else with electronic health records in terms of who
would benefit and who would pay. And you know, I think that it’s a good issue
for this Committee in terms of what level of priority you would put on that.
DR. TANG: So, I understand that so anyone that would be ready to and has
your project set up if we could, if funding could be made available that
MS. HUMPHREYS: I think we need two things. We need, and I’m fairly certain
we can get both, I mean, we can get the one, I mean, we need some participation
from CMS so that we can set this thing up in the evaluation of it in a way that
will give us confidence that people who make use of it won’t be accused of
systematic anything, you know, that sort of thing.
So, we need that, and we need funding. And in my view, we need funding to do
it, but then we actually need a bit of funding to get some people to do some
rigorous testing of it because, you know, it seems to me that I don’t assume
that it’s anyone’s desire that we move from 9 CM to 10 CM and have major
immediate shifts in billing.
DR. CARR: Mike Fitzmaurice?
DR. FITZMAURICE: Thank you for this presentation, Betsy. I think very highly
of the work that NLM is doing. It’s the basis for a lot of decision support
that we may see in five, ten, fifteen years. It’s essential that it be done. I
was bothered a little bit by WHO being in gray and the reason for WHO being in
My question is, will ICD-11 be much more closely based on SNOMED terms,
concepts, and codes than it is now? In other words, there’s a vision that some
people have to build up from the SNOMED codes to ICD-11 so that if vendors go
out and code everything that a physician puts in in the SNOMED, if what they
pick generates a SNOMED code, that it will roll up to payment, to ICD-11, roll
up to DRGs and other classifications. What is the likelihood that WHO and
IHTSDO will cooperate and link together and then we can achieve some of that
MS. HUMPHREYS: Well first off, I just want to say that I think that we have
to achieve something like that because the perpetual maintenance of these
mappings is nuts. I mean, as a long-term strategy for doing anything, it is
something that will eat more resources in the long run and to what good end, I
can’t, so I think we have to get there. I think that it is the goal of everyone
to do that for ICD-11, and I also feel that that is another area where in my
mind we really need some initial work and some testing and evaluation. I
personally cannot regard this as anything, I mean, I regard it as the way we
absolutely have to go. I don’t know that we have demonstrations of exactly how
to do it.
So, I feel like it’s going to take us a while to move to ICD-11 anyway, and
I would really like to see some initial work on, you know, some test areas and
models and whatever that we then could do some rigorous testing of. It is
absolutely the way to go, but I’m not sure that we actually know how to do it.
MS. GREENBERG: It’s actually is part of the collaborative agreement between
the two organizations. So, there’s some commitment there, but as Betsy said,
it’s not just a, as you already said, an easily one-to-one type thing. So, it
really needs a research budget.
DR. CARR: Okay, I think we have time for one last question. Raj?
DR. CHANDERRAJ: As a practical question, I know there are about 99,000 codes
in ICD-9. How many are there in ICD-10, and how many are there in SNOMED?
MS. HUMPHREYS: In ICD-10 CM, there are about 97,000 which is quite a few
more than in, I believe, ICD-9 CM at least the way we count them. In SNOMED CT,
there is something like 300,000 concepts, but SNOMED CT’s scope is much broader
than ICD-10 CM’s or 9 CM’s. So for example, a large piece of SNOMED would be
organisms, you know, all the different types that can get you or whatever. And,
there would be a large anatomy section and pharmacies, so it has a broader
coverage than, it’s more comprehensive in coverage than the other systems.
MS. GREENBERG: This made me think of, go back to this issue of the lack of
synonymy, the lull rate, etcetera. It’s always been my assumption, again, it’s
easier said than done, but that by time to develop these maps in this imperfect
space because ICD-10 is not based on SNOMED terms, but I think it’s reasonable
to think we’re going to be in this space for at least another ten years or
whatever, that doing so would allow us opportunities to improve, to make
changes and improvements in both SNOMED CT and in ICD-10 CM. And, some of that
would increase then the alignment, you know, that would be part of the, I guess
I would just leave with the, and you’re agreeing with me, Betsy.
So I mean, neither one of these can be done because they both do have a very
structured updating process, both SNOMED does and ICD. So, rather than seeing
this, I mean it’s certainly a challenge, but I think we see this as an
opportunity. The more we can get on with this and do this type of mapping and
testing, the more fodder we’ll have for making those improvements.
MS. HUMPHREYS: I agree with you entirely. And, I certainly think that my
expectation would be that there will be, looking at what’s done now and
attempting to match things the way they are now, will absolutely lead to I
hope, sensible decisions about what should be done on either end of this so
that it actually matches up correctly and is useful for people.
And you know, in a much simpler example, which I love in my field which is
library and information science, a group got together years ago to write an
algorithm that would sort in machine the same way you were supposed to file
cards, in the same order. And, they discovered that a machine could not sort
that way because human beings were reading and adding information to what was
on the card in order to sort it properly. And, the people who were running that
committee brilliantly decided that the existing filing scheme was fatally
flawed, and we should just go to one where a machine could sort it correctly.
So, I think that we may have a lot of this in some of these terminologies.
DR. CARR: Thank you very much, Betsy. I appreciate it. And with that, the
meeting is adjourned. Safe travels everyone.
(Whereupon, the meeting adjourned at 2:05 p.m.)