[This Transcript is Unedited]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
(FULL COMMITTEE MEETING)
March 4, 2004
Hubert Humphrey Building
200 Independence Avenue, S.W.
Washington , D.C. 20201
CASET Associates, Ltd.
10201 Lee Highway, Suite 160
Fairfax , Virginia 22030
Committee Members :
- John R. Lumpkin, M.D., Chair
- Jeffrey S. Blair
- Justine M. Carr, M.D.
- Simon P. Cohn, M.D.
- Peggy B. Handrich
- Richard K. Harding, M.D.
- John P. Houston, J.D.
- Stanley M. Huff, M.D.
- Robert W. Hungate
- Eugene J. Lengerich, V.M.D.
- A. Russell Localio, Esquire
- Vickie M. Mays, Ph.D.
- Harry Reynolds
- Mark A Rothstein, J.D.
- Donald M. Steinwachs, Ph.D.
- C. Eugene Steuerle, Ph.D.
- Kevin C. Vigilante, M.D.
- Judith Warren, Ph.D.
HHS Executive Staff Director :
- James Scanlon
Executive Secretary :
- Marjorie S. Greenberg
Liaison Representatives :
- Judith Berek
- Virginia S. Cain, Ph.D.
- J. Michael Fitzmaurice, Ph.D.
- Edward J. Sondik, Ph.D.
- Steven J. Steindel, Ph.D.
- Aldona E. Robbins, Ph.D.
TABLE OF CONTENTS
- Call to Order, Welcome and Introductions – Review of Agenda – John R. Lumpkin, M.D.
- Claims Attachment Letter – Simon P. Cohn, M.D.
- Privacy Letter – Mark A. Rothstein
- Health Statistics for the 21st Century
- Dr. Daniel Friedman
- Barbara Starfield, M.D.
- Dr. Carl Volpe
- Quality Report – Robert W. Hungate
P R O C E E D I N G S [9:10 a.m.]
DR. LUMPKIN: Good morning. Welcome. Just to get started, we will start off
My name is John Lumpkin and I am the chair of the committee and senior vice
president of Robert Wood Johnson Foundation.
Why don’t we go around the room and make introductions. Then we will review
the agenda. We have a couple of housekeeping items and also as we go around the
committee and do introductions, if you have any conflicts with the subject
matter of the meeting that you need to declare, please do so at that time.
So, let’s start off with Marjorie.
MS. GREENBERG: I am Marjorie Greenberg from the National Center for Health
Statistics, CDC and executive secretary to the committee.
DR. LUMPKIN: Jim wants to clarify. That is financial conflicts, not
MR. HOUSTON: I am John Houston with the University of Pittsburgh Medical
Center and I don’t have any conflicts, either emotional or professional.
DR. CARR: I am Dr. Justine Carr from Beth Israel Deaconess Medical Center
in Boston, Department of Health Care Quality.
DR. HUFF: Stan Huff with Intermountain Health Care, the University of Utah
in Salt Lake.
I need to recuse myself from HL7, LOINC, and ICD-10 PCS.
DR. LENGERICH: Gene Lengerich, a member of the committee and from Penn
DR. LUMPKIN: Stan, could you explain that?
DR. HUFF: This is a direction from the ethics committee. So, when I do the
non-disclosure thing, they — because I hold a position as a co-chair on HL7,
they think I have a personal bias. I am also a co-chair in LOINC and so I have
a personal bias. There is no financial interest there. And about two years ago,
I was a one-time consultant to 3M, who produces ICD-10 PCS. I think that one —
actually there is no business relationship any longer. So, when I reviewed
again, that one will probably go away.
That is the explanation, sir.
DR. LUMPKIN: I think the important addition I would make to that, it is not
the issue that you might — that you have personal bias, but it is the
appearance of bias. Because we love and trust you.
DR. HUFF: Thank you for that vote of confidence.
MR. LOCALIO: I am Russell Localio, University of Pennsylvania School of
Medicine. I am a member of the committee.
DR. WARREN: I am Judy Warren, University of Kansas. After talking with
Stan, I don’t know what I have because I am also a co-chair at HL7 with no
financial bias and I am a consultant at the SNOMED Editorial Board with no
reimbursement or anything like that either.
MS. GREENBERG: Just for some clarification, one’s interests are defined
broadly. So, it is not just a question of financial. So, if you have a
leadership position in an organization, then generally there is a waiver that
allows you to participate in all the general matters, but not matters that are
specific to that issue.
MS. BEREK: Judy Berek. I am the liaison for the Centers for Medicare and
MS. TORIARES: Nancy Toriares(?) from the U.S. Census Bureau, reminding
everyone that you have just seen your last decennial census long form. It is
DR. LUMPKIN: What if we long for the form?
MS. TORIARES: Well, they long for $12 million to do it. Reminding everyone
that this replacement, the American Community Survey, starts in July of this
DR. ROBBINS: Aldona Robbins. I am the liaison from the Board of Scientific
Counselors, National Center for Health Statistics, CDC.
MR. REYNOLDS: I am Harry Reynolds, Blue Cross and Blue Shield of North
DR. STEINWACHS: I am Don Steinwachs at Johns Hopkins, Bloomberg School of
Public Health, member of the committee.
DR. STEINDEL: Steve Steindel, Centers for Disease Control and Prevention,
liaison to the committee.
MR. BLAIR: Jeff Blair, Medical Records Institute, member of the committee.
MR. HUNGATE: Bob Hungate, Physician Patient Partnerships for Health, member
of the committee.
DR. MAYS: Vickie Mays, University of California, Los Angeles, member of the
DR. COHN: Simon Cohn, national director for health information policy for
Kaiser Permanente and a member of the committee.
MR. SCANLON: I am Jim Scanlon with the Office of the Secretary, HHS and I
am the executive staff director for the committee.
MR. SONDIK: Ed Sondik, director of the National Center for Health
Statistics. I am entirely conflicted.
MR. FANNING: I am John Fanning from the Department of Health and Human
DR. FRIEDMAN: I am Dan Friedman, an independent consultant in population
and public health information services from Massachusetts. To quote the
chairman of the committee, now I have a teenage daughter at home, so I have
many conflicts everyday.
MS. LORRAINE: Kathryn Lorraine(?) Office of Policy, FDA.
MS. KOPPELMAN: Hi. I am Lacey Koppelman, Office on Women’s Health at the
Department of Health and Human Services. I am actually representing Suzanne
Haynes, who is senior science advisor in the Office on Women’s Health and she
is a member of the subcommittee.
MS. PICKETT: Donna Pickett, National Center for Health Statistics, CDC.
MR. HITCHCOCK: Dale Hitchcock, Office of the Assistant Secretary for
Planning and Evaluation, co-lead staff of the Subcommittee on Populations.
DR. KYLE: Frank Kyle(?), American Dental Association.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics, CDC.
MR. RUMMEL: Nick Rummel(?), HIPAA Compliance Alert.
MS. KAPLER: I am Evelyn Kapler(?) with the Office of Public Health and
Science at HHS.
MS. EVELYN: Brenda Evelyn, Office of Special Health Issues, Food and Drug
MS. FRIEDMAN: Maria Friedman, CMS and staff to the Subcommittee on
Standards and Security.
MS. GRANTON: Good morning. Miriam Granton with the Office of Disease
Prevention and Health Promotion and I staff the Subcommittee on Population.
MS. RYAN: Hi. I am Colleen Ryan. I am with the Indian Health Service and I
also am render staff for the Subcommittee on Population.
MS. WILLIAMSON: Michelle Williamson, National Center for Health Statistics,
MS. WATTENBURG: Sara Wattenburg, Substance Abuse and Mental Health —
DR. LUMPKIN: That is what you get for testing the equipment this morning.
If you hadn’t tested it, it would have worked fine.
We have Peggy, if you would introduce yourself and state any conflicts of
an interest, financial nature. We have stopped with the emotional conflicts.
MS. HANDRICH: I am Peggy Handrich, a member of the committee from
DR. LUMPKIN: Great. Welcome.
Do we have anyone on the phone yet?
PARTICIPANT: Not yet.
DR. LUMPKIN: Richard Harding and Mark Rothstein will be participating in
part of the meeting on the phone. Mark is recuperating. So, we wish him all the
best and hope he gets back on his feet very quickly.
Let me start off with a couple of logistics issues.
MR. BLAIR: How about if they could turn the air conditioner on. That is a
good logistics issue.
DR. LUMPKIN: This is March. This is a federal building in March in
Washington. If we can do something about the temperature in here — it is a
little bit too warm.
We were discussing a couple of logistics issues. The first is if you intend
to leave before the end of the meeting tomorrow, please let me know during one
of the breaks, just so I can manage a quorum. I don’t think we have a problem,
but I just would like to know.
Second is that there is a caravan leaving tomorrow morning at 8:15, for
those of you who want to go out to the wilds of Virginia to get your —
MS. GREENBERG: No, no, no, Maryland. Hyattsville, Maryland.
DR. LUMPKIN: The wilds of someplace out here east, outside of D.C.,
Maryland, I guess, for I.D. badges.
MS. GREENBERG: Opportunity to see the beautiful new NCHS building.
DR. LUMPKIN: And an opportunity to see the new NCHS building that is so
accessible that — who is going to go along so that the cab driver can find it?
Gracie will go along with you.
MS. GREENBERG: This is to get your badges. It does make life easier once
you have the badge.
DR. LUMPKIN: Okay. So, anyone besides Justine going? Okay. So, hook up with
Jackie, she will give you the details also during the break.
MS. GREENBERG: Then if you are — I guess before the lunch break, maybe we
will take a head count for the dinner.
DR. LUMPKIN: Okay. We have two people on the phone. If you could introduce
DR. HARDING: Richard Harding is on the phone.
MS. HORLICK: Gail Horlick.
DR. LUMPKIN: Gail. Welcome.
MS. HORLICK: I have tried — I was the first caller before and they finally
told me to hang up and try again.
DR. HARDING: That is what happened to me, too.
MS. HORLICK: I won’t be able to stay for the whole meeting, but I will call
in when I can.
DR. LUMPKIN: Great. Thank you.
Let’s go through the agenda. The first item that we have on the agenda —
are there any additions or changes to the agenda? Mike Fitzmaurice, who is our
liaison from AHRQ has just arrived. So, I am going to embarrass you by noting
that fact and having you introduce yourself.
Any other changes to the agenda? Okay. Seeing none, let’s proceed.
The first item on the agenda, we were just passed out a copy of the claims
attachment letter. So, I am going to turn that over to Simon.
DR. COHN: Good morning, all.
This is a letter that is being brought forward as a result of hearings of
the Subcommittee on Standards and Security. The hearings occurred on December
10th and yesterday, March 3rd. I think as you all know, claims attachment is
one of the standards under HIPAA. This has been something that has been
percolating around for awhile.
We thought we were going to see one a couple — NPRM a couple of years ago
and then sort of nothing happened with it. In the meantime, the actual standard
was taken back to the standards organizations and was basically retooled. So,
the subcommittee wanted to take a look, find out what was happening, as well as
what needed to be done with all of that. This is really the focus of this
As you know, this is typically an occasion to suggest modifications,
changes, additions or others to letters with the idea that the action would
occur tomorrow on the subcommittee. If there are major modifications, we will
obviously deal with it during the subcommittee break.
So, John, with your permission, we will go over it paragraph by paragraph
on this one. It is a relatively short letter.
DR. LUMPKIN: Please.
DR. COHN: Okay. This says, “Dear Secretary Thompson: As part of its
responsibility under the Health Insurance Portability and Accountability Act of
1996, the National Committee on Vital and Health Statistics is responsible for
studying, selecting and recommending standards for electronic health claims
“To fulfill these responsibilities, NCVHS’s Subcommittee on Standards
and Security recently held hearings on standards for electronic health claims
attachments on December 10th, 2003 and March 3rd, 2004.
“The subcommittee heard testimony from providers, health plans,
vendors, associations and standards development organizations about the need
for both basic and advanced functionality in a claims attachment standard. For
example, such a standard should be flexible so that providers with minimal
infrastructure can electronically transfer claims attachment information to
health plans and clearinghouses. At the same time, the standards should possess
the flexibility to permit users with more sophisticated infrastructures to
fully leverage their investment in information technology.
“Health Level 7 and the Accredited Standards Committee X12 have been
working to develop a HIPAA claims attachment standard that can meet these
DR. LUMPKIN: Any questions or comments on that paragraph?
DR. COHN: I know, it isn’t very exciting so far, is it?
DR. LUMPKIN: Okay. Simon, you are on a roll.
DR. COHN: Yes, I know. I am actually a little concerned that nobody is
making comments so far, but we will encourage that as we go along.
The next paragraph, “The subcommittee heard the need for demonstration
projects and pilot studies to document the benefits, costs, work flow
requirements, implementation challenges, privacy concerns and best practices
associated with the claims attachment standard. NCVHS urges the department to
support…” — and probably we want to say “support and
encourage” — “…several different claims attachment demonstration
projects and pilots that would include a broad representation of affected
“A pilot proposed by Empire Blue Cross and Blue Shield is an example
of the kind of projects that should be developed and funded. NCVHS recommends
that these demonstration projects and pilots occur expeditiously so that the
results and findings will feed into the development of the claims attachment
DR. LUMPKIN: Any questions or comments on that paragraph?
DR. STEINWACHS: Simon, just one — the one sentence that said the pilot
proposed, for the naive person, who doesn’t know what is going on, is there a
little descripter of what that pilot is about? It just sort of hangs here that
there is one organization who proposed the pilot should be encouraged, but it
doesn’t — or am I
DR. COHN: Well, no. You bring up a good question. I mean, we obviously,
have testimony from them in a proposal, which we believe has been already
proposed to CMS, which actually used to be a parenthetical here, but we thought
it was a little strange —
DR. LUMPKIN: Is it safe to assume that what the pilot is is the pilot is a
pilot testing a model of a claims attachment?
DR. COHN: Yes, it includes — I mean, this is basically a Medicare carrier
and Medicaid also, I believe.
DR. COHN: Just Medicare. Maybe testing it with various institutional and
physician providers and how it would all work. Do you feel that there is a need
for a little more explanation of what the whole thing is?
DR. STEINWACHS: I would put a little more explanation or I would drop it,
either way. But, you know, if it is important, I would give a little more
explanation, just what you said, I think would be helpful to someone who has no
idea particularly why you are citing this.
DR. LUMPKIN: Does that sentence really add value to the —
DR. STEINWACHS: We could drop it, too. I mean, that would —
DR. LUMPKIN: Because we say in the sentence before that urges the
department to support several different claim attachment demonstration projects
and pilots. It just kind of sits out there.
DR. STEINWACHS: It may be easier to drop it.
DR. COHN: It might be easier to drop it. I think we will have to — maybe
that is a subcommittee discussion to see if we — I mean, the only reason to
have that in there would be if we want to make sure particularly that that one
gets some notice so that it gets funded and if there are any issues related to
it not getting funded. But I think we can talk about that in the subcommittee
DR. LUMPKIN: Were there other pilots presented?
DR. COHN: No. Some people are talking about pilots, but none were at the
stage where they could come forward and say, yes, we have a pilot put together.
It needs either encouragement or —
DR. LUMPKIN: You could —
MS. GREENBERG: How about if we say we heard about a pilot proposed by
Empire Blue Cross to do x and this is the type that should be encouraged and
funded — or developed and funded.
DR. COHN: I would say the subcommittee can take back this comment and —
DR. LUMPKIN: If it is okay with you, I think maybe we could ask the
subcommittee to work on that.
DR. COHN: That is what subcommittees are for, to work on these sorts of
Any other comments on that paragraph? I mean, the people we are sort of
trying to talk about now, it is like a model that somewhat resembles the
Medicare Modernization Act and e-prescribing with the idea that a new standard
gets piloted before it gets nationally mandated and the results of that pilot
get put into the improvements in that standard before it is announced as a
final rule. So, that is the model here.
DR. STEINWACHS: That would be great.
DR. LUMPKIN: And I think just one other addition. If you do include that,
you might want to add a modifier that does describe them as the Medicare and
Medicaid carrier or intermediary because it sort of — from a casual reader,
why are we picking on Blue Cross/Blue Shield. It is because they have already
got some designation and relationship with CMS as a designated intermediary.
That is an important distinction.
MS. BEREK: I would use contractor because they are both a carrier and an
DR. LUMPKIN: Okay. Whatever the appropriate term is.
MS. BEREK: They will have a new name in a few years.
DR. COHN: So, what do we call it, contracted?
MS. BEREK: If you call it a contractor, it covers both carrier and
DR. LUMPKIN: See, and I still want to call them HCFA. I just can’t keep up
with these names.
DR. COHN: We will ask you to make sure that we are correct on this one, but
the more we get into this, the more likely there is going to be — you are not
going to see that sentence when it comes back.
Next paragraph, “NCVHS understands that a notice of proposed
rulemaking concerning electronic claims attachment currently is under
development. We urge the department to issue this NPRM as soon as possible.
This will motivate the industry to participate in pilot projects and invest in
the needed infrastructure. We also encourage the department to provide for
public comment on the results of the demonstrations and pilots before a final
rule is promulgated.”
Then, “NCVHS wishes to thank you for the opportunity to offer these
recommendations.” Are people comfortable with that final part?
DR. LUMPKIN: Now if I can sort of understand this, our prior recommendation
was to hold off on doing the claims attachment rule until the field was clearer
somewhat. Now we are saying we think the field is cleared up enough to issue
the rule and use that as a basis for evaluating the demonstration projects and
then coming out with the final rule.
DR. COHN: I don’t actually believe we actually had a letter on claims
DR. LUMPKIN: Okay.
DR. COHN: I think that we had held hearings a number of years ago. You
know, it is usually our practice to wait until an NPRM is published and then
make comments about the NPRM.
DR. LUMPKIN: No, we did have a letter on — way back when, we did have a
letter on claims attachment. We did make a recommendation, which didn’t proceed
and I think at some point — and whether we did it by letter or recommendation,
we did say, yes, it is a good idea to hold off.
But that really probably doesn’t have any impact on this. The important
point is is this process that we are recommending, which is to use the NPRM to
sort of set the basis for the demonstration projects and evaluation as they
moved to a final rule is the approach that we are recommending.
DR. COHN: Is that said clear enough here for those to understand?
MR. BLAIR: One of the things that we struggled with for a little while was
to get a balance. We recognized that software developers and folks that would
be doing pilots might not invest until they saw a serious intent by HHS, ala in
the form of an NPRM before they would really move forward on this.
So, we felt it was important for us to encourage that that NPRM be set
forth as soon as possible. At the same time, we were also hearing concern that
people wanted the opportunity to make public comment in the rulemaking process
after the pilots — after the results of the pilots were available and before
the final rule was issued. So, it was the balance between those two things that
we have reflected in the last paragraph.
DR. LUMPKIN: I would wonder if perhaps that — what you just said might be
incorporated a little bit more in that last paragraph because really the
sentence says this will motivate the industry to participate in pilot projects
and I think it would be stronger if we said that what we heard in hearings is
that the industry is waiting for an NPRM in order to move forward and that
there was concern that there be adequate ability to comment on the pilots. I
think that would make it a little bit stronger paragraph.
MR. SONDIK: I have a question on the — it says support and encourage or
encourage and support, I think, might even be better. Where would the funds
come from for this demonstration or pilot? I mean, do you know of a pot that is
DR. LUMPKIN: I think they were going after your budget.
MR. SONDIK: Why do you think I told you I was conflicted?
DR. COHN: Money is always an issue in all of this. I think it is why we use
the word “support and encourage.” “Support” obviously has
many meanings, financial support, leadership, et cetera, et cetera.
MR. SONDIK: Support and encourage, I like that, but the way it was written,
when it said developed and funded, then it made it seem as if there was a pot
simply waiting or that it was very clear how this would be funded.
DR. COHN: Maria, can you comment on that from the CMS perspective?
MS. FRIEDMAN: We just recently got our budget. So, the jury is out on this.
CMS recently just got its budget and I know the folks on the Medicare side of
the house are looking at this. I don’t think they have made a decision on
funding this particular one. But I think the broader issue that we heard in
testimony, that because funding is problematic everywhere, this might be an
opportunity to do a public/private partnership or in some cases we heard that
the industry might be willing to kick in.
MR. SONDIK: My recommendation would be to less specific and if, in fact,
the offending sentence is going to be dropped or the —
DR. COHN: It may be okay.
MR. SONDIK: And I think the words encourage or encourage and support, I
think would be sufficient.
DR. COHN: It sounds like those nails on the coffin of that sentence are —
MR. SONDIK: You raised it.
DR. COHN: That is well said.
Okay. Other comments? The subcommittee will obviously take this back. We
will do some modifications to the letter and bring it back tomorrow.
DR. LUMPKIN: I want to make sure that I understand the looks I am seeing
and the body language that I am trying to interpret. I don’t see any major
obstacles to the intent of this letter just to refinement of the wording. Is
that a fair statement? Okay.
DR. COHN: Thank you.
DR. LUMPKIN: Thank you.
We will move on — has Mark joined us?
MR. ROTHSTEIN: Yes, I am on.
DR. LUMPKIN: Welcome. We are sorry you couldn’t join us.
DR. COHN: Are you okay?
MR. ROTHSTEIN: I am barely hanging in there but I am going to stay on the
call for as long as I can. I would appreciate if someone from the subcommittee
— I know that Richard is not present —
DR. HARDING: I am on the phone, too.
MR. ROTHSTEIN: Yes. Either John Houston or —
DR. LUMPKIN: John is right next to me.
MR. ROTHSTEIN: — Simon could take the lead in discussing the letter. That
would be very helpful to me.
DR. LUMPKIN: Great. Now, we have that under our Tab 4, I think. Thank you.
MR. HOUSTON: “Dear Secretary Thompson: As a part of its
responsibilities under the Health Insurance Portability and Accountability Act
of 1996, HIPAA, the National Committee on Vital and Health Statistics, NCVHS,
monitors the implementation of the Administrative Simplification Provisions of
HIPAA, including the standards for privacy of individually identifiable health
information, the privacy rule.
“The Subcommittee on Privacy and Confidentiality of the NCVHS held
hearings in Silver Springs, Maryland on November 19th and 20th, 2003. The
hearings, the first of several to be held over the next year, were intended to
gather information about the effect of the privacy rule on public health,
research and health care providers, health plans and consumers. In general,
witnesses at the November 2003 hearings reported less anxiety and confusions
about complying with the privacy role than did witnesses at NCVHS hearings in
“Several witnesses said that materials posted on the web site of the
office of civil rights, OCR, were helpful, but they also stressed the need for
OCR to expand its outreach and public education activities so that the privacy
rule can be implemented effectively. It was noted that not all covered entities
and consumers have access to the Internet.”
DR. LUMPKIN: Okay. Any questions or concerns about the introductory
MR. HOUSTON: I had one in particular. In the last paragraph that I read
regarding NCVHS hearings in 2002, I am wondering whether it might, rather than
saying 2002, simply say prior to the compliance deadline for the privacy rule,
because it sort of positions before versus after the deadline in a more
straightforward fashion than 2002.
DR. LUMPKIN: Mike.
MR. FANNING: Excuse me. Wants the text to strike in 2002 and insert prior
to the compliance —
MR. HOUSTON: Prior to the compliance deadline for the privacy rule.
MR. FANNING: Thank you.
DR. FITZMAURICE: On the second paragraph that starts out, “The
Subcommittee on Privacy and Confidentiality,” the second sentence says,
“The hearings, the first of several to be held over the next year,”
it is hard to have something and say it is going to be part of the next year. I
would say, “The first of several to be held over” and insert “a
12 month period,” striking the next year.
DR. LUMPKIN: How about just first of several to be held.
We will proceed on to the public health — let me just get an idea. For
those who have read this, are there enough concerns paragraph by paragraph or
do we want to read the whole thing and then just take — let’s do the section
MR. HOUSTON: “Paid public health. The privacy rule explicitly permits
disclosure of protected health information, PHI, for public health purposes
without need for an authorization. Accordingly, the main public health issues
involve covered entities that misunderstand the privacy rule and limit their
disclosure and their coordination of the public health provisions of the
privacy rule with other provisions.
“In general, the witnesses stated that misunderstanding of the privacy
rule by many covered entities was adversely affecting the reporting of
notifiable conditions to public health officials. At least one witness
suggested, however, that some covered entities were merely using the privacy
rule as an excuse to avoid the burden of public health reporting.
“Of course, it would be difficult to assess the motivation of a
covered entity for being reluctant to make reports to public health officials.
One of the witnesses representing the Council of States and Territorial
Epidemiologists, CSTE, testified about a CSTE survey of state and territorial
public health epidemiologists and Centers for Disease Control and Prevention
bioterrorism grantees on syndromic surveillance systems.
“Thirty-five percent of the respondents said that the privacy rule had
caused major obstruction or delay in disease reporting and 25 percent said that
their disease reporters, health care providers, had a significant problem with
the privacy rules requirement that covered entities account for disclosures.
“Immunization is another concern. Often, school nurses need
immunization information to assess compliance with state laws requiring
immunization as a condition of enrollment in school. State laws vary on whether
a school is considered a public health authority and whether immunization
records may be shared without the authorization of a parent or guardian.
“In many states providers cannot disclose immunization information to
schools without receiving a HIPAA compliant authorization. If authorizations
cannot be obtained for any reason, some children receive duplicate
immunizations. If OCR indicated that states could designate school nurses as
public health officials, the information on immunization could be released
without need for an authorization, thereby benefiting the children.
“With regard to accounting for disclosures, the reporting of suspected
cases of abuse and neglect has been a particular concern for social service
agencies. In many states, agencies receive reports of suspected cases of child
abuse or neglect are prohibited from disclosing the report or the name of the
individual or entity filing it. The state prohibitions on disclosure, however,
apply only to the recipient of the report and do not extend to health care
providers, such as hospitals that file them.
“Under HIPAA, an abusing parent acting as the personal representative
of the minor child may obtain an accounting for disclosures and learn of the
report. This has the effect of discouraging the filing of reports of suspected
abuse and neglect. An exception to the accounting for disclosures requirement
for reports of suspected abuse and neglect would eliminate this problem.”
DR. LUMPKIN: Okay. Let’s split this up into three sections for discussion.
The first one is just the factual discussion of the fact that the major
obstacle of public health is in the privacy rule’s misinterpretation, whether
intentional or unintentional by potential reporters. Any comments on that first
section, first couple of paragraphs?
DR. STEINWACHS: In the last line of the first paragraph, I didn’t quite
understand what coordination — I understood limiting disclosures, but —
DR. LUMPKIN: Accordingly, the main public health issues involve covered
entities and misunderstand the privacy rule and limit their disclosures and the
coordination of public health provisions of the privacy rules with other
MS. GREENBERG: I think there needs to be a comma or something after
disclosures because I was having trouble with that.
MR. FANNING: I had to read it a few times. That is the correct thing. The
verb in front of coordination is involved, not limit. Possibly, the sentence
could be polished up a bit just to make that clearer. I don’t think there is
any point in drafting it out loud here, but the intention is clear and we will
make it easily readable.
DR. FITZMAURICE: With regard to that same sentence, what are the two
different public health provisions that need to be coordinated of the privacy
rule or with other provisions of the privacy rule. It reads to me like people
looking at one part of the privacy rule and we are looking at another part of
the privacy rule, being with public health and they are not coordinated. I
don’t know what that means.
MR. HOUSTON: I think the intent was to coordinate with other provisions of
the privacy rule, such as accounting of disclosures and things of that sort. I
believe that was the intent.
DR. FITZMAURICE: With you saying that, that makes it clear. Thank you.
But I have another question.
DR. LUMPKIN: Okay. Go ahead.
DR. FITZMAURICE: The next paragraph, the last sentence, “Of course, it
would be difficult to assess the motivation of a covered entity…,” I
would put a period right there. At the bottom of the next paragraph, the one
that starts “In general,” the sentence is “Of course, it would
be difficult to assess the motivation of a covered entity for being reluctant
to make reports to public health officials.” Because we might want to take
that sentence out. It would be impugning —
DR. LUMPKIN: It is sort of redundant to have the — okay. That is an
editorial kind of thing. The thought is is that — well, the first sentence
above it says at least one witness suggested, however, that some covered
entities were merely using the privacy rules as an excuse to avoid the burden
of public health reporting. Then it would read, of course, it would — actually
it is difficult to assess the motivation of covered entities period.
MR. HOUSTON: Do we take out the whole sentence or just the —
DR. LUMPKIN: Well, the problem is is that the disclaimer on the first one,
one witness alleged that. We think it is important to report it, but we are not
going to impugn covered entities. It was because, you know, there was a lawyer
in the meeting.
DR. COHN: I guess I am a little confused. I can’t exactly see what you are
doing here. You are leaving some of that last sentence?
DR. LUMPKIN: We are making a suggestion to the subcommittee on how to
DR. COHN: I guess another way of handling it would be getting rid of the
entire last sentence and changing “were” and then the sentence above
“or might be using.”
DR. LUMPKIN: Yes, softening that statement.
DR. COHN: I take a little bit of umbrage at this one, but it is —
MS. GREENBERG: Change this to some covered entities might be using the
privacy rule and then leave out the last sentence.
DR. HARDING: It is pretty hard to hear what you all are saying.
DR. LUMPKIN: No, because actually you are bringing it back tomorrow. So,
rather than trying to — we are not going to try to wordsmith it, but I think
as long as you understand what the issue was —
MR. FANNING: What is the choice now? What is the direction of the
committee? Is it to strike the last sentence or is it to strike all after
DR. LUMPKIN: It is to clean it up. We kind of agree with softening the
statement that is made.
MR. HOUSTON: Here is what I think is — the second to the last sentence
simply says at least one witness suggested, however, that some covered entities
might be merely using the privacy rule as an excuse to avoid the burden of
public health reporting period. And then strike the following sentence.
MS. HORLICK: I like that.
DR. LUMPKIN: Anything else on the reporting issue?
Let’s go on to immunizations. Michael. We are kind of walking through the
public section. Immunization.
DR. FITZMAURICE: Right above there, the one that says 35 percent of
respondents said that the privacy rule —
DR. LUMPKIN: Oh, sorry. Still on public health. Go ahead.
DR. FITZMAURICE: — caused major obstruction and delay in disease reporting
and 25 percent said that their disease reporters, health care providers had a
significant problem with the privacy rules requirement. Significant problem, we
don’t exactly say what the problem is. I would offer a suggestion that the
problem with the burden of the privacy rules requirement. Now, maybe it is more
than burden, but my sense was that they had a problem with the burden of the
requirement. Too much trouble. Therefore, you just don’t want to do it.
MS. GREENBERG: Is the burden created by the privacy rules requirement?
MS. HORLICK: Yes, I like that.
DR. LUMPKIN: Is that true, though?
MR. HOUSTON: There is a discussion later in the letter, which describes
accounting of disclosures. Okay.
DR. LUMPKIN: I just wanted to be careful because we are kind of editing the
summary of testimony and I just wanted to make sure that we were being —
MR. FANNING: I think we would have to go back to the testimony. Now, this
is sufficiently general that it doesn’t go into detail about what the exact
grievance is. Do you want to leave it that way or do we want extract from the
testimony that pins down the issue.
DR. FITZMAURICE: I think it is well-explained later on, but the question
here was we raised a problem and we don’t see what the problem is, by putting
“burden” in there —
MR. HOUSTON: Can I —
DR. FITZMAURICE: — communicate what the problem is.
MR. HOUSTON: Maybe if we defer just until the end of the letter and then go
back to this because there is a paragraph that describes in a little bit more
detail what the burden of the accounting disclosures is or we move it up,
either one of the two.
DR. LUMPKIN: But since it describes it as a burden, I think Michael’s point
is that we say — had a significant problem. I think what he is trying to say
is that the problem is burden, but we don’t say that. So, we kind of leave it
hanging. So, we could just add in and say that 25 percent of the disease
reporters have a problem with the burden of reporting, which I think is the
intent. Then we can leave the two paragraphs separate.
MR. HOUSTON: I am sorry. I thought we had already decided we were going to
put that burden language in.
DR. LUMPKIN: Never mind. We are all on the same page. It is page 2.
Anything else on the public health reporting part?
Okay. We are going to move on to immunization section, that paragraph.
DR. LENGERICH: A couple of things in the second last sentence where it
says, “If authorizations cannot be obtained for any reason, some children
receive duplicate immunizations.” Is there any estimate of what that some
is? And I suppose some already do even without this process and then is it more
than just duplicate? Are these inefficient, harmful, unnecessary?
DR. LUMPKIN: I think that unnecessary is certainly the case but I think
that probably that can be softened — may receive.
MR. FANNING: The summary of the meeting says, “Some children receive
duplicate immunizations —
MS. HORLICK: This is Gail. I think that it is accurate to say that some
children do. I don’t have any sense of the numbers on that. I mean, there are
clearly children that receive duplicate immunizations because they don’t have
access to the records, but I don’t know if you want to soften it anyway because
I don’t know how often that is happening, but it definitely is happening.
DR. LUMPKIN: Well, kids definitely receive multiple immunizations, but the
issue is as this sentence implies that because they don’t get the
MS. HORLICK: The provider can’t disclose the records. So, they give them
the shot again so they can go to school.
DR. LUMPKIN: Or the parent may not have the record. I mean, there is any
MS. HORLICK: Right. That is true.
DR. LUMPKIN: So I think that by softening it, we don’t imply that there is
a one-to-one correlation cause and effect, but it certainly contributes to the
problem that results in kids getting multiple duplication.
MS. HORLICK: Well, that is fine. I see that.
MR. FANNING: I am not sure how you would soften it. It is a fairly
straightforward sentence. Describe the factual event; namely, testimony to that
DR. LUMPKIN: But if authorizations cannot be obtained for any reason, some
children may receive duplicate immunizations. We are not actually summarizing
testimony in this line. We are actually summarizing the conclusions on the
MR. HOUSTON: We say testifiers indicated or —
MS. GREENBERG: No. They indicated they do receive them.
DR. LUMPKIN: Okay. So, you could say the testifiers stated that they
believed that — or whatever, stated, and that takes it off.
I have actually a little bit of concern about the construct and I was
trying to work through this as I listened to this issue. HIPAA provides —
actually I am thinking when it voids the preemption, floor preemption to state
law. As such, state law governs whether or not the restrictions on movement of
personal health information from a clinical setting to a school nurse. So,
HIPAA says — if HIPAA were to say that is okay and state law says
“no,” it is not, it doesn’t really matter because state law will then
So, if we are concerned that state law is prohibiting in some states
movement of immunization information to school nurses. Making a change at the
federal level will have no impact.
MR. ROTHSTEIN: John, this is Mark. I think the problem is that states don’t
realize what is going on necessarily and if the Secretary could indicate that
the states might want to take a look at their state law and see if it does
permit this disclosure, that then the state, if they chose to, could eliminate
DR. FITZMAURICE: The state could also say that within the privacy rule if
disclosure of information is required by law, by state law, the privacy rule
does not override that. So, you can disclose it. If the state were to say we
require by law that you immunize the students and so along with that, we
require by law that you look to see if they have had other immunizations. So,
that disclosure is required by the immunization law already. But if you don’t
state it specifically, it would have to take an OCR interpretation of that
communicated to the states, maybe a Q&A.
MS. HORLICK: I think some states are interpreting it that way but other
states are not.
DR. LUMPKIN: Well, you know, the difficulty I had is I spent three separate
years in Illinois with the legislature trying to change the law, allowing
immunization information to go from the public health system to the schools.
They clearly did not want that to happen. I mean, it wasn’t any ambiguity. They
didn’t misunderstand. They just didn’t want immunization information without
the parents consent going to the schools.
So, in a sense, you know, I totally agree with this principle and I agree
with the concept, but are we trying to make immunization — are we trying to
make state policy or are we trying to clarify federal policy? So, that is my
concern around that paragraph as much as I think it would be neat if that could
fix the problem.
DR. FITZMAURICE: I think you raise a good point. If it is a matter of
confusion, then I think what is here and my suggestion could clarify it, but if
it is not confusion, they simply don’t want that information moving. They want
to protect the privacy of the individual. A judge might rule that that was more
restrictive and, therefore, more protective of the individual’s personal health
information, even if it didn’t protect the person’s health.
DR. LUMPKIN: So, actually what we are looking for is clarification at the
federal level that HIPAA privacy does not — should not stand in the way of
immunization information being provided to schools.
MS. GREENBERG: It may currently, unless the school nurse is considered a
public health authority.
MS. HORLICK: Yes. The interpretation in a lot of the states is that because
— if it is not provided for treatment purposes, then they can’t disclose to
the school without an authorization.
DR. LUMPKIN: Okay. I think that — I am going to remove my objection and
let it go forward this way in that it may have a potential of doing some good
and for those states where they are just bone-headed about it, since I don’t
live there anymore, it is not going to do any harm.
DR. COHN: Actually, I am sort of buying in, I think, to some of your
comments, which is No. 1 that there needs to be probably some reflection that
the suggestion here is necessary but not sufficient, which is I think what you
were trying to say. I think that that is something that somebody needs to
realize. This won’t solve it in and of itself. I guess I am going back actually
to the earlier piece, which is — I mean, I think that the whole disclosure
piece is a barrier. I am trying in my own mind to think of the fact case where,
you know, patients don’t get immunizations without some parental involvement,
at least to my knowledge.
Now, the issue around immunizations is that people can’t attend school
unless they have appropriate immunizations. So, this creates a tremendous
bottleneck, but if you were a parent and having to deal with the patient anyway
and going and getting things — I guess I am trying to think of what the
barrier remains on that one. If you are already having to be involved and it is
your choice about whether to sign an authorization for disclosure versus
getting the kid a duplicate immunization, I am sort of — still, I guess, I am
thinking back on the testimony and trying to figure that one out.
One of the Johns, do you have a comment on that one? Am I missing the point
DR. LUMPKIN: I think your point is is that if the parent has a choice
between getting a second immunization and signing an authorization, most
parents would sign an authorization.
MR. HOUSTON: — is that it was very difficult in certain cases to have
parents be responsive in any way, shape or form, whether it be to sign an
authorization or getting a duplicate —
MR. ROTHSTEIN: Can’t hear, John.
MS. HORLICK: Yes, we can’t hear.
MR. HOUSTON: — and that there were other cases where if the records
weren’t available — I recollect at least in this last testimony we heard not
too long ago, that, you know, that the school nurse was indicating that parents
were driving hours in certain other cases either to get an authorization
signed, especially where they are out in a rural area, where a physician office
may not be close by. So, I think there were some concerns over all that just
operationally you could solve by simply making these
DR. COHN: I actually agree with what you just said. It is just that that
doesn’t say that here. What I was sort of thinking was is that the real issue
here is is that it is, you know, a tremendous bottleneck. It is a tremendous
hassle. Once again, my understanding — and once again, I only have to speak
for the state that I live in, but unless you have your immunizations, you
weren’t allowed in the school. So, it becomes sort of a major barrier there and
that that is really sort of the issue.
It isn’t that — I mean, I guess people could choose to get duplicate
immunizations, but, once again, as I said, you know, if you are a parent and
you are being told that you can’t go to school if you don’t get the
immunization or sign a disclosure —
MS. GREENBERG: The problem is I think it had to be a HIPAA —
MS. HORLICK: — that they couldn’t even do things by fax, but that was the
clarification issue, but I think the major issue, my understanding is that for
a lot of these parents in New Mexico particularly we heard these concerns.
These were just parents that couldn’t take off from work, couldn’t go to the
doctor, wasn’t sure who had their records and maybe the information was in the
doctor’s office. The parent was willing to sign whatever they could, but
because of all these logistical problems, the child was either getting
duplicate immunizations or not being able to start school.
MR. HOUSTON: I recollect, again, this later testimony, though, that there
were still problems in getting parents to even go and resign authorizations.
So, there is a variety of issues that —
DR. COHN: I withdraw that. We will deal with it in the subcommittee level.
MR. HUNGATE: It is a procedural question on my part. It has been a long
time since my kids got immunized. So, I am not current, but isn’t there a
consent form that the parent signs at the time of immunization? Isn’t there
some formalization of the authorization for immunization? And couldn’t the
disclosure be included as part of that so that it was done, the decision made
at the time of immunization?
DR. LUMPKIN: Well, it could be if the parents so chose to do that, but
giving authorization to release the data is not a co-condition for getting the
vaccine. So, parents may choose not to sign it. Providers may choose not to
have that form filled out.
This is tied in with — in fact, a number of states have immunization
registries, if not the vast majority of them because the CDC has been actively
funding these. Another complexity to this, which leads me to say that I think
this is fine, let’s just go with it, is the fact that states that operate
immunization registries could choose to provide that information to schools
because they are actually mixed entities. As such, the entity that collects the
immunization registry data is generally not a covered entity portion. So, the
information is no longer in the HIPAA domain.
MS. HORLICK: John, that is probably true for most every case, but there are
probably six or eight states where the entire state health department — New
Mexico, I believe, is one of them — is a covered entity.
DR. LUMPKIN: And that is the reason why I think we should go ahead with it.
Is there anybody here from New Mexico, who can stand up for their state? Jeff.
Okay. Jeff was silent.
So, I think we are kind of — okay. We see that there are limitations, but
we think this may be the best way to deal with it.
Anything else on immunizations?
Okay. The last issue is on abuse and neglect. Any comments on that? My only
concern on this one — and I just don’t know how to do it — maybe we should do
it in a summary — this one is a bright light issue to me and I suspect would
be to the Secretary. We just need to make sure that this issue doesn’t get lost
in this letter because this is one that probably will drive quicker action than
anything else we raise.
MR. HOUSTON: There is a bullet point in the summary, third bullet point
down in the summary. I think these were sort of taken in order of —
DR. HARDING: Please speak into the mike.
MR. HOUSTON: I am sorry. The third bullet in the summary really does
describe this and I think the bullets were sort of added sort of in the order
in which they were placed within the letter. But, I mean, we could expand or
potentially move that up. In fact, if you wanted to try to emphasize this in
the summary —
DR. LUMPKIN: Well, I suspect that our conversation here with our federal
partners, some of whom might be charged with answering the letter once we send
it will, I think, serve to highlight that point. So, I think we can leave it
the way it is. But I think it is important for us to note that this is an issue
— the issue of children being abused and our protection as a nation is a very
important issue, one that is considered very highly. So, this point here is not
the same, I think, level as all the other points that we raised, as serious as
Okay. Anything else on that?
MS. HORLICK: John, I just want to say that I need to sign off now, but I
will call in for the subcommittee meeting tomorrow.
DR. LUMPKIN: Thank you, Gail.
Okay. Let’s go on to research.
MR. HOUSTON: “Witnesses at the hearing provided frank testimony
describing the potential detrimental impact of the privacy rules research
provisions on research activities. The witnesses at the hearing overwhelmingly
supported the privacy rules intent of aligning its requirements with those of
the federal common rule for the protection of research subjects, 45 CFR Part 46
(common rule); thereby, promoting consistency and ease of compliance. In some
key instances, however, the privacy rule diverges from the common rule in ways
that cause either gaps in privacy protection or unnecessary obstacles to
“Additionally, due to the considerable confusion, compliance with the
privacy rule provisions on research would be helped by clarification of
expanded educational activity. An example of the inconsistencies between the
privacy rules research provision and the common rule relates to preparatory to
research activities. The privacy rule permits PHI to be reviewed by a
researcher for the purposes that are preparatory to research without
institutional review board, IRB, or privacy board approval and without a
“Preparatory to research includes such activities, hypothesis,
definition, protocol preparation and research recruitment. In fact, according
to the department’s August 2003 document, institutional review boards and the
HIPAA privacy rule, the privacy rule permits a researcher, who is a workforce
member of the covered entity, to contact potential research subjects for the
purposes of seeking an authorization as part of the covered entity’s health
“Even if such contact could be construed as coming within health care
operations, the interpretation permits recruitment of potential research
subjects (as an element of research) absent either a waiver from the IRB or
patient authorization and thereby violates a fundamental principle of research
ethics in the common rule.
“The privacy rule permits an authorization for the use and disclosure
of PHI in research to be combined with an informed consent document, although
many researchers prefer to use separate documents. According to the same August
2003 publication, the privacy rule does not require IRBs to review and/or
approve authorizations either as stand-alone documents or when combined with
informed consent documents.
“The permissive nature of this interpretation has created confusion
about whether IRBs have the authority or responsibility to review these HIPAA
authorizations. Because IRBs are charged with reviewing all written materials
provided to research subjects, as well as considering whether privacy and
confidentiality protections are adequate. A revised interpretation recognizing
the authority of IRBs to consider HIPAA authorizations would promote clarity
and help coordinate the privacy rule with the common rule.
“An area in which the diversions of the privacy rule and the common
rule results in burdens on researchers involves general research
authorizations, under the common rule, subjected to such limitations as an IRB
deems appropriate, a research subject may provide informed consent for future
unspecified research, including research using biological specimens.
“Under the current interpretation of HIPAA, however, an authorization
may not be for future unspecified research and a separate authorization must be
obtained for each trial or study. Unless the HIPAA interpretation is changed,
it will be exceedingly difficult to compile research repositories, including
the collection of biological specimens linked to medical records, which are
essential to many forms of research.
“A January 2004 document, “Research Repositories, Databases and
the HIPAA Privacy Rule,” indicates that a waiver of authorization could be
obtained from an IRB, but this additional step further complicates the process.
Several other areas related to research also need to be addressed. Genetic
researchers are concerned that any DNA sample, even if not linked with an
individual, might not be considered anonymous because analyzing the sample
could reveal the unique DNA identifiers of the individual.
“Clarification that unlinked DNA samples are not identifiable would
resolve the issue. Clarification also is needed on the applicability of HIPAA
to indirect participants (individuals who are not research subjects, but whose
PHI may be disclosed by research subjects) in multi-institutional studies. The
witnesses also identified some areas in need of additional outreach and
educational initiatives to counteract the reluctance or refusal of some smaller
institutions to participate in research because of misunderstanding HIPAA and
the standards for the de-identification of individually identifiable
DR. LUMPKIN: Okay. Any comments, questions on the research?
MR. HOUSTON: I do. A couple things that I have on the second paragraph
about halfway down, rather than “in fact,” I was thinking maybe the
word “specifically” might be a better choice of words. The second
paragraph under “Research,” about halfway down, there is a word
“in fact” or words “in fact, according,” the very top of
DR. LUMPKIN: Okay. That is an editorial change.
MR. HOUSTON: Then I did have some changes at the end or this paragraph —
MR. FANNING: Excuse me. Is there a choice to drop the term “in
DR. LUMPKIN: That is, in fact, the case, if I could be specific about that.
MR. HOUSTON: In its place say “specifically,” I would think.
DR. FITZMAURICE: In the copy that I have before me, it has already been
MR. HOUSTON: Then you probably printed out the version I sent out to the
committee that Mark said that we should not hold on — this makes things a
little easier then. I wasn’t sure whether this was going to be distributed
because I think Mark had indicated that he would prefer me to just bring the
comments up when we discussed it today.
If we are looking at that document, I think from my perspective, the last
couple of sentences of that same paragraph, I think, could use some further
DR. LUMPKIN: They don’t change the content, just the wording.
MR. HOUSTON: I believe, though, again — maybe I should read it and —
DR. LUMPKIN: Please.
MR. HOUSTON: These are the last two sentences of that same paragraph.
“Even if such contact could be construed as coming within health care
operations, interpretation permits…” — this is where I added
“…the direct recruitment of potential research subjects and element of
research by researchers who do not have a preexisting clinical relationship
with the patient.” Then I had put a period and then started with
“Absent either a waiver from the IRB or patient authorization, such
practices would violate fundamental principles or research ethics in the common
DR. LUMPKIN: I like your version.
MR. FANNING: Would you please mark it on your text and give it to me so I
can incorporate it into this.
DR. COHN: Do you have a copy of this?
DR. FITZMAURICE: One is a minor one that — at the top, where is says
“Preparatory research include…” that probably should be
“includes,” just with an s at the end of it.
More substantial, the last one — the one you just read, the even if
sentence, it says, “Such contact by the researcher to a potential
recruitee could be construed as coming within health care operations,
which…” — the interpretation is — “…it permits the direct
recruitment of potential research subjects, an element of research by
researchers, who did not have a preexisting clinical relationship with the
patient.” I think that is probably false because a researcher can’t do
that unless the researcher is working directly with the physician, for the
physician or if the researcher has a business associate contact for the purpose
of contacting the recruitees to see if they would be willing to participate.
It is the same as if the physician is doing it either because of the labor
relationship or because of the business associate contract. That is a direct
clinical relationship the physician has with a patient and has extended that,
have the researcher do something as a health care operation for him.
I disagree with describing it as researchers do not have a preexisting
clinical relationship. If the clerk calls me up and says have you had your
immunizations and I give that information, is the doctor’s clerk now having a
preexisting clinical relationship with me? I think it is all covered by being
under the rule of the physician.
MR. LOCALIO: This is Russ Localio. I have a problem I think with the way
you have revised the text; whereas, I didn’t have as much problem with the way
it was. Let’s assume that a physician has an ongoing clinical relationship,
clinical physician/patient relationship with a patient. The physician, the
treating physician then wants to recruit the patient for a research study. The
way I understand it, the physician would have to have IRB approval to do that
contact. It doesn’t matter whether the physician is the treating physician or
not the treating physician. That needs to be clear because I think your point
is here that the way the privacy rules might be interpreted is that they waive
the requirement for IRB approval of research.
MR. HOUSTON: No, what it does do is waives the requirement of IRB approval
for research recruitment activities, which are preparatory to research.
MR. LOCALIO: Fine. I think we are in agreement on that. I think the way you
revised it, it tries to make a distinction between recruitment by the treating
physician and recruitment by a researcher, who is not a treating physician and
I don’t think that is a distinction that should be made. Certainly, I don’t
think it is a distinction that is made by the common rule.
Research is research very clearly defined by the common rule and requires
IRB approval, even if it is the treating physician. So, in other words, the
surgeon who wants to do a follow-up of his patients and not simply call up his
patients and say, you know, come back, I want to do a research study on these
people. That treating surgeon would have to get IRB approval.
MR. HOUSTON: I believe that is correct.
DR. FITZMAURICE: I disagree with the interpretation. I think in order to
obtain the patient’s authorization, somebody has to contact the patient and who
better than the treating physician.
MR. LOCALIO: That is a separate issue. I think here you are just talking
about the IRB would have to approve anyone contacting a patient to recruit the
patient for a research study. The IRB would have to approve contacting a
patient, except —
DR. FITZMAURICE: You mean because of the common rule.
MR. LOCALIO: That is correct.
DR. FITZMAURICE: I don’t disagree with that. I am just talking about the
MR. LOCALIO: Now the privacy rule, let’s assume that that piece is
clarified. The privacy rule would not prevent — if the IRB approved it, would
not prevent a researcher, who was not taking care of the patient from
contacting their patient to recruit the patient for research.
DR. FITZMAURICE: One exception. If the patient does not have a business
associate contract with the physician or is not the treating physician him or
herself, then that researcher is prohibited by the privacy rule.
MR. ROTHSTEIN: — in front of me the interpretation from the document,
clinical research and the HIPAA privacy rule and it — let me read the entire
Q&A, if I may.
Question: If under the preparatory to research provisions, a researcher
identifies subjects that meet the study’s eligibility criteria, how can the
researcher contact the potential participant to obtain authorization after
identifying these individuals?
Answer: Under the preparatory to research provision, covered entities may
use and disclose to researchers PHI to aid in study recruitment. They may allow
a researcher to identify but not contact potential study participants. A
researcher may do so without authorization from the individual under the
following circumstances. There are two given.
First, if the researcher is a workforce member of a covered entity, the
researcher may contact the potential study participant as a part of the covered
entity’s health care operations for the purposes of seeking authorization.
Alternatively, the covered entity may contract with a researcher as a business
associate to assist in contacting individuals on behalf of the covered entity
to obtain their authorizations.
Second, if the covered entity obtains documentation that an IRB has
partially waived the authorization requirement to disclose PHI to a researcher
for recruitment purposes, the covered entity could disclose to the researcher
the PHI necessary for the researcher to contact the individual.
So, basically what the interpretation through this Q&A does is say that
you can solicit — that is, a researcher can solicit authorizations from
individuals before a protocol has been approved by the IRB and because
recruitment is considered to be part of research, what this would allow is an
element of research to go forward before the IRB has even seen the protocol.
That would violate the common rule. That is the point we are trying to make
MR. LOCALIO: Well, I agree with the point you are trying to make, Mark.
This is Russ Localio again. I agree with the point you are trying to make. It
is just that my understanding is the way the proposed language to amend this
was inconsistent with the point you are trying to make.
The point you are trying to make is IRB approval is Step 1 and then all the
rest of the stuff is Step 2. The point that I thought I heard was that there
was a distinction between recruitment done by the treating physician and
recruitment done by a researcher, who is not the treating physician. If Step 1
has been completed, the treating physician would have to complete Step 1 and
get IRB approval to recruit a patient for a research study that he or she was
MR. ROTHSTEIN: Yes, Russ, I agree with you. The original draft did not even
mention a treating physician. That was an amendment that John Houston is
MR. LOCALIO: And, again, I think the original — I agree with the original
language. I have trouble with the amendment because of what I think is a
distinction between treating physician and researcher, who is not a treating
physician. Both have to have IRB approval.
DR. LUMPKIN: I think that if the subcommittee could get the message that —
I think that we are probably comfortable with the language that is right here
and it would be easier not to change it at this point because I think the rest
of us are pretty much lost on this one.
Mark, I agree with you.
MR. ROTHSTEIN: Okay. So, we can go back to the original paragraph 2 under
DR. LUMPKIN: I think that might be the best.
MR. HOUSTON: There is one point, though, that I think, though, is important
then, based upon what Russ had stated, I think, is that there is still the
issue of IRB approval that it isn’t even discussed here, even in the original
DR. LUMPKIN: It is and I think that what is missing from this thing, from
this document, which I think may make it a lot easier, is under
“Research,” we should just say we think the common rules should
govern and that the first — that researchers need to comply with the common
rule and then look to see whether or not there are any additional requirements
that are placed on by HIPAA that doesn’t deal with the issue about the separate
authorizations. I think that one is an upcoming paragraph. But I think we
should probably make that statement somewhere early as sort of a governing
Then this just becomes explanatory rather than a statement of our beliefs.
MR. HOUSTON: When you say the common rule should govern, do you mean that
the privacy rule provisions should be eliminated or should — the common rule
should come first and then the privacy rule should then be applied?
DR. LUMPKIN: You should first apply — you should first comply with the
common rule. After you have assured compliance of that, you should then
determine whether or not you are also in compliance with HIPAA.
MR. BLAIR: New Mexico agrees with that.
DR. FITZMAURICE: Wouldn’t you say you have to comply with both of them and
where they disagree, you have to still comply with both of them at the — if a
common rule is more restrictive, then you go with the one that is more
restrictive. It is not a matter of which one comes first. You have to obey both
of them and where they do conflict, then you need to get a judgment.
DR. LUMPKIN: Yes, and our judgment is the common rule should govern. If
there is a conflict between — well, no, the more restrictive should govern.
DR. MAYS: We do not want them to actually bring up the issue that there is
a conflict between the two and that they should go back and try and ferret out
where there are conflicts because I don’t know that you always want the most
restrictive, but you have to really think about the most restrictive may end up
creating problems in terms of research or vice-versa may create problems in
terms of, you know, burden relative to kind of the clinical operations that go
on. I don’t know if you always want to say that.
I think it is better to send people off to try and determine when there is
conflict that the common rule is the preferred but that they should really go
through and try and see where there is conflict between the common rule and
DR. FITZMAURICE: Vickie makes a good point.
MR. ROTHSTEIN: That is why in the opening paragraph to the section on
research, we note both that where there is a distinction between the privacy
rule and the common rule, one of two things could happen. It could mean that it
is a burden on research or it could mean that there is a gap in privacy
protection and we have examples of both from the testimony.
DR. LUMPKIN: Okay.
MR. LOCALIO: Following up on that, yes, I think there were examples of both
from the testimony because I was at one of the sessions. So, therefore, I would
recommend deleting the word “potential” in the first paragraph in
“Research.” I heard testimony describing detrimental impact.
The degree of the detrimental impact is certainly subject to empirical
clarification, but there was definite testimony about the detrimental impact.
DR. LUMPKIN: Okay. Witnesses at the hearing provided frank testimony
describing the detrimental impact.
DR. VIGILANTE: Just to help operationalize this for me, the line that says
“Authorization may not be for future unspecified research and a separate
authorization must be obtained for each trial or study.” Does this mean
that, for instance, if you —
DR. LUMPKIN: Are you like in the fourth paragraph?
DR. VIGILANTE: I am in the fourth paragraph — no, third paragraph of page
DR. LUMPKIN: We haven’t gotten there yet.
DR. VIGILANTE: Oh, okay. We are going to try to march through so we can
sort of finish up on the first paragraph. Anything we drop potential, we will
come back to that.
MS. GREENBERG: We added potential. Why are we dropping it. I know where you
are, but why are you dropping it?
DR. LUMPKIN: Because Russell commented that they heard actual testimony of
real harm that occurred because of the gaps in privacy. So, it is no longer a
potential. The sentence says, “The witnesses at the hearing provided frank
testimony describing the detrimental impact of the privacy rule on research and
provisions on research activities.
MS. GREENBERG: That is different than on privacy. I think there were people
who felt that this privacy rule’s research provisions could have a detrimental
impact on research activities. Do you feel that you heard that they already had
had a detrimental impact?
MR. LOCALIO: Yes. Certainly, that is at least the way I interpreted it. I
will just give you an example. You have 20 IRBs you have to go through and 10
of them say “no,” because this is the way we interpret the privacy
rule and 10 of them say “yes,” because this is the way we interpret
the privacy rule.
That is a detrimental impact on research because you have a biased sample.
By the way that is a severe problem in research.
MR. FANNING: Is that the direction of the committee?
DR. LUMPKIN: Yes. We are now back to the second paragraph —
MR. FANNING: Could we just ask — Mr. Chairman, you said that you wanted a
general statement about the relationship between the two rules?
DR. LUMPKIN: I have been shown the lack of wisdom of my statement by my
MR. FANNING: Before we leave the first paragraph, we have here joining the
staff of the Privacy Subcommittee, Laura Cutcat(?) of NIH, who will suggest
some factual issues in the first paragraph.
MS. CUTCAT: Good morning. I also work with OCR in the development of all
the HIPAA research and educational materials that have been cited in this
letter. I would like to suggest that the federal common rule for protection of
research subjects be revised to reflect the accurate title of that subpart,
which is Federal Policy for the Protection of Human Subjects.
These policies are codified for HHS at 45 CFR Part 46, not for everyone who
has signed on to the common rule. So, we may also want to suggest adding
codified for HHS before 45 CFR Part 46, just to be clear on what we mean there.
DR. LUMPKIN: Mindful of the time, I am going to try to suggest a procedural
approach to getting through this document. For people who are on the Privacy
Subcommittee, which will be meeting tomorrow, if you have editorial changes,
please hold those until tomorrow. Let’s try to get to the substance of the
document as we move forward.
In the second paragraph we kind of had some changes that John suggested,
but we are not going to go with those. We are going to sort of stay with the
language as it is.
DR. FITZMAURICE: Except for specifically rather than in fact.
DR. LUMPKIN: The second one deals with the separate informed consent, I
think, if I am summarizing that correctly, issue. Do we have any comments?
DR. STEINWACHS: You are on the third paragraph?
DR. LUMPKIN: Now on the third paragraph.
DR. STEINWACHS: I found the first sentence in the third paragraph a little
bit confusing as stage setting for that paragraph. I was just going to suggest
that the committee might want to just eliminate that sentence and just lead
DR. LUMPKIN: I think it is a little bit confusing what the first sentence
does is to say, okay, this paragraph is going to be about burdens, burdens on
DR. STEINWACHS: I believe you.
DR. LUMPKIN: It may not say that clearly enough. So, if you have some
suggestions — word changes —
DR. STEINWACHS: I was hanging up on the use of separate documents versus
one document and it just seemed to me that it was —
DR. LUMPKIN: I don’t think the second sentence talks about the burdens on
DR. STEINWACHS: Are we on the same third paragraph?
DR. LUMPKIN: I am on the fourth paragraph. I really apologize. I am with
Kevin. We are on the fourth paragraph.
MS. GREENBERG: Whether IRBs have the authority
DR. LUMPKIN: So, you are suggesting that we just eliminate that first
DR. STEINWACHS: It seemed a little clearer to me that way, but that is —
you know, as an outsider reading it.
DR. LUMPKIN: Okay. Any problems with doing that?
MR. ROTHSTEIN: Well, this is Mark. I think we need to set up the issue and
that was the purpose of the sentence. Maybe we can make it clearer, but —
DR. LUMPKIN: Okay. Why don’t we leave that for the subcommittee saying that
we want to make that a little bit clearer. Okay.
MR. HOUSTON: Work on rewording the first sentence of this paragraph to make
it clearer or a better lead in to the rest of the paragraph.
DR. LUMPKIN: Okay. Anything else on the third paragraph?
DR. FITZMAURICE: There was an issue that came up and resulted in a letter
from the Secretary from OCR back to some group. The issue was that for FDA
clinical trials, IRBs were reviewing both the informed consent and the
authorization and it was taking up a lot of time and burdensome for research.
The result of the letter that went out said essentially, yes, you are charged
with reviewing informed consent, but you do not — there is nothing that
obliges you to review the authorization form. That was to reduce the burden on
IRBs and to reduce the burden on researchers. I think that they already — IRBs
can already consider whatever they want to consider before they give the
imprimatur because they already can. I am not sure I see what is required by
They have the authority to do anything they want. They are not required to
by the privacy rule, to look at authorizations. But if they think they need to
in order to protect the information, then they can.
MR. ROTHSTEIN: Well, the issue was that there is from the testimony great
uncertainty among IRBs about what jurisdiction is and we thought that including
this tier, that the relief we are seeking is only a clarification that IRBs
have this authority and I think this is something that the Secretary’s advisory
committee will be taking up as well.
So, this will be sort of the HIPAA side of the same issue to try to clarify
for IRBs what their jurisdiction is.
DR. FITZMAURICE: So, are you asking, Mark, that the department interpret
the privacy rule as requiring the IRBs to look at the authorizations or
allowing them to look at the authorizations?
DR. MAYS: Partly what I was going to say is that based on what you just
said, it might be helpful to actually have a statement in there that then the
IRBs would know that they are not required to do it. I think what it seems like
the committee is trying to do is to let IRBs know that based on the typical and
customary act of wanting to see everything that they can and that HIPAA doesn’t
prevent it, but I think it would be helpful in terms of the IRB burden for them
to know that they should not feel obliged to do it because of HIPAA.
MR. HOUSTON: I don’t think that was the point, though. I think the point
was that this paragraph was intended to provide the state with the fact that
the IRB should have the authority if they so choose to be able to. I mean, the
difference between being obliged, having the right and having the authority,
authority says that if somebody — a researcher is working within an
institution and the IRB of record decided that it has the authority to oversee
and review all of these materials, then that researcher would be required to go
through the IRB and have all those materials reviewed, rather than the thought
that the IRB may not have the mandate to be able to review those materials and,
therefore, the researcher could decide him or herself whether to have the IRB
review those materials.
DR. LUMPKIN: Let me just suggest that if you look at the recommendation
that generates from this paragraph, this paragraph really states the problem.
The recommendation on page 4 says, “HHS should clarify that IRBs have the
responsibility to approve stand-alone authorizations for research.
When you read the paragraph, it states the problem. The solution is not
saying that they have to. It says whether or not the issue that is raised in
the paragraph — it is that they are unclear about their authority.
DR. FITZMAURICE: I don’t see anything that gives them that responsibility.
They have the right —
DR. MAYS: Right. I was just going to say now it is totally different. Here
is the recommendation that has the responsibility. They have the right, but
HIPAA doesn’t make them have more responsibility.
PARTICIPANT: Maybe we should change “responsibility” to —
DR. MAYS: Yes, I think that would be good, yes.
MR. FANNING: This is — to the recommendation?
DR. FITZMAURICE: It looks like it is encouraging IRBs to look at all the
authorizations. That is burdensome.
MR. HOUSTON: I think Mark needs to weigh in on this one. Mark, could you —
Mark, was I correct in what I thought we were trying to do, which was to grant
the IRBs the authority to be able to have the oversight of all materials?
MR. ROTHSTEIN: We can’t as a subcommittee or even a full committee tell
IRBs what to do. The Secretary might. In trying to act on recommendations from
the Secretary’s advisory committee, as well as our committee, say that they
ought to do that or are required to do that.
At this point, I think the authority language is fine and maybe when the
Secretary’s advisory committee letter gets to the Secretary, the Secretary may
want to strengthen that and change it from authority to, in effect, require.
DR. LUMPKIN: Okay. Before we — I have to ask one question because as I
have been hearing the discussion, the role of the IRB is to assure that any
individual, who is a subject of research, is adequately protected and their
privacy is protected and that is the purpose of the consent that they give that
is reviewed by the IRB.
If that consent is adequate for IRB purposes, why are we concerned about
the HIPAA privacy authorization, because the researcher has to comply with the
consent that is given in both. As long as they have — if the one that is by
the IRB is more restrictive, then it doesn’t really matter what is in the HIPAA
If one is looser, because the IRB is comfortable with it, then the HIPAA
one is going to apply because it is not a research issue. It is a HIPAA
MR. ROTHSTEIN: The problem as described that I am hearing, John, is that
many institutions do not use a single form, although they may, for an
authorization and informed consent document. So, where there are two documents,
the IRB clearly could pass on the authorization, the stand-alone authorization.
Some IRBs feel that they don’t have the authority to rule on the authorization
because it comes under HIPAA and, therefore, they only approve one and the
researcher is left in limbo because the authorization that is a separate
document has not been approved by anybody and without an approved
authorization, we are told that they are having trouble getting access to PHI.
MR. FANNING: Mr. Chairman, I think what the recommendation here is, the
Secretary should make clear that just because the HIPAA rule does not require
IRB or privacy board review of an authorization form doesn’t mean that the IRBs
can’t address that issue. Is that not the point, Mark?
MR. ROTHSTEIN: Correct, yes.
DR. FITZMAURICE: And I would answer that, but they don’t have to.
PARTICIPANT: Right. That is why authority is a good word.
DR. LUMPKIN: But my question is —
DR. FITZMAURICE: They have criteria that they have to look at, applied to
the research. If to meet that criteria they want to look at the authorization,
they darn well can look at it. But the authorization is a legal document and if
it meets the legal requirements of the HIPAA regulations, then it is OCR’s
purview that either does or doesn’t.
DR. LUMPKIN: But let me understand because I am trying to figure out why we
are in this morass and it is taking up a lot of time. I understand the
conflict, but is it a HIPAA conflict or is it an IRB conflict? Because our
charge is not to advise the Secretary in relationship to IRBs, except as to the
extent that there is a conflict.
Now, are we uncomfortable from a HIPAA viewpoint that IRBs are uncertain
about their authority to review HIPAA authorization?
MR. HOUSTON: And I would say “yes,” because IRBs are specifically
mentioned in the privacy rule as having certain oversight rights or at least
can be involved in the oversight of the privacy rule with regards to research.
DR. FITZMAURICE: They can look at anything they want to to make sure that
the privacy of that information is being safeguarded.
MS. GREENBERG: Would it be correct to say HHS should clarify that nothing
in the privacy rule precludes IRBs from approving stand-alone authorizations
DR. FITZMAURICE: And I would add to that “or requires.” It
doesn’t require them.
MS. GREENBERG: Nothing in the privacy rule either precludes or requires to
approve stand-alone authorizations.
DR. FITZMAURICE: Mark, how do you feel about that?
MR. ROTHSTEIN: That is fine with me.
DR. LUMPKIN: Paragraph 4.
MR. HOUSTON: There is only one — this is a clarifying comment. The
sentence that starts, “Unless the HIPAA interpretations change, it will be
exceedingly difficult to compile research repositories…” I think there
should be a comma after repositories.
DR. LUMPKIN: We are going to leave commas for the committee, subcommittee.
I don’t think there is any objection to that.
DR. FITZMAURICE: On the sentence that says, “Under the current
interpretation of HIPAA, however, an authorization may not be for future
unspecified research and a separate authorization,” I would add after that
or IRB or privacy board approval, must be obtained through a trial or study.
DR. LUMPKIN: Read that again.
DR. FITZMAURICE: “Under the current interpretation of HIPAA, however,
and authorization may not be for future unspecified research and a separate
authorization must be obtained for each trial or study.” I would insert
and a separate authorization or IRB/privacy board approval must be obtained for
each trial or study. Because you don’t need authorization if you have IRB or
privacy board approval to get access to the information.
MR. ROTHSTEIN: Well, that is what the last sentence goes to clarify.
DR. FITZMAURICE: But you can clarify it right there in that middle sentence
so that people don’t get misled.
PARTICIPANT: That was one of my concerns.
DR. LUMPKIN: Okay. We are going to add that in.
DR. VIGILANTE: Just for clarification, is the — the next sentence that
says, “Unless HIPAA interpretation is changed, it will be exceedingly
difficult to compile research repositories, including the collection of
biological specimens linked to medical records, which are essential to many
forms of research.” Does this include, for example, linking de-identified
genomic data to de-identified tissue to de-identified clinical variables, such
as, you know, the nutrition of this particular subject, the co-morbidity,
survival, but linking all their elements, but having them all de-identified, is
that viewed as problematic?
MR. ROTHSTEIN: If they are all de-identified, that would not be a problem.
However, we also received testimony that we get into the next paragraph, that
some genetics researchers think that even sort of anonymous DNA samples are
not, quote, de-identified because it is the ultimate identifier and could be
sequenced to identify only an individual. We are trying to, you know, clarify
that in the next paragraph, but I think your concern is one that has been
expressed to us by researchers and in the testimony.
DR. VIGILANTE: But the position of — the current position, is there
clarity around that or not? In other words, if you de-identified the tissue, if
you have tissue samples linked to DNA, linked to clinical outcomes and they are
all linked to each other but they are all sufficiently de-identified. Is that
seen as problematic?
MR. ROTHSTEIN: I think the interpretation of the rule would say, no, that
is not a problem. Researchers, I think, are mistaken and that is why we call in
the next paragraph for clarification that de-identified genetic material is
not, quote, identifiable.
DR. LUMPKIN: Okay. I am going to do a little bit of a reality check here
because we still have a fair bit of the document to go through. What I am going
to suggest is that because the Privacy Subcommittee is meeting tomorrow
morning, there are no other conflicting subcommittee meetings and I wonder if
we would feel comfortable with having this discussion continue tomorrow morning
at 8:30 and then having those parties who are interested or have additional
comments to make about the document, bring those to the subcommittee meeting
tomorrow morning for — bringing the document back to us in the afternoon.
MR. BLAIR: So, we are not precluded from attending, but we are not required
DR. LUMPKIN: That would be correct.
DR. STEINDEL: John, the only problem that I see with that is that we have
arranged for some of the new members to go to Hyattsville for badge photographs
at that time, assuming that there would be no conflicts.
DR. LUMPKIN: Then I would ask those who have additional comments and
concerns, perhaps to share those with John, if they are going to be going out
and getting the — because we do have a full agenda for the two days and I
don’t think we can continue to delve into this. Obviously, there are some
issues that — and I don’t want to make a slight because I think that some of
the work, including the commas, are important clarifications and we don’t want
to because of the press of time to minimize that input.
I think having fuller time at the subcommittee will do that.
MR. HOUSTON: The only thing I would ask if they come to me, that probably
we should have John Fanning around also.
MS. GREENBERG: If you have a cell phone on your route out to Hyattsville,
you could call in to the meeting.
DR. LUMPKIN: No, this is like the good old days with the Privacy
Subcommittee. Only the older members remember those days, back before we had a
Okay. We are scheduled — we are going to take a ten minute break and then
we are going to come back with our special panel presentation.
DR. LUMPKIN: Let’s reconvene. There are a couple of lessons for those of
you who are new or relatively new to the committee. First is, as Kathy Coltin
can tell you that and Lisa Iezzoni, who were members of the committee, is that
if you are involved in a major project even though you cycle off the committee,
you may not be able to get away as quickly as you think.
As two of our next panelists will tell you that even though you are gone,
you are never really gone. We have a panel that is going to begin to address
the issue of the health statistics for the 21st Century, which all of you have
in front of you. It is a major report of our committee, which Dan Friedman took
the lead on and sort of hammered through, despite all sorts of major obstacles.
We appreciate that. So, we have a panel. I am going to ask the panel
members to introduce themselves. I don’t think I know the order. Dan is going
to go first and Carl is going to go last. So, by process of elimination,
Barbara Starfield is going to go in the middle. Just a brief introduction
again, although we had introductions in the beginning.
DR. FRIEDMAN: I am Dan Friedman —
DR. LUMPKIN: And as you know before, I always interrupted you and said can
you speak into the microphone for the people on the web.
DR. FRIEDMAN: I thought that I was.
I am Dan Friedman. I am a consultant in population health information based
in Massachusetts and a former member of the committee.
DR. STARFIELD: I am Barbara Starfield. I am on the faculty of the Johns
Hopkins School of Public Health and a former member of the committee.
DR. VOLPE: I am Carl Volpe with WellPoint and I have never been a member of
the committee. I am going to be talking about the private sector view.
DR. FRIEDMAN: John, how long would you like me to — about how long would
you like me to speak for?
DR. LUMPKIN: We are scheduled to go until 12:30. So, I think the betwixt
the three of you, if we could look at maybe 15 minutes apiece.
DR. FRIEDMAN: Before I get started, let me mention that — let me emphasize
that any of the obstacles or barriers that John alluded to in developing this
report in no way resulted from either other members of the committee or staff
to the committee because, first of all, I and the other people who worked on
the report really appreciated the collegiality and the support from both the
committee members and the staff, which was really just super, as well as from
NCHS, which was a partner in developing the report and the Data Council, which
is also a partner.
I am going to speak about three things. First of all, at Marjorie
Greenberg’s request, I am going to provide a brief summary of the 2002 NCVHS
report on shaping a health statistics vision for the 21st Century. Marjorie
asked me to do that. I will try to keep it very brief. Marjorie asked me to do
that because there has been the addition of several new members since Barbara
and I cycled off and since the committee was released.
Second, I am going to present a very brief overview, very brief, of state
health data systems within the context of the report.
Third, I am going to build on the report and to describe three very simple
tactics for addressing some of the issues around state health data systems.
Even though the discussion is going to focus on state health data systems, let
me add that some of these issues really could apply just as easily to national
and federal systems. You have the report in front of you. One of the things
that the report did, which was very brief and very simple, but we thought was
very important, was for the first time actually define health statistics in the
public health literature.
The report defines health statistics as numerical data that characterized
the health of a population and the influences and actions that affect the
health of a population. What is important here, the take-away point, as it
were, is that our definition of health statistics focuses on population health.
We also defined the health statistics enterprise as the infrastructure and
activities necessary to produce health statistics. In this country, that
includes public and private organizations and individuals at all geopolitical
levels that perform the processes of health statistics. Again in this country
in particular, it is a very decentralized enterprise and it includes
organizations that collect, analyze and disseminate data on the health of
populations and the factors that influence health.
The mission of the enterprise, as it is defined in the report, is to
efficiently provide timely, accurate and relevant information that can be used
to improve the nation’s health. Again, a focus on population health as the core
part of health statistics.
One important part of the vision that we laid out in the report is the need
for an overarching conceptual framework for health statistics, an overarching
population health conceptual framework. In the report, we emphasized that that
conceptual framework is important because it helps to focus the enterprise on
needed data and it helps to guide the enterprise in identifying new and
emergent issues. It focuses on health, the population and the community, rather
than on solely much more proximate influences on the population health.
It emphasizes the distribution and level of health rather than just average
health, delineates major influences on health and helps to define a research
agenda for improving the population’s health. In the report,we lay out a
conceptual framework schematic. Barbara and I happen to like this schematic. At
the same time, I think what is — I don’t want to speak for Barbara anymore —
I happen to like this schematic. Having said that, there are other schematics
that perform exactly the same functions and the functions are basically laying
out in an easily accessible fashion what are hypothesized influences on the
health of the population. Here, we include a variety of community attributes,
contextual influences, including the natural environment, the cultural
environment and the political context.
The report includes a more detailed version of this. The report also
includes ten guiding principles for the health statistics enterprise. I am
going to run through these extremely rapidly with almost no commentary.
First, enterprise-wide planning and coordination. Second, broad
collaboration among data users, producers and suppliers at all geopolitical
Third, rigorous policies and procedures for projecting privacy,
confidentiality and security. Fourth, flexibility to identify and address
emergent health issues.
Fifth, use of data standards. Sixth, sufficient detail at different levels
of aggregation. Seventh, integrated streamlined data collection for multiple
Eighth, timely production of valid and reliable health statistics. Ninth,
appropriate access to and ease of use of health statistics.
Finally, continuous evaluation of the data.
Now, let me spend just a few minutes on state data systems. What I am
presenting very quickly are an overview of archetypical state data collection
systems. This is not an average picture of all state data systems. It is not a
modal picture but it is based upon my own personal knowledge. It is based upon
information that I collected from several states and I am going to — very,
very quickly — I am not going to read everyone of these, but I am going to
quickly run through it.
What is important to notice here, it starts with claims and billing, client
case management, AIDS treatment, WIC within claims billing, from AIDS treatment
to WIC within client management to each of these bullet points in most states
are separate and discrete data systems, not in all states, but in most states.
Each bullet point is a separate industry, discrete data system.
Now we can move on to your reports and then to encounter data. Again, each
of those sub-bullet points are separate and discrete data systems.
Environmental health data, asthma to water quality. Facilities surveys and
reports, each of those bullet points for data systems. Licensure, health
facilities and services. Licensure, health professionals, everything from
acupuncturists — I left off here, tatoo artists, but in those states where
they are licensed, they are also usually a separate data system for tatoo
artists, veterinarians, podiatrists, et cetera.
Occupational health, population-based surveys. Again, each of these — the
behavior risk factor surveillance system, community health surveys, HIV family
of seroprevalence surveys, PRAMS, pregnancy risk assessment monitoring systems,
youth risk behavior surveys, discrete data systems. Professional association
memberships, registries, reportable diseases and conditions.
Now, what we have in what I am positing as the archetypical state are
roughly 152 discrete state health data systems. One state I recently visited
had 127. Another state that has a fine meta-data site up on the web, states
which shall remain nameless, but the former state health officer chairs this
committee, had 159 data sets listed.
Now, there are — these roughly 150 data sets do have some archetypical
issues around them, some archetypical questions that we can raise that cut
across the states and then cut across the data sets. The first question we can
raise is is there a unifying conceptual framework. A second question we can
raise is is there interoperability among the data sets and a third question we
can raise, is there communication among the data sets, among the data
providers, data collectors and data analysts.
What I am going to suggest here is that one thing that the committee can do
is try to identify relatively simple techniques for overcoming some of these
issues, relatively simple techniques. What I am going to suggest is three
techniques as examples.
The first tactic, first technique is conceptual and it consists of learning
what data we have and learning what data we don’t have. In fact, when you have
got 150 discrete data sets, you really don’t, my feeling is, have a
particularly good sense of really what you have and what it adds up to.
In the interest of time, I am going to skip over here and say that one way
of learning what we have is to focus on models of population health and to
essentially use a schematic model of population health as a schematic, as a
template for evaluating data holdings, literally for plugging in variables from
data set to data set into the model and saying what information do we have
about the influences on population health, what information don’t we have,
essentially using population health models as an organizing scheme for learning
what we have and trying to be more reflective about what we have.
This is an approach that we really haven’t done in this country. The
Canadian Institutes for Health Research in conjunction with the Canadian
Institute for Health Information in Canada is in the early stages of conducting
a project somewhat related to this to basically develop what will probably be a
meta-data site for researchers.
A second tactic is methodological and, again, this is what I would suggest
is a very simple tactic. The 152 data sets for the most part do not have common
identifiers. They don’t have common variable definitions. This is despite
HIPAA. They do not often have common content on their identifiers. Certainly
they don’t have common codes and common transaction standards.
Now, what I am suggesting is not the end — I am going to emphasize this
particularly for Jim here — what I am suggesting is not a national health
identifier. I am not even supposed to mention those words here. What I am
suggesting is that archetypical state data sets can start using just a simple
set of common identifiers, commonly defined with common content, common codes
and common transaction standards and just using common identifiers. That
relatively simple step would go a very long way toward facilitating linkage and
facilitating intersectoral research.
The third tactic is a procedural one and it is somewhat more speculative.
The 152 archetypical data sets has different data set holders, not necessarily
152, but certainly more than one or two different set holders. Different data
set release protocols, different data set mandates, different data set
governing legislation and different data set programmers, analysts and
Now even though these 152 data sets are held in all likelihood in most
states by the state health department, at the same time, the health statistics
enterprise, as I said, consists of a great many different organizations. Now,
we can think of the governance continuum for producing data, health data for
policy purposes, as basically being on a continuum, ranging from governmental
organizations, like NCHS, the State Health Data Centers, AHRQ, Australian
Institute for Health and Welfare and somewhere loosely speaking in the middle,
quasi-governmental organizations, like the United Kingdom Public Health
Observatories, Canadians for Health Information and non-governmental
organizations, like the Manitoba Center for health Policy.
Now, clearly there is no single correct organizational model for turning
data into information for policy and the choice of organization, the models,
obviously, needs to be determined by the local and the national climate. Having
said that, certainly at the state level one thing I think that we need to
really start thinking about are different kinds of organizational frameworks
and one of the things that I would suggest we start thinking of are virtual
organizational networks as a new point on the organizational governance
Ongoing, highly structured collaboration among participants, organizations
and individuals within a clear legal or contractual framework, virtual
organizational networks that would build on the strength and capability of each
participant organization that would distribute some of the functions and
responsibilities of the health statistics enterprise to where they are most
appropriate that would maximize collaboration and intersectoral approaches.
Now, one of the requirements for a virtual organizational network is to
have a real nerve center, a strong organization or body in the middle, an
entity in the middle that would help to organize the enterprise for assessing
population health information needs, establish priorities, convening
participants, coordinating and planning of participants and facilitating data
sharing and linkage.
In the interest of time, I am going to close here and questions at the end,
DR. STARFIELD: While we are getting up the slides, I just sort of want to
get a sense of the table. I know some of you, but I am not sure I know how you
are going to answer the question that I ask. I am going to ask you to try to
identify your thinking, whether you regard yourself as basically clinically
trained, if you sort of think about individuals, whether you are population
trained, that is, you basically don’t think of individuals. You think of
populations or whether you think there is no useful distinction between
clinical thinking and population thinking because after all populations are
simply aggregates of individuals.
I was trained clinically. Okay. I originally was trained to think about
individuals. I have shifted over the years, but anyway, what about yourselves.
How many of you just tend to think clinically, tend to think of individuals?
How many of you tend to think — how many of you think of yourselves as
population thinkers? And how many of you think there is no useful distinction
between the two?
All right. Let’s talk about population health and its implications for
health statistics and we will see where you fit.
Challenges of 21st Century health systems — by the way, how many of you
have this presentation in front of you? That is good. I have a few corrections
to make. I would hope, please, that you make those corrections when I point
them out to you.
Challenges of 21st Century health systems: One, the poor and worsening
position of the United States population health indicators, 15 of the most
industrialized countries. For about seven or eight of the major health
indicators, the United States is last in rank on all of them. That includes
infant mortality, probability of death at under age 65, at under age 5,
probability of death at ages 15 to 60; a healthy life at age 60, years of
healthy life lost, percent of total life expectancy lost and several others.
The United States is last of the industrialized countries.
I might point out a couple of things, that the position of the elderly in
the United States has fallen in rank. We were among the top two ten years ago
and we are no longer among the top two. We are worse off than that.
The third thing I might point out that was surprising to me until I thought
about it and that is the percent of the population over 60 has actually fallen
in the last ten years in the United States. It is not true in other countries,
but in the United States it is true.
Okay. So, the poor and worsening position of the United States on
population health indicators, I have some thoughts about why that is the case,
but that is not the topic of my talk. So, I will save that for some other time.
The equality of health services and there are several things that might be
said about poor quality of health services. One is responsiveness to needs. Now
I want to distinguish responsiveness to needs to responsiveness to consumer
demands because they aren’t the same and I am talking about responsiveness to
I might point out that when evidence started to come through about the
dangers of unopposed estrogen and even with the hormone replacement therapy
with combined therapy, that those who were taking those medications did not
agree with recommendations to stop taking them. They did not believe the
evidence. So, there was continuing demand from postmenopausal women for these
things and the appropriate course of action was not, of course, to go with
demand, but, in fact, go with needs as we knew from the evidence.
Comorbidity, we have always had comorbidity. It is increasing or at least
the recognition of it is increasing and that makes us think twice about
disease-oriented guidelines. If we are interested in population health and
population outcome, we probably want to go well beyond disease-oriented
The third one is coordination of care and that has always been a challenge
for us, but it is increasing now, particularly with the increasing use of
specialists, largely by virtue of consumer demand, rather than consumer need in
my view. People are going elsewhere more often and there is increasing need for
Overuse. Overuse is something we probably didn’t recognize before ten years
ago. It is a major problem. Thirty-five percent of the increased costs in the
United States health care system are a result of overuse and unnecessary use of
services. What I would like you to add in is safety. It is related to overuse.
But I would like you to add in there safety.
Now, we didn’t really have much population-based data on safety. You could
read it in the newspapers, but you couldn’t read it in the medical journals
before five or six years ago.
Best estimates is that adverse effects and errors account for a third
leading cause of death. There is a big difference between the second leading
cause of death, which is from cancer, and the third leading cause of death,
conventional third leading cause, which is stroke. Estimates of deaths as a
result of adverse effects and errors is closer to the rate of cancer deaths
than it is to stroke deaths.
Okay. That is the quality of health services. Now there are a couple of
challenges in quality of health systems and one is that population outcomes. We
need to think about population outcomes, not aggregations of individual
outcomes and the need to think about disparities. We didn’t do that before five
years ago, think about disparities.
I want to particularly distinguish quality of health services from quality
of health systems. We focus in this country on quality of health services.
Virtually all of our quality assessment systems are on health services and we
focus almost not at all on quality of health systems.
I was interested to read an article by Revo(?) in a recent issue of Family
Medicine, a supplement, and where he said, “Practicing effectively in
today’s health system, teaching systems-based care in medical school and in
residency-based teaching…” We have actually never done that before. It
was in the Journal of Family Medicine. The family physicians seem to much more
— we need to think population than other clinical specialties.
Is population health the sum of individual health than I already told you.
What he thought about that, I think the way he thinks. Just think about it. Is
the mortality rate, the number of deaths over the population? Shake your head.
Is it? The number of deaths divided by the population, is that the mortality
rate? Yes, no or maybe. It is. Well, it isn’t. I mean, technically it is, but
nobody ever uses mortality. That is over population. At least age adjusted.
All right. Now, what about births less than 2,500 grams over the number of
births. Is that the low birth weight ratio or rate? Most people would say
“yes.” And I guess it is, but what does it mean? What does it mean
when the low birth weight rate for the African American population is twice
that of the — I mean, would you make policy based on the low birth weight
ratio as a number of low birth weights? Not if you are interested in
disparities, you wouldn’t. It is not meaningful.
What about the number of diabetics or the number of hypertensives in the
population, divided by the population, is that the illness rate? No, it is not
the illness rate. Okay. It is diabetes over — but that is not the health of
So, clearly population health is not the sum of individual health. We have
known that like — you know, we have known about mortality rates for a couple
hundred years. Right, Ed? I mean, you have to age adjust it. You want
information to be useful, right? Yes, you want it to be useful.
We haven’t known about low birth weights for that long, but probably at
least 75 years we have known that you have to think about low birth weight in
the context of different population groups. But we still don’t think about
health except in terms of diseases. More about that later.
The distinction between a focus on individuals and populations or
subpopulations is part of the distinctions among the branches of medicine. Now,
I apologize to those who saw these slides the last time I was here, but I will
go through them quickly.
Clinical medicine asks what disease might this patient have and how should
it be managed. Clinical epidemiology asks what is the relative likelihood that
this patient has or is at risk for this disease and what is the evidence to
support its management? Now, this is relative risk oriented thinking, right,
for both of these. Relative risk oriented. Given a constellation of risk
factors, what is the likelihood that something will follow from that? Okay.
That is relative risk.
Social medicine, why does this patient have this disease at this particular
time and how might this affect management, taking into account the
constellation of a variety of factors that influence the occurrence and
severity of illness, the genesis of illness.
Community medicine, is this disease important? If so, how important is it,
to whom and what is the overall benefit of management to the community. For the
first time now, you are beginning to see the notion of attributable risk,
community attributable risk, not relative risk, that is making the decision.
How important is this in the population?
You know, the risk factor may be very important for a particular disease,
but in the whole scheme of things, it is not that important when you talk about
the relative frequency of different diseases in the population.
Finally, public health, what characteristics are most salient in improving
overall health and the distribution of health in populations and what does
evidence suggest should be priorities for intervention. Attributable risk, both
in the terms of genesis of illness and in the terms of the effectiveness and
impact of interventions.
So, here, you tend to see now why thinking about health of populations
disease by disease might not be so important. It is the relative balance and
frequency of the diseases and the impact of the diseases in populations that is
All right. Now, clinical epidemiologically and social views towards health,
this is the diagram that goes along with that. I have a different diagram than
Dan presented, but we were really motivated and coming up with a diagram that
is in the report by simplicity and so we sacrificed a few things in developing
that diagram. It essentially says what this diagram has, but it says it in a
little more simple way. Let’s start over here on the right.
You have health. We have health over here and notice that I have divided
health in two. One of them is average health, which is the kind of thing we
usually think about when we think about population health and the other one is
distribution of health in the population. Clinically, we think this way. We
think about genetic and biological characteristics and we think about
sociodemographic characteristics, but from an individual point of view, we are
also interested in developmental health disadvantage and those of us, who are
pediatricians, really put a lot of focus on this because what happens earlier
in life certainly happens in terms of manifestations, certainly reflects in
Then if we are interested in social medicine, we are interested in the
impact of these things on health and distribution of health, physiologic state,
material resources, social resources, behaviors, panic stress and health
services received. Most of social medicine focuses on this. Almost all the
literature on social medicine, on social epidemiology focuses on the
relationship between social resources and resulting health and distribution of
health, in particular, in the last decade.
Now, if we are community oriented, we are thinking at the community level,
occupational and environmental exposures, material resources, social resources,
behavioral, cultural characteristics, psychosocial characteristics and health
system characteristics, not health services, but health system characteristics,
which all influence things at the individual level. Now, what this diagram that
the diagram in the report doesn’t have is it has got the interactions, all of
these have interactions. They work both ways. That means that any of these,
disadvantage in an of these things makes for greater vulnerability in other
things because there are interactions, multiple vulnerabilities. Very important
in the context of what we will deal with later is comorbidity.
Then some people will think of political and policy contexts as influencing
the community level characteristics. That is an individual level diagram or at
most a social medicine and a community medicine diagram.
Now, the community medicine and public health use towards health looks
pretty much the same, but it focuses much more heavily on the political and
policy context, as influencing these community level contexts. I have left out,
mainly because I couldn’t fit them in, all the individual level
characteristics, as in the last slide. And you have health care again as
average health and distribution of health, but you are not only interested in
this, you are interested in equity and — I am sorry — correction, please.
Let me go back to the last one. This division of health is occurrence of
illness and severity of illness. Different things influence occurrence of
illness and severity of illness. That is why in describing health, I am
thinking that we need to think about occurrence of illness, incidence and
prevalence and about severity of illness, disability and death rates.
Now when we get to the population level, we are interested in this average
health in terms of occurrence and severity, but we are also interested in
distribution of health in populations. Okay. And no longer are we thinking now
of individual health disadvantage, but we are thinking of historical health
There was a very interesting article in the British medical journal a
couple of years ago that showed that the most important thing that influenced
the health of different communities in London was what they looked like a
hundred years ago, controlling for a whole lot of other characteristics, what
the communities looked like a hundred years ago, more described the health of
those communities than any other characteristics and you have got the community
characteristics, the policy characteristics and the political context.
So, health system characteristics here, again, is key in terms of influence
on distribution of health and on average health. Now, what are the implications
for data systems? First of all, is linking individual or aggregated individual
data with a contextual or ecological data, what is the difference between
aggregated data and ecological data?
The contextual data at a community level really means an aggregation of
individual level data. For example, average income of the population is a
contextual level variable because it comes from adding up individual incomes
and dividing by the population, but pollution level is not an aggregated
measure. It is a true ecological characteristic. So, we have got two kinds of
things in the community level to take care of things to deal with, things that
are aggregated from individuals and things that are characteristic of the
community and the new literature on multi-level analyses doesn’t really
adequately distinguish the two, but they are very important because they have
got a lot of policy implications.
It is much more technically feasible to deal with population level in a
community than it is to deal with individual incomes and changing individual
incomes. So, there from a policy point of view, it is important to make that
Okay. Linking individual aggregated data with contextual or ecological
data, so we have to link the clinical information with the systems information.
Implications for information systems at the area level. We need characteristics
of areas in which people live and work because we are now beginning to get lots
of studies that show the impact of those community level characteristics on
individual health and distribution of health.
Social and political or power characteristics of people in an area, the
whole concept of what do we call it, social — social capital, yes, is a very,
very fuzzy capital. Some people use it as an aggregation of individual social
MR. BLAIR: Just a clarification because I am struggling on one piece and I
am not sure that I really understood what you meant when you said was it the
community a hundred years ago, individuals a hundred years — help me
understand what that means and —
DR. STARFIELD: Jeff, in that study in London, what they did was they
described the health of the community in 1890 and 1990 and then they looked at
the characteristics of those communities and they found that what determined
the constancy of health was, in fact, the community itself, not the
characteristics of the individuals in the community. Okay?
MR. BLAIR: Thank you.
DR. STARFIELD: So, implications for data information systems, the
characteristics of areas, the distribution of social and political power in
those areas, which is much broader than what Putnam called social capital,
which talked about individuals participating on bowling leagues and the
characteristics of the health system.
Implications for data and information systems from the point of view of
health services, I want to take a few minutes on this. Now, this is health
services, not health systems. The first thing I want to mention here is
problems, patients’ problems.
I think that ultimately if we were really interested in improving health,
what we would do is to deal with patients’ problems, not deal with diseases,
deal with patients’ problems. You want people to function better in the
community. I am going to give you a definition of health in a minute. What is
the purpose of health? Not to do away with diseases. It is doing away with the
problems that interfere with people’s lives.
So, we have got to be more problem oriented. I know you have discussed at
some length in this committee the issue of problem coding and you have
discussed the ICPC and I am very pleased, John, that you wrote a letter on
behalf of the committee to the Secretary about the ICPC. I think if we are
interested in improving health, we need to think about problems, rather than
diagnoses. Of course, they are related. But they are not always related. So, we
really have to think about doing away with people’s problems, things that
interfere with their life.
For example, I am going to use the data that Craig Gayon(?) and also that
Hank Inga(?) had been working with. Shortness of breath is a risk factor for
the diagnosis of the COPD. Any clinician knows that. But for patients, who
eventually have COPD, it is an infrequent presentation. Now a clinician thinks
of shortness of breath related to COPD, but, in fact, it is a relatively
infrequent presentation of COPD; the difference between relative risk and
The second thing is diagnoses. We make a very big thing about diagnoses,
not only in this country but pretty much everywhere in the world, which I think
is largely a function of increasing specialization, physicians, clinicians, who
are interested in specific diseases or organ systems, but we have lots and lots
of comorbidity. I mean, you know, among the elderly, the rate of comorbidity is
something like 70 percent have three or more conditions at one time.
I mentioned before the irrelevance of most clinical guidelines that deal
DR. STEINDEL: Barbara, could I get a clarification? When you say
“problem,” do you mean the patient’s perception of the problem or the
physician’s interpretation of —
DR. STARFIELD: I mean the patient’s perception of the problem. We have
already shown with studies — and Don may have mentioned these because he was a
co-author on a couple — that the patient is more likely to get better if the
patient and the physician agree on what the patient’s problem is.
Management, is it going to be disease or morbidity oriented? Right now, it
is disease oriented. And reassessment, is that going to be disease or problem
oriented? Are we going to look for changes in manifestations of diseases? Or
are we going to look for changes in the way people respond to their diseases
and deal with their diseases?
Okay. Now, implications for data information systems, disease morbidity or
health oriented, comorbidity and the concept of health. Now, comorbidity,
diseases, risk factors and influences are not independent of one another, as I
showed you from that previous diagram. Data systems must allow for coordination
among different providers and types of providers. These are Dutch data, right?
Do you know these from Vandanocka(?). Maryanne has done a terrific job with
What you have here along the horizontal access is the number of diseases,
0, 1, 2, 3, 4 or more than 4, in a population. What you have on the vertical
access is the ratio of the observed — now expected comes from the frequency of
different diseases in a population. Observed is what you observe in terms of
number of diseases in the population.
The line, the black line, shows all ages combined and you can see that a
much greater proportion of the population in general has four or more — more
than four diseases than you would expect by chance, distribution or diseases in
the population. A smaller proportion of the population has fewer numbers of
diseases, 1, 2 or 3 diseases, than you would expect if diseases were randomly
distributed in the population and a slightly greater proportion, in fact, have
What Maryanne has done here is to divide it by age; 0 to 19, 20 to 39, 40
to 59, 60 to 79 and 80 plus. Now, notice here even in the 80 plus group, a
greater proportion of people have no diseases than you would expect by random
distribution in the population. Over here, only a slightly greater proportion
of 80 or older people have more than four diseases than you would expect by
random distribution of the population. Disease is much more common in the
elderly. So, you expect more and, in fact, you observe only slightly more.
Look at children. Look at the proportion of children that have more than
four diseases than you would expect by random distribution in the population.
That is pretty striking. We tend to think of kids as healthy, but, surely,
there is a proportion, a substantial proportion of children that have more than
four diseases than you would expect if diseases were random in the population.
So, comorbidity is a real challenge to a health system that we aren’t even
thinking about meeting. This is what we call the ACG system and it shows that
there is a way to take diseases that people have and fit individuals into one
lock based upon the constellation of diseases that they have. We can do it if
we want to do it.
What is health? Health is the extent to which an individual or group is
able on the one hand to realize aspirations and satisfy needs and on the other
hand to cope with interpersonal social, biological and physical environments.
It is a resource for everyday living, everyday life, not the object of living.
It is a positive concept embracing social and personal resources , as well as
physical and psychological capacities. I wish I had made up this definition,
but I didn’t.
It is not the WHO definition, but it is WHO URO(?) definition. They came up
with this in Copenhagen and it was then subsequently buried by the WHO Geneva,
but it resurfaced again in Ottawa in 1986 and I think now is certainly pretty
well accepted in a lot of circles as a definition of health.
How is population health measured is a question. Infant mortality rates
versus neonatal and post neonatal mortality rates. A big difference between
neonatal mortality rate and post neonatal mortality rate, not only in terms of
when they occur, but in terms of what their causes are. Influences on neonatal
mortality are very different than influences on post neonatal mortality, really
different. Post neonatal mortality is quite responsive to health services, but
neonatal is only responsive to neonatal intensive care services. Big difference
in terms of making policy, in terms of what you want to influence.
Yet, it is not common for us to distinguish the two. No international
statistics really look at the neonatal separately from post neonatal. Mortality
rates versus age adjusted mortality rates, we have talked about that, car
specific mortality, life expectancy, years of potential life lost. The GAO gave
this one the highest choice, 17 different indicators. They thought this was
probably the best one to characterize population health, disease occurrence and
severity. We have talked about that. All diseases, target disease is very
popular in the United States and worldwide.
Self-reported health and — there are others, but these are the most common
ones. Which of these represent population health?
Alternatives to categorizing population health.
PARTICIPANT: You said you don’t know, is that what you said?
DR. STARFIELD: I said I don’t know but I am not willing to make the
But I actually have my preference. I have shown you a scheme for
characterizing morbidity case mix. That is one possibility if you are
interested in diseases. The DALEYs are a good approach, I think, the Disability
Adjusted Life Expectancy Years, but as used by WHO, they basically boil down to
disability from specific diseases. I think as used, it has a lot of
The one I want to mention that is the profiles of health, derived from
combinations of separate domains and this is something we have worked with in
children and I sort of think it is neat. Criteria for defining profile types.
We have divided health into four different domains in children. One of them —
let’s see if I can remember that
— we have actually divided them into six domains.
One of them is disease rates. One of them is achievement of social
expectations, given stage of development. Another one is comfort or discomfort,
which is symptomatology and signs. Another one is perceived health,
satisfaction with health and self concept. Another one is vulnerability, what
you have that suggests you have compromises to your health. And another one is
resilience; that is, what do you have or what circumstances are you in that
suggest you are more resilient to insults to health.
We have taken these different domains for children and we have divided
populations of children into one of these groups. Excellent health is defined
as excellent health on three or four domains with no domains of poor health. We
have trichotomized each of these domains, lowest, middle and high and then fit
children into these boxes. Good health is at least average health on all
domains with excellent health on no more than two domains, et cetera. Worst
health is poor health on three or four domains. We found the validity, all
kinds of validity, from these characterizations of health and we also find
predicted validity in terms of from one year to the next.
So, I think this approach is I think rather promising if you are looking
for a way to characterize overall health. Now, dealing with disparities, we
don’t deal well with disparities in this country. Other countries do a lot
better. One of the reasons we don’t deal well with disparities is because, you
know, the — what is it called
— health of the population 2010. Is that the title? Healthy People 2010 —
talks a lot about disparities, but it doesn’t give you any idea of what you
should do about them because it has no feeling for what it is that is causing
the disparities, given that diagram of influence. There is very little in
So, whereas other countries like The Netherlands and like U.K. are thinking
of disparities in terms of what is the best way to deal with them, that is much
more advanced than we are in this country. But one of the things at least we
could do is do stratified analyses rather than statistical adjustment. What we
almost always do is we put race, ethnicity, income group into a regression and
we say controlling for these things, this is what we find in terms of health;
whereas, what we really should be doing is look at different levels of health
within major important population subgroups.
Now, from the point of view of social and political context, an important
future direction is understanding the variety of influences on health. The
National Center for Health Statistics is moving us a long way in this direction
by making it possible to link individual data with area level data. That is
right, isn’t it? Yes.
I think that AHRQ is doing that as well. Information systems moving towards
characterizing health, towards characterizing contexts in which ill health
develops, to characterizing comorbidity, as well as disease, to characterizing
disparities, which are systematic differences in one or more aspects of health
across population groups to find geographically, sociodemographically or —
there is another one — bridging clinical medicine and public health to use
knowledge about health, risks and resiliencies.
Summary. Population health is not the sum of individual health. Averages
don’t represent population health and provide information about the context in
which systematic differences in health occur and how they can be remedied. Why
population health is not the sum of individual health in any useful sense is
because comorbidity individual measures of health don’t represent health and
because of the non-random distribution of health; that is, age, geography,
social differences cause systematic differences in population subgroups.
There are a few more slides, but we don’t need to do them.
DR. VOLPE: While we are waiting for this to come up, I am Carl Volpe. I am
with WellPoint. Dan didn’t give a quiz. Barbara gave a quiz during the
presentation. My quiz comes after. So, take good notes.
While we are waiting for this to boot up, I do want to give you a context.
I am with WellPoint and I am very much going to talk from that perspective.
WellPoint is the second largest health insurer in the United States. Arguably,
we are the most market oriented health insurance company in the United States.
We believe in markets very strongly and you are going to see that influence in
my presentation today.
Quite honestly, I struggled when I read the report five, six, eight, ten
times, to try to figure out how to link the report to our work. So, I do a bit
of a disclaimer here that I am going to be talking a bit about WellPoint. You
are going to see WellPoint come up in a very short time. This is not an
advertisement for WellPoint, but it is an opportunity to contrast because, I
mean, clearly several health plans are represented at this table today.
If you have seen one health plan, you have only seen one health plan. A lot
of the work that I have seen in the development of policy on the government
side really is a reflection of what we see out of integrated delivery system
models, group health plans. We are not that, not at all. About three-quarters
of our business are PPO business. We do not own hospitals. We don’t have
physicians on staff, except for medical management.
So, everything we do is through contract. It is possibly called a virtual
health — a virtual network model. All right. So, let’s see if we can get
going. This is actually the materials that Dan presented. Okay? The mission to
particularly provide time and accurate relevant information and a conceptual
framework and the primary purpose of data is to characterize the health of the
population and influences and actions that affect the health of the population.
What I did is I said, well, what is WellPoint’s mission? How does WellPoint
view the world and that is on the next slide. I am going to read a little bit
of this because I think it is important to understand. The WellPoint Companies
provide health security by offering a choice of quality branded health and
related financial services designed to meet the changing expectations of our
diverse customers through a lifelong relationship.
Now, you notice the words are highlighted in red. That is the way it is
presented in our company material. We highlight those words. What is our
conceptual framework compared to what we just saw.
We focus on individual health security not necessarily individual health
status. We differ from some of our brethren on that. But that is how we focus
our work. We recognize the importance of individual and purchaser choice and
changing marketplace expectations. We reinforce the notion of a lifelong
relationship. Underpinning all of that are our data systems. And what is the
primary purpose of our data? To characterize the transactions between
purchasers of care, employers, payers and individuals, payers, that is us,
health plans, and providers of care as these transactions relate to services
provided to health plan members, a very, very different focus than the kinds of
things you have — yes?
DR. MAYS: Can you just tell me what you mean by quality branded health?
DR. VOLPE: I skipped over that because it was a little bit too — I mean,
how we sell our products. We have many Blue brands. We are Blue Cross of
California, Blue Cross of Georgia. It turns out that the Blue name sells
product. So, what we do is we make sure that as we are selling our products, we
brand it and brand it appropriately. For example, we don’t sell anything under
WellPoint. We sell nothing under the name of WellPoint. All of our services,
but I apologize for doing too much on the business side.
Then what I did is I contrasted the population health information model
that we saw earlier with our view of the world; for example, level of analysis
as to population. Our level of analysis for the most part are individuals and
purchasers. Now, those purchasers can be small companies to a hundred thousand,
200,000, 300,000 lives. Okay. On the population health side, it is enterprise
wide planning and coordination, multiple producers. Our information systems are
solely designed to facilitate the business functions. That is what they are
Information needs are typically driven by research questions and require
empirical data collection. That is not true for the most part. Our data systems
are, again, for business operations and may or may not be empirically derived.
For the most part they are not. There are a couple of examples I am going to
give in a minute where they are, but for the most part they are not.
The expectation of public access and public ease to data information, there
is virtually no expectation whatsoever from our view that there is public
access to information. Quite honestly, some of it we explicitly keep from the
public for competitive reasons. The data on the population health side are
census data, federal and state requiring, and CMS claims. The question mark
there is I think it is emerging as a stronger area.
Our side, there is a piece missing, eligibility systems and then claims,
pharmacy and laboratory results data and CAPHs and HEDIS surveys are really the
two types of surveys you tend to see from our side, which are more empirically
derived. The data information is critical to — and I think is probably the
most important part, data information is critical to the assessment of existing
public policy and to support new policy development. From our side of the
world, there is no expectation at all that our data would be used for policy
development and, quite honestly, there is a strong bias against the use of our
data for policy development.
Let me give you an example. Recently, we were approached by a large
research entity, who has a contract with a subset of HHS and they wanted to
look at a subset of our data. When we talked to them and we spent a lot of time
talking with them, our typical strategy is how does it benefit you, how does it
benefit us. Ultimately, the benefit would have been a public policy benefit. We
couldn’t derive a benefit for ourselves from the data. So, we chose not to do
the study, not to do the work.
I know that may sound a little cold, but that is the nature of a company
like ours, which I said is very market-oriented and our goal is really toward
that end. I failed to mention that we are publicly traded and that will give
you a more — a stronger context for understanding who we are.
There is agreement on the value of data to measure improved population
health. We do have that interest. We do try to make our data sets available
when we can. The need for data standards clearly we are consistent with the
committee on that. We need standardized data. The need for strong but
reasonable — and I should underline “reasonable” — privacy and
confidentiality standards. As a user and in terms of our work, the issue of
reasonable information is important to us. Then the need to recognize the
influence of certain disease states on population health.
What is going on with health plans right now, especially the large ones,
like ourselves? I think also I did want to mention that we are the second
largest. We are soon to be the largest one in the United States. If the merger
goes through this summer, we will have some 28 million members, again, under
this virtual network model for the most part.
A lot of us believe that health plans don’t have a long term future solely
as claims preparer and payment negotiators. That is really not where it is
going to be into the future. We are getting a lot of pressure from large
purchasers in particular. They are extremely concerned about their employees’
health within the context of improving productivity, reducing absenteeism and
reducing their financial burden, but they are talking about it and they are
talking about public health and they are talking about population health.
As you can see, we are devoting significant resources now — I think every
plan in the United States is doing this — to characterize health status of our
members so that we can target our work.
What I would like to do now is switch and just talk about four different
examples. I was going to skip the first one. Barbara, quite honestly, based on
your presentation, I decided to talk a little bit about it. Four examples of
how we are broaching and looking at population health at the margins, which is
really the — I want to talk about our health improvement programs and I said I
am going to talk about ours, but largely every plan is doing this as well. I
want to talk about quality — physician quality improvement and quality
measurement, a project called — and then early epidemic detection, areas that
we have been involved in.
Our traditional disease management, focused on medical interventions to
improve the health status of the individual. We are looking at our health
improvements, again, to improve health status, but it is really to strengthen
healthy behaviors to help the member live their life more fully from the
I don’t know how many of you have ever seen this slide, but virtually every
health plan is showing it. It is the way we tend to look at how we need to
focus our energies in terms of our product design, in terms of virtually
everything we do where 8 percent of the membership costs — about 70 percent of
our dollars go to support 8 percent of our members. Twenty-four percent result
in about another 23 and 68 percent is virtually nothing.
Now, if we switched this in terms of actual number of claims that are
submitted, the distribution changes slightly. Okay? A typical health plan like
ours, we get about a hundred million medical claims a year for 15 million
members and that does not include drug claims. But you can see from our
perspective, our focus is really on that 8 percent.
Our approach has been — and, again, others have done this as well as early
engagement with our members, understand the entire portfolio, the disease
burden. I will talk a little about that more in a minute.
We want to work with our treating physicians. It is not done independent of
them and we are looking for a cost quality paradigm and that is sort of laid
out at the bottom where our goal, of course, is to get very aggressive with our
sickest members to help them achieve some incremental change in health status
and then lessen their need for health services.
How are we doing that? It is really through a combination of disease
management, wellness and prevention programs. We have a strong intermediary
role with our members and others do the same thing to provide information. But
we are playing with this notion of these ABCs of medical management, which
really is advocate, provide education and personal support to our members. We
have individuals who do this. Work with the members to change their behaviors
and we serve as a concierge actually to navigate the health care system. We
have teams of people who do this.
Let me give you an example. We have a woman in her early eighties with
heart disease. She is in our disease management program and so we were talking
to her and one of the questions we asked her is what do you want to do. What
she said is I am in my eighties. I have always loved to travel. I want to
continue to travel. I don’t want this disease to stop my ability to travel.
So, we said, well, we talked to her and we said what do you think you need
to do. She said, well, will you have an exercise physiologist talk with her and
we talked with her doctor. The notion was, yes, let’s see if we can get her
into an exercise program. This is Southern California. So it couldn’t be any
exercise program. We had to get her into a Pilates program, but we did.
But that was her decision. She made the choice. We had a ten year old boy
with asthma, who really, really wanted to baseball. So, we worked up a strategy
so that we could get him on a team so that he could baseball. Those are the
kinds of things that we — we are trying to make it meaningful for the member
as opposed to just treating the disease.
Physician score cards, I am not going to talk a lot about this, except if
you have seen the literature and you talk to folks — I have spent a lot of
time talking in this area — when you talk to folks about health plans doing
quality measurement, you are typically talking with a health plan that has an
integrated delivery system model, group model. So, a lot of the work is being
done in that area. Well, the reality is for a vast majority of the people in
the country, they are not in HMOs. They are in PPOs. We are playing, as well as
others, are looking at how you do quality measurement from the perspective of a
Clearly, quality measurement standards are very, very helpful for us, but
one of the things we are struggling right now is the notion of an attribution
model. We don’t have enough data on it for an individual physician. So, what we
are doing is when a person sees — under a PPO model, may see four or five
different doctors. If it is appropriate, every doctor gets credit for every
other doctor in terms of developing the metric, if it is appropriate. There are
clearly some crossovers that we don’t allow to happen.
We think that while it may not be the greatest statistically, we think it
starts to build up our base and gives us a database on which we can start
giving doctors information. We also think that ultimately if — we have
sufficient data, when an individual physician starts sending patients to a
specialist and they can start looking at each other’s scores and especially if
there are incentives and financial incentives associated with this, which is
where we are going, that ultimately we will be able to have physicians start
talking with physicians about the quality of their care, which we believe is
probably the most important strategy.
These are some of the measures that we have under consideration. I don’t
think any of you who are physicians or clinical researchers would be surprised
at anything that you see here.
Quickly, about antibiotic strength, I am preaching to the choir when I talk
to you about growth of antibiotic resistance and the inappropriate use of
antibiotics in the United States. There is an organization called the Council
for Affordable Quality Health Care, which has most of the large health plans in
the United States working or part of it. We came together in several parts of
the United States and we pooled our data. That is interesting because of the
antitrust issues. We had a third party come in and we sent our data to a third
party so that could develop practice patterns for physicians that cross brands
and cross companies. Then what we did is we took that information and we
provided it back to the individual physician. These are your practice patterns.
We didn’t see it and the individual plan did not see it. The entity that we
hired gave the information out.
Obviously, we had algorithms to determine whether their practice patterns
were appropriate or not. We communicated with them. We did this with the CDC.
You can see a whole host of other activities that we did to move into this
The last one I just want to mention is bioterrorism and SARS. We are
actually in a very unique position to help in the early detection of epidemics,
partially because of one of the things we do that drive hospitals and docs
crazy. That is the prior authorization process or the prior notification
process. It turns out that within our plan and, again, we are the second
largest in the United State s, every individual who shows up at an ER, at an
emergency room, we need to be notified as soon as they show up or within hours.
We can literally — and we have pharmacy claims that we can run against,
which are available to us and many of the large companies have nurse advice
lines and when a member calls in, we record and we note with that member
exactly what the diagnosis was or potential diagnosis, what at least the
symptoms are. Well, we actually have the ability — and have done this — where
we can aggregate all that data and on a daily basis, we can run data sets to
look at legalization patterns and we can actually measure changes in
utilization and because we have zip codes for the most part, we can actually
get subsets as opposed to full state changes.
We just did a pilot project with CDC on this. They were pretty excited
about it. We were pretty happy with it. But, again, it is another example,
well, although we are playing at the margins, you know, our mission is not
public policy, we can contribute to the effort.
Areas for future collaboration, I don’t have to go through these. You know
them. We desperately need standards on quality outcome measures. We are
particularly concerned about standards on electronic medical records. That
would be enormously helpful to us and to you. We are not there at all, not even
close, and it is holding us up. I think it is holding the nation up in terms of
Race, ethnicity, socioeconomic position, it is clearly something you folks
have considered. We have considered as well. So, my final conclusions, the
vision outlined in the study probably has little utility for us as health
plans, quite honestly does. Okay? Our data can help articulate population
health, but it is a secondary function and it will remain a secondary function.
Now, among us, how strong we play will differ. I mean, there are some plans at
this table, who play very, very strongly on the public health side, very
strongly in trying to address issues of health.
Others, like my own plan play very little, but for all of us, I think it is
a secondary function. The marketplace is beginning to value population health
information. There is no question about that, but the application is going to
be less rigorous and less empirical than this committee and academic
researchers might like. I say that because I was in a meeting last week and one
of my co-workers was expounding on population health and how he was building
models for population health to sell products. I asked him to define
“population health.” His answer was, oh, well, you know what it is.
You know what it is. He was not able to do it.
I think you are going to see the words “population health” move
into the private sector. Large employers are now talking about it, but they are
not talking about it the way you do and they are not talking about it with the
empirical rigor that you do. So, if I can give one warning, don’t let the
perfect be the enemy of the good. I think you are going to get something out of
it, but you are not going to get it probably at the level of good empirical
Now, one might choose that bad research is worse than no research. I am not
sure I believe that, but that is up to you. I think that we are going to move
ahead, all health plans, around quality measurement and health improvement and
I think that for the most part health plan involvement and population health is
going to continue to be driven by the marketplace.
DR. LUMPKIN: Thank you.
DR. STARFIELD: You know what the definition of “public health” is
is a successive redefinition of the impossible.
DR. LUMPKIN: Okay. Any questions?
DR. FITZMAURICE: A question of Carl and then others may want to join in.
You said that health plans think of population health differently than we
do. We think of population health, we carve it up and we look to see what are
the parts of the population that have different incidences of diseases. I
imagine you do a lot of the same thing with your health plan population, that a
lot of the methods would be the same. You are just looking at a population that
has health insurance and we look at those, who don’t have health insurance.
DR. VOLPE: I think that is a piece of it, for sure. But we historically
have not taken a broad-based view of the health of the population, even defined
as the employer base. For example, an employer will come to us and say what are
the six most significant diseases that are costing us a lot of money. That is
not a — I would not argue that that is not a population health approach. That
is the kind of information we would provide to them.
DR. FITZMAURICE: But something like, let’s say, you are looking to say one
of the things that pops up is tooth decay in your younger sponsored
beneficiaries and you say all right, we can give them some fluoride toothpaste,
give them a price break or we can go to the public health people and say could
you step up the fluoride in the water to the optimal amount, whatever that is.
You can use population health as a tool to improve your cost benefit
DR. VOLPE: I think that is right. I think to a much lesser extent we have
done that. I think that we are not there in terms of doing that yet. We could
do that. That is a mission of mining our data to help feed public policy and —
DR. FITZMAURICE: And would benefit both.
DR. VOLPE: Would benefit both. That is right.
DR. MAYS: I want to ask a couple of questions. I want to start with Carl
and then I am going to ask Barbara and Dan.
I want to get back a little bit to this question because I guess I was kind
of surprised because it was — in trying to determine, for example, a business
case in terms of your profit line or the best health outcomes, why wouldn’t you
really want to focus on population health? Do you see that as a route to which
you could improve, for example, outcome?
DR. VOLPE: I think that is a great question, but let me tell you how we
think of it. We think of it as how do we design products that will be purchased
in the marketplace and be perceived as affordable and valuable. That is the way
we talk about it. Okay?
Let me give you a — some of our colleagues, for example, are very, very
concerned about having good services for people with AIDS, very concerned about
that, very concerned about it being publicized because of adverse selection.
People with AIDS will want to buy their product. So, in some cases, having
structures that provide good outcomes actually disadvantage the marketplace. As
I said, I shaved the horns down for this conversation but really it is the way
it is looked at.
DR. MAYS: Barbara and Dan, one of the things that this presentation is
designed to do is to really ensure that the committee thinks about population
health. One of the things for the two of you — and particularly Barbara
started out talking about the committee she chaired — what should like privacy
and confidentiality or standards and security take away from your presentation?
DR. STARFIELD: Do you mean that do the standards for privacy interfere with
the population approach? Is that what —
DR. MAYS: How might they think differently or how might they work
differently if they were really to embrace the model of population health that
is advocated in the 21st Century report?
DR. STARFIELD: I think they would think about how we could identify
individuals without identifying individuals.
DR. FRIEDMAN: I would certainly agree with what Barbara has said and I
guess I would go on to say that I think that in public health and certainly at
the state level, we don’t do as good a job as we should and as we can of making
our case for why the use of the data are important to the nation as a whole. As
an example, in Australia and in Canada, surveys — population surveys ask
respondents whether those data can be linked to administrative and census data
sets and there is surprisingly high, by our standards, high positive response.
I think one of the reasons for that is — I think part of the reason for
that differential is that in those countries there has been a better — there
has essentially been better marketing of the importance of population health
information, as well as better marketing of the confidentiality protections
that are wrapped around those data.
MR. HUNGATE: I want to try to talk to both Barbara and Carl and Dan around
a problem that I face in my job in — I chair the group insurance commission.
WellPoint is one of our vendors. We have no measures of health status across
our population. We have 250,000 members. If I treat it as a population, I don’t
have comparable data about my population and other populations. So, I can’t
We have demanded that all the plans that supply us pool their data and
begin to do the physician profiling, but I would ask you, how do you deal with
the variation in risk in patient population by physician in your profiling?
DR. VOLPE: We are just beginning to look at that right now because we are
concerned with it ourselves. I think this is a fairly nascent area for a lot of
us and we are struggling quite honestly. It is the right question to ask. I
don’t have an answer right now, but I know we are looking.
DR. STARFIELD: There are ways to do it. I mean, there are case mix systems
that do that.
MR. HUNGATE: That characterize the health of a population?
DR. STARFIELD: Right.
MR. HUNGATE: But in order to do those, we would have to have our vendors
use common instruments to do that, right?
DR. VOLPE: That is right.
MR. HUNGATE: It would have to be a specified — and we would have to then
have methods that are in common with what DPH had or the state wide population
so that we could have common data.
DR. VOLPE: That is right.
MR. HUNGATE: What are the Blues doing to facilitate that generation of
common data in the way that there is a public/private kind of partnership
around that? What is the action plan to solve the deficit?
DR. VOLPE: Well, let me ask whether there is a perceived need among all
health plans right now to have that because that will —
MR. HUNGATE: There is a perceived need to get one customer —
DR. VOLPE: Clearly, clearly, and not just Massachusetts. We have had
similar requests in California and Texas and we have seen consortia develop in
other states, where they have come together. Now, they have funded that as well
because this — from our perspective — and most health plans were willing to
do it for you, but if it is above and beyond what we do, the question is will
you pay for it.
MR. HUNGATE: Usual response.
DR. VOLPE: That is a fairly typical response. That is exactly right because
our missions are different. I mean, ours is not a public health mission. As a
purchaser, you can tell us what you want, clearly.
DR. LUMPKIN: we are just about out of time. I have a few people to ask
questions. So, let me sort of pose one of my own and then get to the last two I
had on my list. One of the things that we are seeing as we look at the data is
the number of clinicians who are using electronic health records are now
starting to increase. The numbers we have seen nationally are around 10
percent, give or take 5 percent. As that number increases, would the answers
that you give or Barbara gives or Dan about the functionality of the system
change as the cost of data acquisition for more robust clinical data decreases?
DR. VOLPE: It is not the number who are using it. It is how similar the
systems are and the standards. We are on the verge of a Tower of Babel here.
That is why it — that point was on my slides. I think if we can get some
standards there — and I am not sure health plans are the right entities to do
that. I mean, we have complained bitterly about HIPAA, but there has been some
real value to HIPAA.
So, I think that — I worry about the Tower of Babel because I am
experiencing that right now.
DR. STARFIELD: The answer is unquestionably yes, but the value of it will
be concepts in which the analyses, the concept under which the analyses occur
and how we use the information, what questions we ask of it.
MS. HANDRICH: I kind of have a reflection on what you all have said and
then a statement that I took away from this presentation in terms of our
approaches. I think about the things you said, Carl, about the difference
between the way companies view population health and the more general kind of
academic view of public health. There is a company in Wisconsin that
characterizes something they did as a population health approach and that was
that they paid every employee $50 to contribute towards a complete physical
exam and allowed for the employee to report the findings of the exam to the
company and concluded from the findings that the majority of conditions
presented were lifestyle related.
Then they made some decisions about how they wanted to proceed with
wellness and so forth in the future, very different kind of approach in terms
of population health. In general, we have a lot of very visionary and positive
companies in the states, but they are not at all interested in reporting data
or supporting the plans that helped them, reporting data to the state for
population health purposes, not at all interested in doing that.
I think the challenge of what all three of you have presented is — you
know, one challenge is how do we find a convergence of interest so that for
public policy purposes and for private sector purposes, we can begin to address
some point where collection and agreement about data could really move forward.
DR. STARFIELD: You have described it as an academic approach, but it is a
DR. VOLPE: Just as one interesting point, a perfect example of that, we
provide coverage. He owns the data. We don’t because of the nature of the
relationship. So, it gets even more complicated. For example, if we get asked
to make some data available, typically, we have to take out what we call our
cell-funded business. We can’t provide those data because we don’t own it. The
companies themselves own it. So, it is much, much more complicated than any of
us would like, I think.
MR. LOCALIO: Dr. Starfield, I would like to ask you to clarify something
you mentioned. I thought you said that you estimated that adverse events from
medical care were the third leading cause of death in this country. But I have
here the 20 leading causes of death, United States 2001, from an organization
called National Center for Health Statistics and that does not appear.
I also have something called the 2001 United States Unintentional Injuries,
Causes of Death from an organization with the initials EDC and I see nothing
about adverse events from medical care there. I would like to ask you do you
think that is a deficiency in the way we report this type of information?
DR. STARFIELD: Yes, I think it is a deficiency in the way we report the
information. Anybody, any one of us who have had experience with a parent dying
knows that in a substantial proportion of those cases the sign out diagnosis is
not what caused the death. I know it is supposed to be the underlying cause of
death, but it isn’t even that.
MR. LOCALIO: Thank you. You can come and visit us anytime.
DR. LUMPKIN: Boy, that was a softball question.
We actually are out of time. I know this has been a very good discussion.
This is really part of the process that — you know, one of the things that we
want to do as a committee is that when we issue reports we just don’t want to
sort of drop them off, you know, on a stump and run, but that this is a very
important component of our work is to monitor the implementation of the
As you know we have from time to time monitored a number of issues. The
NHII Committee is actively engaged in revisiting our 2001 report. So, I think
these all represent efforts on behalf of the committee looking at this
Ed, as a partner in the development of this report, I just want to give you
an opportunity if you want to make a comment.
MR. SONDIK: I think it is exactly the right thing to do in terms of the
follow-up. I think this particular panel has been one of the most stimulating
that I have heard. Now, that may be because I am so close to it in a sense, but
I think what was offered here were new frameworks, new thoughts on the absolute
fundamentals, like what is health and what is happening in the business sector,
how they are using information. I think we ought to keep this as a kind of a
benchmark, if you will, in terms of panels and come back to this type of thing.
I think the whole purpose of having this vision is so that it gives us a
target. We know what we are building toward and I have said this before here,
that I think prior to having the report, we really didn’t have an articulated
vision. We were basically doing the usual thing of you go year to year. You try
to deal with the demands from all of your colleagues and so forth and meet
those on an annual or on a couple of year basis.
This gives us a longer term vision, but the vision based on what we heard
here today, gets a little blurred, I think, from having heard this. The vision
I don’t think is quite so clear as it was prior to — for me personally
actually, prior to this panel. I won’t keep us here, but I think the
fundamental question that Dan poses in the diagram that shows the influences
and, secondly — and I think one purpose of health statistics is to help us
understand those influences, yet we don’t in our data collection do it in an
optimum way to do that. I think that is a very important point.
I think, secondly, the point of view as to what health is, it couldn’t come
at a more timely point for me because, in fact, CDC, as many of you know, has
been going through a futures exercise and Dr. Gerberding is very, very serious
about this and I am actually at this point involved with several other people,
thinking about goals and the question is how do we formulate those goals. This
was kind of like, you know, filling an inside straight, although I am not sure
I know what that card is that was dealt.
But it is very, very interesting and then to hear you, Carl, also talk in
effect about health, what is health — and I was particularly struck by the two
slides on the perspectives in transition and the medical costs and how — I
would like to talk to you more afterwards about how that actually plays out,
but I think if this is, in fact, what is happening in health — and I am not
sure that I see it at all and I am quite sure that I don’t see in most cases
the effects of this, to a minor degree in tobacco, but not much else — then I
think this is something the committee needs to keep abreast of.
I come back again just to how all of this that we deal with here is being
used. I think that is the critical thing. So, I thank you all for being here
and I thank whoever put the panel together. I think this has been excellent.
MS. GREENBERG: Debbie Jackson actually put it together.
DR. LUMPKIN: With this, we are going to have a break for lunch. We will get
back at — not at 1:15, but because we are breaking a little bit late, about
MS. GREENBERG: Before you break — well, actually, we are all coming back
DR. LUMPKIN: Yes, we are all coming back here.
MS. GREENBERG: Be thinking at lunch as you are eating up there whether you
want to join us for dinner, where the food is going to be considerably more
[Whereupon, at 12:42 p.m., the meeting was recessed, to reconvene at 1:40
p.m., the same afternoon, Thursday, March 4, 2004.]
A F T E R N O O
N S E S S I O N [1:40
DR. LUMPKIN: Okay. We are going to get started. We are behind schedule, but
we are ahead of schedule. I shudder to say that because there is a corollary of
Murphy’s Law that says work expands to fill the time allotted.
There is going to be — the agenda shows that tomorrow morning the Privacy
and Confidentiality Committee was going to be in Room 305. That is going to be
changed to this room because we have invitees. After a fair bit of research in
order to assure that there is going to be adequate time for discussion, it will
convene at 6:30 for Jeff and the rest of us may get there later.
MR. BLAIR: I will read all of the controversial paragraphs for you.
DR. LUMPKIN: Marjorie needs to know how many people are going to go to the
dinner. Please indicate by a show of hands.
We have the phone line hooked back in. Is there anyone on the phone? Okay.
We will continue to monitor as the day goes through.
The next item on our agenda is the Quality Report, which we have seen
before in a different format. It is the new revised, revamped and better
MR. HUNGATE: I believe in the Senate vernacular or the House, I have heard
someplace about the third reading of the bill and the third reading of the bill
is intended to be the final reading of the bill.
DR. LUMPKIN: But unfortunately in the same vernacular, it is possible to
move an item from third reading to second reading. So, please, we won’t adhere
to that vernacular but we will — I appreciate the sentiment.
MR. HUNGATE: But in order to make sure that we came out with a report that
was accurate and correct, we went back and consulted Confucius and we found
that he said, “If language is not correct, then what is said is not what
is meant and what ought to be done remains undone.”
MS. GREENBERG: Have we ever asked him to serve on the committee?
MR. HUNGATE: I want to start by thanking Kathy Coltin for all the work that
she has put into this and commending Susan Canaan, the writer, for her help,
Gail Janes, Julia Holmes, Peggy — Justine spent her time on the phone as we
worked on these things yesterday. So, thank you all for bringing this to where
it is, which I think is a very elucidating, if you will, status report,
milestone of where we are. It is a nudge to important priorities. It contains a
lot of crosscutting issues.
Now, if I could — I am not going to use any power points, but if I could
just get you to look at the page of contents for just a minute and talk a
minute about the structure of the report.
There is an executive summary, which is followed by a matrix and the matrix
follows the order as shown in the third section, “Specific Findings and
Recommendations.” That section is organized into four groups with the
major headings of Assessing Health Care and Health Outcomes, Reducing
Disparities in Health and Health Care, Building the Data and Information
Infrastructure to Support Quality and Balancing Patients Interests in Quality
Now, that organization largely evolved from trying to make sense of how we
meshed this report into the other activities of the committee, the various
subcommittees and workgroups that exist, so that each of these has a natural
affiliation with other organizations of subgroups of the activities. The first,
Assessing Health Care and Health Outcomes, largely is in partnership with
Standards and Security, it seems to me. You know, that is where we are and we
talked about some of the content there, had some unresolved issues, which we
should talk a little bit about here.
What I want to do in this discussion is first let you know about a couple
of things we found yesterday afternoon in the workgroup discussion. One was
that we realized after reading again, that we had really focused in the report
in hospitals and physicians and quality measurement and had omitted the whole
section of long term care, chronic disability. It wasn’t intentional, but it
So, we will acknowledge that in the introduction, acknowledge it in the
recommendation on functional status, because especially important in those
areas. Then finally in the next steps, we will put in a comment relating to
dealing with that.
Another content that got dropped somehow in our working through it was in
the early versions, we talked a little bit about Leap Frog and Fact as coherent
activities on this agenda that are also occurring. That somehow didn’t get
through to this copy. So, we will add that in in the section in support of the
So, back to the structure, the specific findings and recommendations
headings are then used within the matrix, which is the summary recommendation
of the recommendations, which is on page 5. Now, what I want to do in the
context of this discussion is there is no way we can give the specific words
the attention that we do in individual letters. It won’t work. But that said, I
am still assuming that each of you is sufficiently dedicated to the activities
around quality that you have read the report.
If you find things that don’t seem to say quite what they should say,
please give that input directly to Susan and we will incorporate that this
afternoon. The things it is very important for you to point out are any factual
errors that occur or any wrong decisions, just bad decisions. Because if we
said things that we shouldn’t say, we shouldn’t say them. So, those are where I
would like to make sure that we have solicited all the feedback that we
That said, I would just like to leave the floor open for discussion. I have
not had good luck in trying to go down through recommendation by
recommendation. So, I am reluctant to do that.
DR. STEINWACHS: Bob is trying to tell you that we on the Quality Committee
do learn and adapt our behaviors appropriately or at least we adapt them.
MR. HUNGATE: The recommendations read quite differently than they did
before. They don’t make hard recommendations. They more point — they more
highlight the agenda issue that needs to be addressed. Vickie and I have
already talked a little bit about some joint work around the section, Reducing
Disparities in Health and Health Care, which is really a very common agenda
between the Quality Workgroup and the Population Subcommittee.
So, we will be talking about that. There were — Simon and I have had brief
conversations here and I think we should have a little of that conversation now
about what the next steps are specifically around assessing the health care and
health outcomes. As we talked in the committee meeting, workgroup meeting,
yesterday afternoon, we felt that we probably were going to need a two day
period of discussion with representatives of plans, providers, payers, the
spectrum of people involved in the content that we have talked about.
Now, in the section here in support of the recommendations, there are many
organizations that are listed in that section of the report that are doing
things, especially related to assessing health care and health outcomes. So,
some of those other organizations it would seem to me to be appropriate content
for that same discussion. So, that is one next step that the committee felt it
wanted to put into this discussion as well.
I would now like to pretty much throw it open for discussion and let you
folks pick up wherever you wish in the content, given the guidelines that are
DR. LUMPKIN: Okay. We are going to throw it open for questions.
DR. COHN: Well, first of all, I want to compliment the workgroup because I
think this is a much improved document over previous versions. There are
actually a number of questions and issues I have as I walk through them.
Specifically — and I am not wordsmithing at this point — it is really a
question of do we really support that recommendation? Is that really what we
mean? Or what is the rationale for things?
I actually just wanted to raise — I think this is an issue that Bob and I
have been talking about — I mean, before we get into the substances, more the
process piece and I think you have heard me before on this one and I don’t know
that there has been anything in this process that has changed my initial
comments about it. So, let me just reflect on it because it really is a key
decision that this committee needs to make about what to do with this report.
Now, one option, and I know that Bob and the workgroup and others on the
committee would dearly love to get this thing passed so that it is no longer a
version — as I say that, I really am very sensitive to that. I actually, I
think, made an enemy of life from Lisa Iezzoni from my insistence that a
previous report actually be correct before we passed it.
But having said that, I just do have to observe that — I mean, I think
that the tone and the level of the document is actually much better than it
was, but I am still left with sort of going, gee, these are issues that were
sort of pulled out over four or five years. It would be very useful to me just
to get some reassurance from the industry and from others out there that these
are still all pressing issues and that they haven’t been either supplanted by
other things — in other words, having groups come forward and saying, yes,
this is the right agenda, even before we talk about how to do them. That has
been something that has been sort of conspicuously absent from this process.
We have had panels at the full committee of sort of what I would describe
as narrow experts, but I mean people with — probably that is the wrong term —
probably experts who have great depth in their areas, but I can’t remember
seeing people that might actually either have to implement this or might be the
people that we would expect to derive value from all of this ever having a
chance to reflect on them about these recommendations and see if the corpus in
total is the right corpus in total in 2004.
So, you know, as I said, I mean, you know, my preference — and it is just
because I am a cautious person — would be to say, gee, accept this as draft.
Hold those hearings. Get some either agreement, disagreement or whatever and
then with that knowledge, go to the final. Now, as I say that, I am just
bringing this up for conversation because I know I am making lots of enemies
even bringing that up. But I just think it is something that we need to
discuss. And I do want to stay Vickie’s friend at the end of this meeting.
So, anyway, let me just put that on the table and as I said, I just think
it needs to be discussed.
MR. HUNGATE: I agree with him. I think it should be discussed.
DR. LUMPKIN: Okay. Let’s focus our discussion on process. There is a — I
don’t remember the count — there are a number of recommendations.
DR. LUMPKIN: 23. And the issue that Simon raises is do we as a committee
want to vet these recommendations in some process before we adopt those
recommendations. That is the issue before us for discussion.
DR. MAYS: I guess I want to make a suggestion that might help us to think
about how to do that because — no, no, no, I think what Simon says is
legitimate. When I looked at the recommendations, certain groups of them
cluster in ways in which I think that the other subcommittees might be able to
say whether or not — because we have all been in the field in recent times.
So, I think some of the subcommittees could actually say — give some opinion
on whether or not they think that these are still the cutting edge issues for
I think that might be helpful. I thought that would be useful to hear, you
know, like in terms of the privacy stuff, in terms of the race/ethnicity stuff,
in terms of some of the others. So, we might out of the 23, Simon, at least be
able to comment on some subset. Then we could see how much is left that we
don’t know. I think that would help in terms of allaying, you know, the
concerns that exist.
MR. BLAIR: Here is an idea to see if maybe we could figure out how do we
wind up getting the feedback that Simon has properly pointed out without being
in a situation where we go into, you know, a significant redo or significant
delay. So, my thought was while there may be some of the items that may be
dated and need to be updated or revised — the last four years, there is
probably not major change in that, but if we focus on the value and utility of
the document and if we had something like a reactor panel or two reactor panels
or something like that of people that have expertise in — I think you divided
it into four domains in the matrix. Is that right?
MR. HUNGATE: That is correct.
MR. BLAIR: So, maybe if you had expertise in each of those domains, four
reactor panels, for — and the criteria might be more on whether the
recommendations are actionable, whether they are understood and actionable so
that we sort of build on what we have and make it more effective. Now that I
have said that, I don’t know if I have veered too far away, Simon, from your
thought about what you might be looking for with having it vetted.
DR. COHN: Maybe I will respond here. You know, you bring up an intriguing
idea, which is — I don’t know whether it is getting that reactor panel, but
maybe it is sending it out and asking people to comment or respond. I don’t
know at this point that I need a couple of experts in the field necessarily to
tell me it is more — you know, I mean, I am thinking about the American
Hospital Association, NCQA, I mean, other people that, I mean, that might need
to review this and tell us whether we got it right or not.
I don’t know that it has been through that sort of a secondary review.
Certainly, the IOM does that routinely on documents before they finalize them.
I guess my view is rather than — I am trying to think whether the actionable
is the key piece here. I think, the question to me is, one, is it
understandable and maybe, two, is it important anymore because I even wonder
about that in some of these recommendations. You are right, maybe actionable is
the third piece, though I do know we were going to try to engage people and to
figure out how to make some of these happen, assuming we can identify that they
are commonly understood and important.
I guess your idea of somehow doing it in a way that doesn’t take lots of
subcommittee time, I mean, is actually a very intriguing idea.
MR. REYNOLDS: Being new and not understanding the process completely, as I
red this, it appears if you look at the last two columns, Column 3 tends to
focus on what I would think would fall predominantly in the Standards
Committee; whereas, Column 4 tends to fall into more information that would be
in electronic health records, the CHIs, the EHRs and those things.
It seems to me that any hearings or anything that went on would almost be
the jurisdiction of both committees to hear it because it really is very
disparate groups and provider organizations that tend to do this. Your billing
office tends to do a lot of the stuff when you look at the 837 and so on.
Whereas, in the medical records world, when you are looking at the other
things, the actual parameters are selected here.
I understand the recommendations but I need to understand better what our
process would be because it seems like it is almost two committee
recommendations firmly entrenched but, yet, I don’t see a process as to how
DR. LUMPKIN: Let me sort of toss something into this mix. Let me share my
concern. First of all, I am not really thrilled about the idea of holding more
hearings. I think we have held four years worth of hearings. The subcommittee
has sort of come up with their best approximation. What we really say in here
isn’t necessarily saying this is exactly what should happen. I think I might be
inclined, more inclined — and this can be done in a number of ways. We could
adopt it provisionally, send it out for comment, collate the comment and then
present it for a final vote in three months or the meeting after that. Or
alternatively just send it out for comment.
But I think that conducting more hearings is not going to necessarily be
the most productive. The Quality Subcommittee has other items that ought to be
on its agenda and they ought to be getting to work to it and I think we need to
move beyond this document in some sort of way that we can assuage the concerns,
but at the same time set some sort of agenda for work.
MS. GREENBERG: I agree with you and I am — I mean, I think there are a few
options as you said, but one as to what to do with the report at this meeting,
one is to just defer it until you get the comments or have hearings or
whatever. Having followed all of this for a good four years would not be in
support of that, but, you know, that is for the members to decide. Another
would be to accept it as a committee report.
A third might be for the full committee to accept it as a report of the
Quality Workgroup and recognizing that — and commend the Quality Workgroup for
having done an excellent job as I think they have and then recommend future —
how to take the process forward, which could be hearings. It could be sending
it out for comments. It could be both of those. I think even in that context it
could be sent to the department as having been accepted by the committee as the
report of the Quality Workgroup with these future plans. That would be an
alternative if a majority of members aren’t prepared to endorse the report as
I think we should find out whether members are ready to endorse the report
as it is, but you don’t really want like a 9/8 split or whatever. I mean, you
want a consensus type of approach. So, I think to defer the report for this
future work, I personally don’t see any benefit to it, particularly if it is
couched in — if the acceptance of the report is couched in terms of, you know,
this is what we have heard. This is what we have processed and now we are going
to try to get more — get reactions to it, rather than, you know, we absolutely
think everyone of these things has to be done and done within 12 months. I do
think that there would be value to the field for this summation of the work
that has been done by this workgroup and by this committee in the quality area
over the last four years for it to get out.
I think it is well put together. I am sure there are some things in here
that everyone — there is somebody who does agree with something in here, but,
frankly, this is very consistent with things that are being talked about in a
lot of different circles, quality circles, standard circles, electronic health
records circles. I don’t see anything here that is really completely out of
touch with those.
So, I think if we want to say how can we kind of get more visibility to
this area and move this quality agenda along, I think we need to move this
report along. I think if the goal was to try to kind of — I don’t want to say
dumb down, but get the report down to the fact that there is nothing in it that
everyone out in the health care industry doesn’t completely agree with, then
there won’t be a report.
I think what this represents is an important body of work. So, you know,
again, I mean, I think an alternative would be to accept it as the report of
the Quality Workgroup with the following plan if it can’t be accepted at this
point by the entire committee.
DR. LUMPKIN: Let’s go with Jim, Jeff and then Vickie.
MR. SCANLON: A couple of more procedural options if possible. Again, the
situation I think Simon was alluding the processes of the IOM. It is often the
case where a workgroup has done, a committee has done what it believes to be a
cumulative assessment of the state of the art, but really needs to get a sense
of priorities on what do others think. I think there are ways to do that. I
think, again in addition to accepting the report as a subcommittee report by
the full committee could be provisional acceptance by the full committee or
acceptance as a working paper, which further — you know, in other words it is
fixing the findings as of that date. It is asking for guidance and comment on
the priorities and on the direction and it is to be refined as a result of a
review and comment process.
But I, too, agree. I think we need to ratchet in the progress that has been
made and really start getting some feedback now. There are very specific
recommendations. there are others that are more directions. I don’t think it
would be helped by hearings, as much as it would be by a broader
circularization and asking for comment and guidance and priorities on — are we
sensing — is this where we are now? This is a cumulative set of
recommendations. Does this reflect the priorities and the needs now?
MR. BLAIR: Unfortunately, I have been away from my office now since
Saturday morning. So, I did not — and I apologize — I did not have a chance
to actually read the report before this. So, this question you might be able to
answer directly or maybe it is going to prove unnecessary.
But my thought is that maybe one of our next steps is to look at each of
the recommendations, identify who or what — whether the NCVHS would be taking
action on those if they are directional or whether it would be directed to —
some of these would be directed to health and human services. In short, I think
we ought to focus on how do we take these recommendations and move them to a
point where some action or follow-up is taken up on the recommendations and
just don’t say it is a nice document, but action is taken. Maybe that would be
something that NCVHS could take on as its task is to do that. Either that or
maybe because I didn’t have a chance to read it, maybe it is clear already in
the document. So, you tell me whether —
DR. MAYS: I guess I want to kind of piggyback on some of the things that
Marjorie said. In the report, there are some areas that it is almost like what
they have suggested is work that overlaps with some of the other subcommittees.
In particular, I will just talk about mine. I would hate that what we would
have to do is kind of wait because I think we were kind of slightly planning
that that is an area that we would want to move forth on.
I can speak from having been in hearings that at least that area is still
well alive and is something that, you know, probably we might be able to make a
contribution on in terms of, you know, moving forth on it. So, it is kind of
like for me, Quality has some other things on their plate that I would actually
love to see them be able to do. That is one point.
The second is when I looked through it, I wondered about whether or not
there weren’t also some things that could be pulled out and put into letters.
There is the report in and of itself, but if you really want to make a change
— I mean, we have talked about this before, that sometimes things can get lost
in reports, but there were some items that I thought — and, again, that would
be for Quality to determine — whether or not they can pull them out and send
them to, you know, either the professional organizations — I don’t know that
they all go to the Secretary, but there are other places, I think, that they
go. So, you know, that was something, you know, I wanted to also put on the
table for consideration.
DR. LUMPKIN: Let me perhaps — and this could be not a helpful idea, but I
am going to toss it out anyway. As I was kind of skimming through the report
again because it was a little bit too long to sit here and read it through to
check on it, but if we were to develop a concept of a candidate recommendation,
that we have a series of 23 candidate recommendations with pretty much the same
text that we have here, with maybe one short paragraph in the beginning and
stating that these recommendations really set the agenda for the committee in
relationships for quality.
Some of these recommendations we believe can be processed fairly quickly
and would end up in letters as a final recommendation. Some of these may
require some work by one or more subcommittees to really bring to fruition how
it can be implemented, if it could be implemented. But it short of sets an
agenda for which there are 23 items in this agenda in which we as a committee
think that we need to address.
Let’s take one of the most controversial, which is, well, one of those that
are more controversial, and it has to do with those that affect the 837 for
which there is an S deal. We have had a lot of debate about the fact that these
recommendations, whether it is to change the character of the race and
ethnicity field, to change the identifier of whether or not a procedure existed
prior to admission. All these are fairly complex.
By calling this a candidate recommendation, we have put a stake in the sand
that this is an issue that needs to be addressed. The 837 may be the place to
put it but there could be other alternatives, but it says at some point in our
agenda as a committee that we want to address this issue of bringing out this
data in a way that is meaningful to allow it to be functional. There are things
that may impact our discussion as part of CMS’s initiative on pay for
performance. All the sudden some of these criteria may seem important to
include in a transaction standard because they will be based upon
So, I think there are factors that come into play, but it would allow us to
move the report, set the agenda without coming out and saying that every single
one of these are things that should be implemented today because we believe
that people are ready.
DR. COHN: I am trying to think of what a document that lists candidate
recommendations looks like. I guess I was — I am not sure I am against what
you are describing. I have been sort of resonating with what Jim Scanlon was
describing, which was sort of — you know, accept a recommendation
provisionally, get it off the plate of the Quality Workgroup because I feel
that they have probably seen this so many times that they don’t want to ever
see this one again. That is an overstatement. But certainly bring it to the
full committee with the idea being that we would over the next several months
get feedback from a variety of others in written form.
I agree with you. I don’t think we need hearings on this topic right now.
With that information, basically decide in June, does it stop being provisional
or what? I guess I was sort of thinking that that made almost more sense than
trying to describe all these recommendations as candidate recommendations
because I don’t know what to do with that exactly. I would much rather have it
be provisional, have us accept it, have it be able to be put on the web and
then further discussion going from there.
DR. CARR: I like John’s idea because I think that as a I read this, there
is either too much or too little. We have to either go in greater depth or very
superficially. I don’t know that either one of those in a report of this kind
will give folks the information that they — that adds value. I think that
exactly what you said, like the ethnicity is a big issue. Where it goes is also
a big issue. Trying to have a paragraph addressing that doesn’t do it. But I
think to give credit to the work that has been done so far, it is quite
reasonable to say here are candidate recommendations and fast track the ones
that are able to move forward and really go in depth on the ones that have a
lot of stakeholders, there are a lot of good ideas.
MR. HUNGATE: My comment is related. It is a procedural one and I think that
one of the problems that we have with this report is that it is extremely
broad. You know, it covers the waterfront. That makes it really hard to get
each of the pieces in full agreement with the rest of the structure. So, if we
think of this as a report back to the committee of candidates for other
subcommittees to assess the priority of with us working in conjunction to
whatever way is appropriate, then the actuation comes from a more knowledgeable
One of the problems we had in getting the recommendations was that we
weren’t always sure that we had the best knowledge in the full committee to put
here. But we didn’t have a mechanism for improving it. So, the sense — go
ahead. I think I have said what I —
DR. LUMPKIN: Jeff, then Marjorie.
MR. BLAIR: A question. It is my understanding that the Institute of
Medicine is going to be asked, if they haven’t already been asked to come up
with performance indicators for quality, safety and efficiency. Are there a lot
of these recommendations that would be helpful to them as they begin that
MS. GREENBERG: Well, these aren’t measures. These are enablers so that you
can develop measures.
MR. BLAIR: I do understand.
DR. LUMPKIN: My guess would be, having sat on Institute of Medicine
committees that if we have a report out there, they would look at it as part of
the data gathering initiative and it would give them some guidance. So, I think
that is sort of the answer, although it is probably not as specific as to drive
what they will do.
MS. GREENBERG: Does Don want to say something? I will go after Don if he
wants to say something.
DR. STEINWACHS: Simon is in between.
Let me just say one thing because I thought — there were two pieces here I
liked very much and I was hoping maybe they would fit together. One was the
idea that the committee might accept this as a report of the Quality Working
Group and so, therefore, it is a product and the idea that you were putting
forth, John, about this being a possible or tentative recommendations or, you
know, it seems to me for the entire committee they are, if you accept this as
the product of the working group and to then try and move that along.
I would like to find that kind of middle ground. I know Bob and others
would about achieving some closure and then we can direct our attention much
more specifically to certain areas because it has been, I think, difficult.
Those of us who came along after a lot of the hearings had already been held to
wrap our arms around this and try and be responsive. So, if we could find that
middle ground, I would really be a very strong advocate.
MS. GREENBERG: I agree and I think it would be — well, I defer to Bob on
this, but my sense is it would be better with, you know, a little wordsmithing
here or there or whatever, but to accept this report as it basically currently
is as a report of the workgroup and then for the full committee in its
distribution of it or transmittal of it or whatever, to characterize these as
seen as candidate recommendations that need to be addressed, then to have to
change this report and call them candidate recommendations within the report or
dilute the report as it is because I think the report — it is not that these
are the only things that people — that the Quality Workgroup heard or that
were heard in these meetings. These are the ones that have been agreed to and
that have been prioritized to be included in here in various ways. Sometimes
the language is stronger than others. Other times, it says we need to find a
mechanism, which, you know, really opens it up as to what that mechanism might
But I would myself be sorry to see, for example, pulling things out and
putting them in letters or diluting this by calling these all candidates
because I think the report was great effort and I personally think it does a
good job, pulls this altogether in an organized way around some very important
themes. That has been agreed to by the Quality Workgroup. But I think the
workgroup would realize that that may not at this point have the same level of
comfort in the rest of the committee, but — so, I mean, that is my sense
having watched the process because so much good information came out of those
hearings over the years . I think it has been synthesized very well and I think
as a product of the Quality Workgroup, I would hope that it not be diluted too
DR. COHN: I was actually going to support what Don was saying and I
actually thought you said it very, very well. I am not sure exactly what to say
now that Marjorie has commented.
MS. GREENBERG: I was agreeing with Don.
DR. COHN: I don’t think you were actually. I think you were saying that we
put the term “candidate” or whatever wishy-washy — I mean, whatever
words that we might use in the transmittal letter and my observation is
generally transmittal letters get lost very quickly and you are left with
MS. GREENBERG: I was thinking if you send it out to people for their
comments you would make clear — if that was the next step, you would make
clear that the full committee sees these as candidate recommendations.
DR. COHN: Well, that is what I am saying is I actually thought — I liked
John’s suggestion that we need
— I mean, it may not need to be more than a paragraph some in here and it
would be something that could be handled by wordsmithing or whatever after any
sort of a vote, that some way to indicate that, you know, that this appears to
be the right direction. We think that these are the right things. We call them
candidate standards or candidate whatever recommendations —
MR. HUNGATE: These are not standards. These are not recommendations.
DR. COHN: I am in the wrong subcommittee. Sorry.
MR. HUNGATE: We have identified data gaps and need. We haven’t said how
they need to be addressed.
MS. GREENBERG: There were some suggestions.
MR. HUNGATE: We made some suggestions, but, you know, it is — and we have
clearly deferred to the content of other groups as being the critical part of
that and every time we get into it, we get into the multiple priorities of the
multiple groups and we are unable to — that is the problem I have got.
DR. STEINWACHS: When you look at our grid, we do have in the center of that
something we call a recommendation. So, I think that it does —
MR. HUNGATE: That is true.
DR. STEINWACHS: It is an assessment from the hearings, what we saw as
needs, that we felt needed to be responded to. So, I think going down the
pathway Simon was going, if we could put into the introduction of this that
makes it clear this is a set of findings from the working group that have been
accepted, hopefully by the committee and we see this as a blueprint for next
steps by the committee and then the working groups or something like that or,
you know, sort of takes it out of maybe actionable outside of the committee.
I think that was the issue. This is sort of a report that says we hope the
committee acts on these things and takes them up, not that this is something
that is going to the Secretary as itself for action.
MS. BEREK: You are sitting there reading the report trying to figure out
what to do. I think one of the issues here is that we are now talking about
hearings that almost no one in this room was around when they were held and I
think it will — if we just keep going over it and over it and over it, we will
not improve it because almost everybody here wasn’t there and isn’t sort of
bringing anything to it.
I think we owe it to the group to say we will accept this and give them the
opportunity to go forward either with some stuff from that report or other
things they would like to work on.
So, I think the kind of compromise of putting sort of an opening, it could
be a letter that is found in the report if you want to make sure it doesn’t get
lost that says — you know, we just need to find some step that says we know
this is not a hundred percent finished, but if we try to finish this, it is
going to take the rest of our lives and that is just sort of a feeling I have
got. I have attended a few of the Quality meetings and it is very hard to sit
and try to figure out what to do with recommendations when you weren’t there.
That is going to get worse and worse and worse. It is not going to get better.
DR. LUMPKIN: Let me suggest this because while I was sitting here not
paying attention to my duties as chair, but kind of thumbing through the
document to try to find some examples, I think that we can by putting a crafted
paragraph at the beginning and the end and with some modification do in the
document itself what we are concerned that the letter may not do.
But rather than trying to convince the whole group, let us take this to the
Quality Workgroup and with the charge to come back with one of two deliverables
tomorrow, one being a cover letter if we don’t think that we can do some simple
modifications of the document.
The reason why I say that is if you look at this, look at Recommendation 16
or Recommendation 4, Recommendation 4 says, “Operating physician requires
an existing data element for operating physician to be reported for the
principal inpatient procedure.” I know why that is there. I think it is
important to have it there, but there is no way you can read that as being
anything other than a recommendation.
Recommendation 16 says, “Adopt ICD-10-CM for coding the classification
of diagnosis and health conditions in administrative transactions.” I
don’t care how you may disagree with that, we have already taken a position on
it and we ain’t backtracking because we had enough of those meetings. So,
having said that, I think that in that context we have actually used some of
these as candidate recommendations and some of these have been vetted and have
resulted in letters and recommendations to the department.
So, if it is agreeable, maybe we can try to hash this out in the workgroup
and come back with one of two deliverables, either the letter plus this current
report of a slightly modified report, which I think we can do fairly easily
with just a few sentences and a few extra words that you can do a find and
replace to do.
DR. MAYS: That sounds reasonable. There is only one other piece that I am
concerned about and it is kind of what we have dealt with throughout, which is
Quality has always been the group that had their hearings — not their
hearings, but had their materials report to the full committee. So, what would
be really useful is if you all come back also with a process of what you want
to happen after this over because if things come in — I mean, it is like and
the workgroup is already off on something else, it is like is this going to be
something that the full committee does? Is this something that the workgroup
does? I think if you all would just think about whatever you propose, giving us
a process of how we are going to deal with it, whether we are going to deal
with it as a full committee and make decisions about things or whether or not
this is going to be again the workgroup and then the workgroup comes back to
the full committee — because this is not the end of it from what you are
saying You are going to put it out there in the world and then what is going to
happen is, you know, you are saying to some extent did we get this right, are
there other things?
So, somebody is going to give you some feedback and I just want to be clear
when that happens, do we have a process by which we will deal with it and is it
going to go to the workgroup or are we going to be back to the only way we deal
with this because then it gets to be do we have time on the agenda? It gets to
be it has to come to the full committee.
DR. LUMPKIN: And that actually could be done on the grid.
MR. HUNGATE: I think from the grid we can designate the content, subset of
this organization that deals with the feedback. I would even — you might want
different kind of feedback from different kind of groups from different parts
of the grid. It is not clear to me — it is clear to me how this could help
inform our work here. It is not clear to me how we prioritize the pieces inside
it in an effective way.
DR. LUMPKIN: But if I could suggest that what that additional column in the
grid would do is suggest the responsible pieces of the NCVHS for accomplishing
those tasks, which is vetting those recommendations and then as we do all of
our other work, the subcommittees will have to determine the priorities of
I mean, to be truthful, if there is anything that is dealing with standards
between now and when we complete e-prescribing, it is just not going to make it
on the agenda because we all recognize that we have been charged by Congress to
do something and do something quickly, but there are other items that may
appropriately fit within the venue of either the Quality Committee at the next
meeting, may look at all these items and say, okay, let’s prioritize them. Now
we have got the report under our belt, let’s look at what we believe as a
Quality Workgroup we want to address.
So, I think there are ways to do that and what we can do in the grid is
just sort of determine what is the portion of the NCVHS that needs to
prioritize that particular candidate recommendation.
MR. HUNGATE: The intent of the first column and category was to provide
DR. LUMPKIN: And so, like I said, the work will be very easy to do.
MR. HUNGATE: Well, the thing that may not be quite as easy is that we have
got multiple groups named.
DR. LUMPKIN: No, I don’t think that is a problem.
MR. HUNGATE: You don’t?
DR. LUMPKIN: No. It just means that the prioritization is going to have to
take across groups and that is the reason why we had the hearings for those of
you who were there then,which aren’t very many was because it was so cross
DR. COHN: It sounds to me like we are maybe getting close to being done
with the process things. I was actually hoping we could actually talk about
some substance for the recommendations.
DR. LUMPKIN: Time-wise we have — we don’t have a document from
populations. We actually have until 3 o’clock.
DR. MAYS: I do want to make — what the report should be about because
everybody here is really new.
DR. LUMPKIN: Okay. So, we will save ten minutes at the end. So, we have
another 20 minutes by my watch to go through and identify issues.
DR. COHN: One of the values of being a long way from here is that I get to
read all of this stuff on the airplane. So, it allows me to — and Vickie
DR. STEINWACHS: Simon, I would be willing to buy you some drinks on the
DR. COHN: I just had a — once again, I think this document is much
improved but there are a couple of things that I looked at that I wanted to
actually refer back to the workgroup or at least have you consider and this is
probably some areas for revision. If it is okay, I will just run through — I
mean, there are not that many of them. I mean, one is on page 31, the
functional status coding and also referenced in the executive committee
I think the recommendation is fine. The recommendation has to do with
functional status coding. It is to review the available options for coding
patients, functional status in administrative transactions, EHRs and other
clinical data sets and recommend standard approaches by 2005.
The body of the text only references a 2001 NCVHS recommendation on
functional status. I think we would all be reminded that literally a month and
a half ago we accepted a CHI report that had a number of action steps related
to identifying functional standards for functional status coding and I think we
need to reference that, incorporate it and that really needs to be part of the
next step. Trying to just keep the various recommendations, things that have
approved and why in a line. That obviously includes evaluating SNOMED and other
things in addition to ICF here. That was just Recommendation No. 1.
MS. GREENBERG: CHI was mentioned in the matrix, but it didn’t get picked
DR. COHN: Let me ask a question and this is just because of my own
confusion regarding Recommendation 12, which is race/ethnicity data for patient
— this is on page 35 and it says, “Investigate how best to capture race
and ethnicity on a standard provider transaction and we recommend adoption of a
consensus-based approach. This is once again a question and I don’t have the
answer to this one. I had initially been reading this as a recommendation that
we needed to get comprehensive race and ethnicity data and that this was
somehow a proposed recommendation about maybe how to do it or whatever and that
that was really the issue here.
I have heard others commenting that really somehow the issue is the
perception of the provider, how he perceives the race and ethnicity and I guess
I need to understand this one better to understand whether I support it or not
because I had thought and had previously before commented that this is an issue
that we don’t have good race and ethnicity data. If, indeed, this is a way to
identify what a provider perceives as a patient’s race in this and ethnicity
versus the patient’s race and ethnicity and identify some sort of something
going on. I guess — can you clarify this one?
DR. LUMPKIN: Let me suggest before you put Vickie on the spot, there is a
report from the Institute of Medicine commissioned by the department. That
report has undergone — and a workshop report and that final report has gone
through external review and is now — the reviewer comments are being
addressed. So, that is fairly close to completion.
AHRQ is very actively engaged, has had some initial meetings with health
plans in addressing the issue of collection of race and ethnicity data. There
is a fair bit going on so that really in the context of identifying this as a
place that our committee can work within the time frames of saying this is, in
fact, a candidate, the clarity will be applied to the water of this
decision-making. So, I think it needs to be vague to give us room to come up
with a recommendation that will be more specific and consistent with currently
DR. COHN: Okay. I think you have responded to my bit of confusion, which is
Now, comment about — it is the supporting part for Recommendation 20,
which is interoperability of clinical data systems on page 39. I am not certain
that I know the right wording here, but I don’t think lack of standards for
open systems architecture is right there. I am not sure if I know what the
right word is. But I would urge you to maybe talk to Harry or Jeff and see if
— it seems like the wrong concept for all of this.
Now, the final question I have and it is actually the page for and I
apologize — I am trying to think of how best to ask this one. Recommendation
17, which is standard clinical codes for procedures, I guess the
recommendation, which I support, which is create a mechanism for efficient
mapping, querying and aggregating procedure information across care systems is,
I think, something that we all support. I am not sure it is the same as the
title, which is Standard Clinical Codes for Procedures. I guess I would just
ask the committee to — the workgroup to review the couple of paragraphs that
relate to that recommendation because it is sort of quite — maybe you need a
different title for that recommendation.
That is what I saw as I was looking through.
MR. HUNGATE: The recommendation as stated and the content preceding it
don’t seem to talk —
DR. COHN: No, it doesn’t seem to connect quite. I think that was — I
thought I had more, but no. I will stop there.
MS. CANAAN: Is the issue just the title of Recommendation 17, the language
of the title or is the preceding text?
DR. COHN: It is a combination. It is both. I will just have you all look at
it because I found there was sort of a disconnect as I was reading through it.
MR. HUNGATE: There were two substantive changes back up in the matrix that
Marjorie had suggested that we should — that did not get incorporated that I
think need some discussion.
On page 6, in the content of the Recommendation No. 7, relating to
functional status codes, a suggestion was made that it should relate to
existing data systems as opposed to evolving data systems.
MS. GREENBERG: This is No. 7, Recommendation No. 7 on functional status
codes and you see this third column is whether it is existing or evolving data
systems. In other places when you are talking about doing things on the 837, it
is existing data systems. This included consider revising the data sets, NUBC,
NUCC, X12, et cetera, for a new standard transaction. So, it seemed that it
wasn’t just for evolving data systems, that it is for existing data systems and
probably evolving data systems as well. The reference was to the 837 in NUBC.
These are existing data sets. So, it seemed misleading to say evolving here
MR. HUNGATE: That seemed correct to me, that the observation —
MS. GREENBERG: Then it should say existing and evolving.
MR. HUNGATE: Existing and evolving seems most correct. Okay.
She also suggested that in the next following recommendation — well, No. 9
actually, in the far right hand column, that CHII should be replaced by HHS
MS. GREENBERG: I didn’t think CHI was doing anything with survey sampling
approaches. So, I felt the Data Council was the appropriate group.
DR. LUMPKIN: Any other changes?
MR. SONDIK: I just have a question on 12. Recommendation 12. Recommend an
adoption of a consensus based approach as opposed to an arbitrary and
DR. LUMPKIN: Recommendation 12, which reads, “Investigate how best to
capture race and ethnicity on a standard provider transaction and recommend
adoption of a consensus-based approach. Given that we are going to call this a
candidate recommendation, I think we can — investigate how best to capture
race and ethnicity on a standard provider transaction.
MS. GREENBERG: Just end it there.
DR. COHN: Can I ask another question, John?
DR. LUMPKIN: Sure.
DR. COHN: Actually, I think this is sort of a differentiation between the
— what was on the executive summary sheet versus in the body of the
recommendations. Recommendation 12 on page 35, that is about investigate how
best to capture — okay, I am sorry — these are options to be considered. I
guess I was looking at the options to be considered on — relating to Option 12
and you had mentioned — I think you had mentioned previously new patients and
I think I had asked, well, gee, how do you identify a new patient and you came
back with, well, it is a CPT&EM code for a new patient service. So, I was
sort of coming back to you, well, are you only going to have physicians code
this? Is this not also something you would want to get institutional providers
to code this or — I am looking on the top of page 8 under your potential
options to be considered. I was just referencing your proposed potential
options to be considered. If, indeed, the area you are targeting is really just
like — not providers, but physicians, then this is a solution but you appear
in your recommendation to not be at all — you are sort of saying, gee, we
don’t need to come up with a solution for hospitals.
MR. HUNGATE: So, that is narrower than it needs to be.
DR. COHN: It showed up here. It wasn’t in the body of the text. So, I had
sort of missed it initially.
MR. HUNGATE: Thank you.
DR. LUMPKIN: Any other — Gene.
DR. LENGERICH: In Recommendation No. 7, it says by the year 2005. Is there
a particular reason for that? It appears that is the only one with a date.
DR. LUMPKIN: I think we can lose that by 2005. It is in the grid and also
on page 31.
MR. HUNGATE: Which recommendation number was it?
DR. LUMPKIN: 7, page 31.
MR. HUNGATE: So, strike —
DR. LUMPKIN: Just “by 2005.”
MS. GREENBERG: I think the reason was — is that the others, you already
know — an approach is already suggested in this one and this was sort of
consistent with the CHI. There isn’t an agreed upon approach recommended yet.
DR. LUMPKIN: But if we now have these as candidate recommendations, if no
committee prioritizes this recommendation, having it by 2005 becomes
meaningless. So, if it gets prioritized then —
MR. REYNOLDS: One other consideration that could be in all these ones that
deal with the race and ethnicity is why not consider recommending capturing it
at the data point? For example, if you are talking about a payer when they
enroll someone, Medicaid enrolls people, Medicare enrolls people. That doesn’t
DR. LUMPKIN: Actually in Medicare you can, in fact, I think it is captured
in Medicare. The difficulty is in capturing it at the point of enrollment is
that when a provider sees a patient, they don’t necessarily have enrollment
data. So, that is the conundrum that is associated with how to do it. The plan
may have it. The provider may not and it depends by service.
MR. REYNOLDS: I am not sure why that matters because once a transaction
goes somewhere, whoever it goes to —
DR. LUMPKIN: Well, because if you are going to — let me give you an
example. If you are going to make a difference in trying to reduce racial and
ethnic disparities, if you have a provider, an electronic health record, that
data may be pertinent at the time of service delivery. It may be pertinent at
the group level, but not necessarily once it gets to the plant.
I am not precluding one or the other, but I think that there are a number
of things that makes this such a complex issue.
MR. HUNGATE: So, we are committed to go away and do our work and come back
DR. LUMPKIN: That looks like the plan.
MR. HUNGATE: John, do you think we can do it in an hour and a half?
DR. LUMPKIN: To my mind, in the interim, rather than going and
eavesdropping on Simon, I will just try to jot some thoughts down and walk
through the documents on what changes I think need to be made.
MR. HUNGATE: The committee was quite of the content here. I think it is a
nice recap. The prioritization is not perfect. You know nothing meets
everybody’s expectations, but it is a pretty good working list of some things
that are important. I think that is the contribution.
MS. CANAAN: Kathy Coltin is not here to receive your plaudits, but she is
the one who is responsible for the very large majority of the writing and the
intellectual work on this report.
DR. LUMPKIN: Thank you. We are now ready to adjourn, but I did promise
because Vickie needs to give us a little ten minute set up. See, she thought I
DR. MAYS: No, I didn’t. I have already moved the mike.
What I wanted to do because we have lots of people here, who were not here
when Populations began. I mean, I am kind of learning listening to this
discussion before when Populations actually began its hearings. I think it
might be useful since we don’t have the report at this point in time to at
least give some background so that you have a sense of how it started. We are
going to wrap ours up so that we don’t have to worry about — there is not
anybody sitting at the table that were at the hearings.
We began in February 2002 to hold a set of hearings. The purpose of the
hearings were really to determine whether or not the OMB guidance on the
collection of data on race and ethnicity was actually working sufficiently. We
wondered the extent to which, for example, there might be variances that were
given to agencies. We wondered, for example, whether or not some of the groups
who were very — some of the racial and ethnic minority groups, who were very
concerned about how the collection would be done. You know, we wanted to hear
So, we began in that February with a set of hearings in which we invited in
— and that is what the report will be on the next time — is our federal
agencies, who were conducting national population-based surveys. The reason for
that was, number one is to find out how they were asking the questions; number
two to find out in terms of the small numbers that they might have, how they
were allocating the categories for race and ethnicity and how they were
Then, third, we also were interested at that time in finding out, since we
are interested in health disparities, what questions besides race and ethnicity
were being asked to determine whether or not there were health disparities in
particular populations. So, that is the background in which we began the
We moved from that hearing on population-based surveys to hear from the
states and then the states we also wanted to hear about vital statistics and
then we moved on to do specific racial and ethnic minority subpopulations. The
report that will come to you is a report that will just talk about the
collection of data on race and ethnicity and the measurement of disparities in
population-based surveys. I just wanted to give you just a little bit of
background as to what you can expect. We learned our lesson because we came
here before with 66 recommendations. So, we now know better than 66. Actually
what has happened is over the time is that some of those have been pulled out
and sent in other letters. So, I think we have a — at least my own and the
committee’s learning curve here, I think that we have a much better sense of,
you know, trying to come in with recommendations that actually can make a
difference as opposed to just a pie in the sky. We will have some pie in the
sky, but, you know, we won’t have 60 pies in the sky.
DR. LUMPKIN: Jim.
MR. SCANLON: Before we break up into the working sessions, you will notice
that we have added a number of new staff to the various subcommittees and that
is the result of a departmental solicitation of all of our agencies for staff
to serve on the subcommittee staff.
So, please welcome the new staff when you get into the breakout sessions.
These are all very capable folks. We have always had good staff for the full
committee. Earlier today, we had a representative from the Census Bureau. She
is acting as liaison today. We invited her to the full committee, but whatever
suits the committee’s purposes, whether we want someone on the — we clearly
want someone to be at the Populations deliberations as well. The census folks
said let’s see what the best person might be.
We are actually happy to have the new staff. Please welcome them in your
DR. LUMPKIN: Okay. Standards and Security, I think, is here and Populations
will be in 305 and followed by Quality.
[Whereupon, at 2:52 p.m., the meeting was concluded.]