[This Transcript is Unedited]


Subcommittee on Populations

Working Session

May 13, 2005

Hubert Humphrey Building
Room 705A
200 Independence Avenue, SW
Washington, DC 20001

Proceedings By:
CASET Associates, Ltd.
10201 Lee Highway, Suite 180
Fairfax, VA 22030
(703) 352-0091


  • Donald M. Steinwachs, PhD, Chairman
  • Justine M. Carr, MD
  • A. Russell Localio Esq, MA, MPH, MS
  • Carol J. McCall, FSA, MAAA (By telephone)
  • William J. Scanlon, PhD
  • C. Eugene Steuerle, PhD
  • Audrey Burwell, OMH
  • Dale Hitchcock, ASPE
  • Nancy Breen, PhD, NCI, NIH
  • Leslie Cooper, PhD, NIDA, NIH
  • Miryam Granthon, OPHS
  • Jacqueline Lucas, NCHS, (By telephone)
  • Edna Paisano, HIS
  • Vicky Mays


  • Call to Order: – Donald Steinwachs, PhD, Chairman
  • Welcome and Introductions
  • Discussion of Populations Subcommittee draft report on racial and ethnic data needs

P R O C E E D I N G S 9:07 AM

DR. STEINWACHS: Okay, why don’t we get started? For Carol’s sake and I guess for the recording’s sake we ought to go around and introduce ourselves, and this is a meeting of the Subcommittee on Populations of the National Committee on Vital and Health Statistics, and our agenda today concerns review and hopefully approval of the report on populations issues and also discussion of future agenda items.

I am Don Steinwachs of Johns Hopkins University, and let me turn to my left.

DR. CARR: I am Justine Carr, a member of the Committee and the Subcommittee and I am from Beth-Israel in Boston.

MR. LOCALIO: I am Russell Localio, a member of the Subcommittee. I am from the University of Pennsylvania School of Medicine.

MS. PAISANO: I am Edna Paisano from Indian Health Service.

DR. MAYS: Vicky Mays, University of California, Los Angeles, consultant, previous member.

DR. BREEN: I am Nancy Breen, National Cancer Institute.

MS. ROGERS: I am Ann Rogers. I am a writer helping to develop the report.

MS. MENDEZ: Good morning, everyone. My name is Miria Mendez. I am with Office of Minority Health and a staff person to the Populations Subcommittee.

DR. SCANLON: I am Bill Scanlon. I am a member of the Committee and a member of the Subcommittee.

MS. JACKSON: Debbie Jackson, National Center for Health Statistics, Committee staff.

DR. STEINWACHS: Okay, welcome, everyone. I want to particularly thank, Carol, do you want to introduce yourself?

MS. MC CALL: Carol McCall with Humana. I run our Center for Health Metrics.

DR. STEINWACHS: And are you in Chicago?

MS. MC CALL: Yes, I am in Chicago today.

DR. STEINWACHS: Is it beautiful weather in Chicago?

MS. MC CALL: No, not at all, cold and raining.

DR. STEINWACHS: You should have come here. It is chilly, but not raining.

I really want to thank the Committee. Vicky has continue to move our very important report and recommendations ahead and this has been a lot of work, and I am reminded how busy Vicky’s life is because every time we have interchanges she is leaving town for another meeting, trying to fit this in.

I, also, want to thank Ann Rogers who working with Vicky has been trying to make this happen and all the staff and I know Miriam has been particularly involved in this aspect of the Committee’s work and we very appreciate that.

So, the goal today is to review the report and I thought maybe, Vicky you have some opening things you would like to say so that we might then have a discussion sort of of general issues and sort of a general discussion of the way the report is laid out and organized and then we will do specifics after that. That is a good strategy, and luckily someone pointed out to me on the agenda here that we ought to have a break around midmorning. So, there may be a moment at which you can relax.

So, Vicky?

DR. MAYS: Okay, what I was going to do just to start is to say what the guiding principles were, kind of the direction and then to kind of lay it out as a way maybe to work, but that depends upon people’s comments, but I thought it might help us to figure out what to focus on because I know for us what to focus on in terms of helping us to complete it and so that is why I am going to put that out there and then you all can do whatever.

We tried in writing the report to kind of stay away from jargon-type terms. Partially we did that because once you get the appendix that tells you, for example, who it is that came to our hearings you will see that we are better off if what we try to do is to not have a lot of jargon.

So, in that sense as we go through if we add things I think I just want us to just be mindful that our constituents are also a lot of community-based organizations, community policy people.

So, we want to just lay it out in the way in which the report is very useful for a wide range of individuals other than just sending it to the Secretary’s office.

The goal as we talked about it on the phone call in terms of the strategies, we have accepted the recommendations. So, I don’t think we are visiting those today, but in terms of the strategies are they clearly understood to do what by whom? So, if at any point they are, I think we should adjust it. The next is people wanted us to be clear about is this something that requires funding. We don’t want the agencies to have these as the states often end up doing these mandates that are not funded.

So, if it should require additional funding I think we should try to say it so that that is clear and then what was raised on the telephone, and I think, Justine, it may have been you , when you asked about tracking and follow-up; so the question in here would be if a recommendation would benefit from asking that something be tracked, a timetable be inserted or some follow-up, I think that that is very useful so that you can get feedback.

There are some questions that I want to throw out there that for me in going through the report again we had to get a little distance and go through the report again. Some of them were issues throughout, and I think it is something to kind of decide.

Some of this will be new stuff for your agenda or to be dropped or what you but we talk about subpopulations quite a bit. The question is should we really be defining those. I mean I think we may be an inside group, and some people may think subpopulations are just a few things and other people may not realize how broadly we are talking.

So, I put that on the table. Should we keep territories in the report? We kind of used the term but it is very loose. The Committee has sent recommendations up in the past about the territories.

In our hearings with some of the API groups it was very clear that we should continue to go to those places maybe to find out more about that.

DR. STEINWACHS: Vicky and I have a personal agenda to sort of hold meetings out in the Pacific in the territories..

DR. MAYS: So, I just put that on the table as to, and again it is just more an issue of the integrity of it. Have we stepped beyond our boundaries by just kind of quickly throwing in territories or should we take it out?

What do you want to do about the American Indian, Alaskans Native recommendations that we have removed? Do you want to try to work on a letter for the June meeting? Do you want to schedule it for later? I think that should get put on the table. Edna is here. Also would you like to take time to review the recommendations that did not make it into the report? Do you want to check and see if you want to put anything back in?

Do you want to decide what is part of your future agenda on that and again, I think, Don that is up to you.

One of the things that I did consciously remove was the electronic health record pharmacy database that was made as a suggestion before because after reflecting and in preparing the report I actually went back to some of the hearings and the transcripts, etc., and I just wasn’t sure that we have a something, mandate, support, etc., to do it but I think that it would fit very nicely potentially as a letter.

The suggested approach that I would put on the table but of course this is in the absence of having discussion of kind of how the report sits with people is that what would be important is for us to make sure that we have the strategies nailed down because the editing of the other part, if the strategy doesn’t make sense rather than spend a lot of time on the editing of the other part it just, you know, it is almost like if we run out of time we could always do the editing of the other part but the strategy needs to be crisp and clear and understood and then if something is a problem you can task us to write about it and that is just editing but i would hate to leave today not having discussed each strategy and making sure that those kind of three criteria that I talked about have been met for those strategies.

So, for us in trying to finish it having that clear would be I think the best thing. Then the language that accompanies that strategies can be reviewed and then decide if there is anything that is left out. So, that is kind of all I had to say as kind of the beginning.

DR. STEINWACHS: Let us open it up for discussion, I guess. You suggested maybe w start and talk about the strategies. It seems to me then sort of the organization of the report which is around those strategies but really the organization and then move to specifics within that.

So, are there comments or any concerns or issues around the statement of the strategies?

DR. CARR: First, congratulations on just a tremendous reconfiguration and vetting. Even a person such as myself was able to feel very knowledgeable reading it because it taught and it led in a really good way, and so, what I would say in terms of the, it seems like there are two major things and maybe three things we are trying to do. One is we are making the case for data integrity, the data definitions, how you collect it, how you define it and also linkages to sources of that data, and the second thing on data integrity is the addition of the subpopulations, and then the second recommendation has to do with data integration and manipulation and we are making the case that we need to cultivate researchers and get people along and we need to focus on who does the manipulation of the data; is there expertise; do we have the proper resources and again, the linkages; and then the third thing that of course is all towards solving the issue of disparities,

So, I put those out there because it helped me organize. Occasionally the things that will facilitate the second recommendation or things that are left over from the first are appearing in the second, and my one sort of overall recommendation would be that we just try to keep each thing within its category because I think it is very well stated and sometimes repeating it doesn’t add anything. Stating it once it was really great reading it. I just very much appreciate all the work that you did.

DR. MAYS: Thank you. Could I just ask because there is one other piece that maybe it is not coming through or maybe it is like secondary and not primary and that is this role of coordination.

I mean because there are several things that it doesn’t require money; it just really requires that sometimes it is even a committee, but it really requires someone where we are saying at the level of the Secretary to task that committee to pay attention to this and to ask them to get back by a certain point in time. So, if that doesn’t come through —

DR. CARR: I was thinking of it sort of in the range of the first part is data integrity. We know what we are doing. We have the definitions and we have identified all the populations.

I mean part of that is linkages. Like we have this data set linked to that but the other part in terms of the data manipulation and integration I think maybe there we could make the case that it is not going to happen unless we mandate that there is integration. So, we need researchers and folks that are educated. We need to cultivate them. We need to facilitate their research and I think there you could say, “And we need to mandate that it is not siloed but integrated.

MS. MC CALL: I would say that integration is a term that works on two levels. One is at the data level and one is at the work level.

So, we can take the theme of integration and we can say, “You have to integrate the actual sources, but then the work has to have integrity, and the work also has to be integrated so that party A knows what party B is doing.”

DR. STEINWACHS: So, focus more on integration than coordination as a word. Vicky, is that it or not?

DR. MAYS: Yes, I think, well, I don’t know. It is like what do you tell the Secretary to do. We always have these discussions about what is it you tell the Secretary to do. So, you would tell the Secretary —

DR. STEINWACHS: You tell the Secretary carefully.

DR. MAYS: Yes, I know..

MS. MC CALL: I think the term “coordination” for me is a little bit different than integration. Coordination to me implies that they ar actually trying to accomplish something together as opposed to integration or even the word “synthesis.” They need to be aware of each other and so that they take advantage of each other’s results and through that awareness that kind of a composite picture can be prepared for the Secretary.

DR. BREEN: One thing, and I am not sure if it came through in this but I think it needs to is I think what we are asking the Secretary of DHHS to do is to take a leadership role and I noticed we got something by e-mail in DHHS other day.

People who work here may have seen it from Secretary Leavitt saying that he wanted electronic health records and they were going to have a group get together. The public sector and the private sector would be represented. They were going to set standards. They were going to encourage this in various ways. They were going to make interoperable software available publicly, freeware and so that is the kind of thing the Secretary can do, and so I think that is the kind of leadership.

I mean when I read that the other day I was thinking that is kind of what we want them to do for the issues that we are talking about here. So, I think the leadership part is important because integration and coordination is probably going to happen on a little lower level.

DR. CARR: And I think it opens up with saying, “Taking a leadership role,” but I think they are not incompatible, leadership role in setting the expectation that there is integration and coordination.

DR. BREEN: Right. That is what we need to do. What you said is what we need to do, but I think what we are asking the Secretary to do is to take a leading role in doing that.

MS. MC CALL: Right, and I think that is great. I mean there was an announcement yesterday of $50 million being made available with regard to electronic medical records and there was a statement in there that said you know, “Dear Private Industry, you have got 9 months to come up with standards,” and if you don’t I will,” not me, Carol, but me, the government, and so the opportunity to set the bar on some of those things, there is obviously a stepping up to a leadership position but with respect to this work here there is an opportunity to inform those standards as well which is really what the first point that we are talking about is really about.

DR. BREEN: Carol, was there any indication on how that $50 million was going to be distributed? How were they going to make that money work?

MS. MC CALL: The advantage of me being here I will see if I can tease some of that out because I am online.

DR. CARR: Wasn’t that Wheeler’s budget that was first cut and now this was reinstated? That was the same number.

DR. STEINWACHS: I am guessing. Otherwise they will beg, borrow and steal it from different agencies. There are ways to scrape up $50 million.

Are there other, you know, I wanted to continue this discussion sort of on the strategies and so, one is I guess the way we talk about the strategy is to emphasize the leadership role around areas of coordination and integration which gets down to more of the data level.

DR. SCANLON: This has been an education for me and I think I appreciate all you have done to do that and in some respects these are the thoughts of a cold reader coming at it from the outside perspective because the specifics of the recommendations in some respects are within the purview of the Secretary himself or herself and so they can do these things and particularly I guess I reacted to the title versus the body. I mean the title says, “In federal surveys,” and I think in some respects the body is broader. The discussion we have been having right now is much broader and there is a question then sort of if you go broader do you have the support for everything that you are arguing. I mean I think that is a key. I mean sometimes when we were moving away from federal surveys there would be questions that would come up in my mind thinking about again as a cold reader in which you are trying to convince me that we should do something and have you given me enough information, and I think the answer may be sometimes no, you haven’t.

Have you given me enough information about how we should approach this from the context of federal surveys and the use of the federal survey information? Yes. So, if we want to go with this broader context we need to think about what it is going to entail.

In sort of the broad context of a leadership role I think that is much easier to develop and to work out some of the specifics in terms of other kinds of recommendations about they should lean exactly in this direction.

MS. BURWELL: I wanted to speak to the comments just spoken about integration and coordination because from the Office of the Secretary’s perspective there is a mandate now that might not be enforced and this is something the Committee might think about asking the Secretary to revisit and strengthen. In 1907, there is the inclusion policy for race ethnicity and it was a directive to all heads of the operating divisions and staff divisions to collect race ethnicity data in accordance with the revised OMB standards and to use it in the analytic and dissemination processes.

So, it is still in effect. The change in administration has not negated this and this is something we might ask the Secretary to revisit.

In addition to that the data council is operated out of the Office of the Secretary. It is headed by the Assistant Secretary for Policy and Evaluation for Planning, Policy and Evaluation and it is at organizationally one of the highest levels it can get to in the Office of the Secretary.

This Committee, again, might ask the Secretary to strengthen some of the coordinating and integrating relationships that the data council subcommittees have and here is one in point. There is a data strategies committee that is meeting now to work on methodological research for improving data collection and dissemination with the underscore on dissemination because there are tons of data that are collected. It is just that through sampling designs and privacy and confidentiality concerns that these data cannot be released.

There is methodological research under way to try to improve some of these activities and integration, there is a subgroup or working group on data integration strategies that took the National Health Interview Survey some years ago and made it the sampling frame for the rest of the department surveys.

So, there is work under way and the Secretary and the ASPE should be encouraged to strengthen these coordinating roles and integrating roles so that the left hand indeed is working with the right hand when we are trying to accomplish some of these things.

DR. MAYS: May I just comment on that because especially like if you read this report the criticism is it has been there a long time. They even went back and said that the 1999 report that the data council itself did; so, it is almost like part of what I don’t want to do is rubber stamp those things because if the Secretary’s Office is willing to first take a look at whether there is structure and whether they are working properly that would be I think a bigger role for me rather than just saying, “We know you already have it,” because remember we have also had 3 years of letters that have gone up which always get cc’d to the data council and all these other places.

I mean I am sometimes reminded and I mean Simon reminded me of something that I think is important. Sometimes things take a long time to happen, you know, because I am like, “When will our stuff get taken care of?” and he says, “It takes a little time.” This is probably the result of some of the letters, this happening, but I really would like it not to go back to the level of the OMB in the sense of quite often at the federal level what is used is the minimum.

I mean we actually in this keep pushing for it to be beyond just the minimum level. So, it is almost like around here when you say that it is like most of the data that is used is at the level of the two questions, kind of the minimal. So, I think we are kind of asking them to do something a little different, and that is going to be my hope, is that they will look at that committee and say, “Do we have all the right people? Should Census be on it? Should we be talking to the Social Security Administration?”

So I kind of want them to take a bit of a new look, the same group but maybe task it to do something different and maybe add a few new people.

DR. STEINWACHS: Let me also bring us back to Bill’s comments. I think Bill was sort of raising the issue of are there areas where we could broaden the report in terms of going beyond talking about just federal data systems where we certainly are on our soundest footing but Gene?

DR. STEUERLE: As one of these people who came in late after the process I don’t feel like I really want to dramatically change things here but I remember sitting in the Treasury Department when we would get requests to do things. In fact we would even get congressional demands to do studies and then people would just take them. You know we didn’t have the staff and we had a list of 20 studies. We weren’t getting half of those done and there were 10 more in some congressional bill and actually when I got in a position to do something I tried to do something about doing more about it. So, I am as worried about the process as I am about the words in the document because the words ask for a lot of things and I guess what I had said and this is more a practical experience is I wanted people when the Secretary gets this to walk away saying, “Okay here are 50 things I should do, but here are two things I have got to do right now as next steps,” and probably correlated with that is especially actually do these because these are low-hanging fruit or I have got a committee that is going to identify what I need or whatever it is. I am just trying to think of as much, I am as concerned about the process of working back through HHS.

I suppose another analogy I have is I often believe that when you are a consultant to an organization as we sort of are in the end what you are really doing is taking the best ideas that are already within the organization and giving them focus that somewhere or another internally in the organization they can’t get. It is not often that we come in with totally new things. I am sure, Vicky you feel like 90 percent of what you have got in this document is stuff someone in HHS told you they could do or Census or some other place.

So, I am not sure how to play that out. I guess my concern is as much what we do with the document and does it really lead to some one or two steps, whether it is a committee that is set up or some low-hanging fruits or two or three things that we really say need to be done in addition to all the things that we really want to be done. I don’t know whether that makes sense or not but that is really more my concern than even re-editing this.

DR. STEINWACHS: Are there some things here that we can sort of make sure are very much out front and center?

DR. STEUERLE: You want somebody to take any document and remember the Secretary gets whatever he gets, you know 100 across his desk. So, you are lucky if he even reads it. So, it probably goes to some staffer to read it and summarize it and give him some bullet points. So, are those bullet points going to have here are two or three actions you can take?

DR. CARR: The other thought I had was in reading through the eliminating health disparities there is overlap with some but not all of the things and so to the points that were just made if it is fleshed out in detail in the disparities do we want to step up and say as was said there, you know, or to further move this along, this is the strategy we propose.

DR. SCANLON: Maybe I will offer a different alternative to what might have been suggested in terms of expanding because I really think that right now what we have goes beyond surveys and you used the word “federal systems.” Both sort of going beyond federal surveys and using federal systems makes me nervous because particularly we started talking about administrative systems and having come from that world in dealing with Medicare and Medicaid we are talking about a hugely complex undertaking and I sort of would question sort of kind of the feasibility of doing some certain things, and we need to think about those seriously before we recommend them to the Secretary to consider anything like that, but following up on Gene I think it is right. If the list is too long it is going to be completely ignored, and so trying to focus and maybe federal surveys in the context of we have this very big important issue about sort of disparities and we think we can make progress on these particular items within the context of federal surveys and that I think we might be able to be more effective.

Right now it is the issue of the disconnect between some of the points that we have raised about administrative data, about vital statistics that they are not in the context of federal surveys and I mean for the person that is summarizing this for the Secretary they may become bullets and the question is then well if the Secretary feels at a loss or the department feels more at a loss in terms of which one of these things do I give priority to which one of them do I even think about sort of implementing. So, that is actually sort of a narrowing proposal as opposed to expanding, but we have got to have consistency between I think the title and what we set up as expectations and we went off at the start of this morning on a valuable but an even broader discussion about leadership and I think we can have a general concept of that but then I would agree with Gene, we need to be very specific about what we are saying for the immediate steps in terms of being a leader.


DR. MAYS: I just want to remind us what the hearings have covered so that it may be that if there isn’t support there it wasn’t we didn’t hear it; it may be that it isn’t written into it, but we actually had a hearing that invited in the states,the vital statistics people and so it was like we had a hearing that we actually looked at those issues because part of what happened is we started with the population-based surveys and it became clear that there are things that they cannot do and began to move to well, that is the only place we have a problem and we did hear from the biostatistics people. We have heard a bit also about some of the administrative data.

As a matter of fact I think that is where we started. When I first came on we had just finished, and I am looking at Debbie because I remember being given; you think this report was something; I remember being giving something that is like that thick about Medicaid and we did presentations. We actually heard from people very, very early on this issue.

So, there is some material for it. I think the question is whether or not the title is throwing us off and we need to just get better about the title or if indeed and this is where the devil is in the details that we need to actually discuss the administrative side and get a sense of if that is where it is really weak because I will admit I know what my strengths are. I am a survey researcher. So, I mean I could probably do that in my sleep in terms of the writing. It is like it has been my work on the Committee where I have learned a lot more about the use of administrative data and stuff like that. So, it could be that it is my weakness and not the weakness of the you know.

DR. SCANLON: Not that you didn’t have hearings on these different topics but perhaps each one of them in some respects would generate an entire report by itself. That I think is the problem. You can’t build a case and really also kind of knock down some of the counter arguments in the kinds of sort of brief references that we make and that is where we are. I mean most of these things that I am concerned about are that people will immediately think of some potential issues and we haven’t talked about them, and you may have heard about them in your hearings but again they are not here, and I don’t think you want a 150-page report coming out this afternoon and so it is a question of making sure that we identify the importance of the overall question and then if we are going to provide something that is actionable be very specific and very supportive to the greatest extent possible, potentially weaving in sort of in terms of a conclusion a list of other areas that we should think about or we can think about the in the future but not in the sense of thinking about them you know that you are going to act on them today but that these are areas that can be explored for the future.

DR. MAYS: I guess for me I don’t know the level of detail or concern. So, having been the ones at the hearing and having been the ones knowing how some things are moving particularly with the work that AHRQ is doing, and it is too bad they are not here, it would almost be better rather than just saying that we totally take them off the table to visit what the concerns are and then decide to take it off the table as opposed to at this general level to take it off.

DR. SCANLON: I will give you an example. We raised the issue of using claims and potentially having sort of raised the information and we also talk about though in a broader context that we really need to train people to be able to get this kind of information. Okay, the thought that immediately comes to my mind is we have got about a million health care providers in this country who have X million numbers of workers and the question is how do we make this work so that we get good reliable data from that sector and this is actually an issue that came up when I was on the committee before 20 years ago, you know, should we be trying to get claims information to include race and ethnicity information and the question is is there a better way and the issue is through potential linkages to other data sets is there a better way to be able to look at disparities through claims information without having the race and ethnicity on the claim and so it is that kind of set of considerations that we don’t talk about but we raise this issue and i think there is the potential of someone thinking in the terms I just outlined and then they dismiss what we were talking about.

DR. MAYS: If they know what the health plans are doing and I actually talk about some of the feasibility data, some of that is actually about to be answered. I mean there are three different studies that are actually looking at how do you — they are actually going to make the case of how do you ask it, and not only how do you ask it but what do you tell people in order to get them to participate. I mean Aetna just has come out of the field, HRET, Robert Wood Johnson. I mean there are like three or it is actually more than three because the American Hospital Council is going to do something also. So, right now they are poised to be able to answer that and that is I think at least for me it was directing the Secretary to realize that there is quite a bit of work going on and to participate and not keep standing outside, to participate.

I don’t think we have it perfect, but to participate in understanding that those are going on and the leadership role to really keep that process afloat and to bring it just into its own work for the department as well as doing whatever it can to continue to encourage the private sector to also do that but you know, again, it is kind of like I don’t think there is an answer, and I agree that if I were sitting on the other side I would raise it, but I also, in here tried to really point to that the private sector and the department has actually asked AHRQ to be I think a big participant in a lot of this.

So, I think the department is on its way to some of this, but I think it is really complex, also..

DR. STEINWACHS: Let us continue the discussion trying to identify what are broader questions or concerns and then I think we can come back specifically .


DR. BREEN: Yes, I wanted to kind of respond to the discussion that was going on and particularly some comments that were made because I am concerned that we don’t get too narrow with this, and so I think your point, Bill is really well taken that it sounds like we are limiting it to federal surveys and then by the end of the report we are talking about state surveys as well, including those plus administrative data and we don’t talk about the private sector particularly but we would like to see all administrative data collecting this information so that we could build a giant database at some point and that is kind of implicit, I think in what we are talking about, but I guess I feel like maybe we should take federal out of the title rather than limit ourselves to federal surveys for a couple of reasons, and I think the issues of both prioritizing and what is feasible that came up are really important to think about and we should maybe have some narrow recommendations that we feel are the most important.

I thought that the two recommendations that we ended up with were actually, they are very broad recommendations but you know there is a lot that could be done within them and I like that kind of recommendation because where I work and I don’t know where you worked, Gene in the Treasury Department, if you were like the Director or something but as a researcher who works within the National Cancer Institute what we usually do with these reports is if they come to our attention we can use them as mandates to do thing that we feel are important and so we have already, you know, CDC is working with the states for the behavioral risk factor surveillance system. That is not really a federal survey system but it is a collaboration between states and the fed.

The National Cancer Institute and CDC and the Indian Health Service all support the California Health Interview Survey which is a state-based survey run by their Department of Health and Human Services and so in fact NCI has been discussing the possibility of supporting more of those and actively trying to provide technical assistance in order to encourage those kinds of county level large population-based surveys in the states so that we know what is going on in the Mississippi Delta and so that we would know what is going on in Alabama and so that we would know what is going on in these places for which we really don’t have any local data, and so, I think it is important to leave that open and to have the possibility of moving those sorts of agendas forward in this report in addition to having maybe some very specific things that the Department Secretary can pick up on and run with and say, “Look what I did,” like the electronic database stuff.

So, I think it needs to have both those levels because you are right, the Department Secretary is not going to think long hours over this report. It is going to have to be some real quick stuff that he or an aid can pick up on.

DR. STEINWACHS: Vicky is going to be devastated if you keep saying that the Secretary is not going to spend long hours thinking about this report.

DR. BREEN: That is why we are spending such long hours to try to boil it down to something that will catch his attention.

DR. STEINWACHS: Well said.

DR. BREEN: But then there are lots of people in the bureaucracy who will pick up on some of these pieces and I think the coverage right now is really pretty nice. So, I would tweak it but I don’t think I would do a major overhaul.


MR. LOCALIA: I am not sure that I am adding very much except I do agree that the title doesn’t do justice to the content and what I thought we might want to work on when we reach this point in the process today is to try to take pieces of the bold print recommendations and put those in the title.

One point I want to stress and that is people have talked about federal initiatives versus state, local, private in terms of the sources of this information but a nation’s infrastructure goes across all of those and if you use that term or a similar term you an embrace all of the pieces in the report that we have touched upon because the Federal Government can be instrumental in making sure that the infrastructure works.

So, we have roads which include federal highways, state highways, county roads, town roads and all of that and it is all part of the highway infrastructure and the same thing with railroads.

So, that is my recommendation. Let us look at the bold headings and see what we can do to get at least the spirit of those into the title so that it covers it.

DR. STEINWACHS: Any other broad discussion? Vicky had raised, Edna that we currently do not have the American Indian recommendations here and I guess just in the back of my mind I thought that maybe that was better to leave that for a separate letter and to work on that but not on our future agenda but I just wanted to make sure because I know you weren’t able to participate in the call when we were talking about what was in the report or what had been part of the report and is not in it.

MS. PAISANO: I think at some time there would need to be another letter or something. When I read this draft of the report I thought it was very well written and it picked up some of the ideas that came across from the hearing on American Indians and Alaskan Natives but there was a couple of things that because of the trust responsibility with tribal governments an important piece is the consultation and I think again the Secretary of HHS took the leadership in relating that message to other agencies as well in work that they do.

DR. STEINWACHS: Do you want to sort of take the next step and start going into more details? The title may be the first thing and strategy and then — yes, Vicky?

DR. MAYS: I can tell you what our old title was which somehow got switched off and maybe it fits. It was Health Data Needs of Racial and Ethnic Minorities in the United States and Territorial Regions. So, it is just very broad and then I think in one of the revisions we got very specific.

Then we had kind of a super heading which I think we were going to get rid of which was that we weren’t sure where disparities were going. So, we had a super heading that said, “Determining Health Disparities in Racial and Ethnic Minorities,” and we decided, I think that was kind of when we thought that disparities was like a moving target. So, we took it off and it was going to be just health data needs of racial and ethnic minorities in the United States and territorial regions. So, that may be too broad but that was the other end of the title.

DR. STEUERLE: I like this title without the whole bottom line.

DR. BREEN: What about just deleting federal?

DR. STEUERLE: Again, this is great. I think a tremendous amount of the administrative data —

DR. BREEN: Not population based?

DR. STEUERLE: This is from my other background more or less in this area but there is a tremendous amount of information in the administrative data that basically is already collected. It is there and it is just simply wasted. There are lots of issues in privacy and other things as to how you get access to it. So, that is the reason I don’t want to say surveys.

DR. BREEN: May I propose, when I did the exercise that Russ suggested we do and he might have done it differently, do you mind if I do this, Russ? So, I took some verbiage from some of the two recommendations and came up with strengthening the capacity of the health statistics infrastructure to enhance the quality, reliability and completeness of racial, ethnic and linguistic data in the US.

DR. STEINWACHS: Would it help to put the ones up on the screen? I have a hard time keeping them all in mind and so we have one idea to slice off the bottom of the current which you said, Gene, take the population surveys off of that and take the current.

The other, Nancy really brings in lots of the elements of the two recommendations. I can handle two bullets at one time. The third bullet comes and it is all gone. I am back to scratch again. Gene is much better than I am. I can see that.

DR. CARR: Is linguistics, does that word definition mean the same as language?

DR. STEINWACHS: You are talking to an engineer. You are in real trouble now. Ask someone here. Does linguistics mean the same as language?

DR. CARR: I thought linguistics was a different kind of discipline.

DR. MAYS: My understanding of it is that some people use the term which I have used interchangeably as primary language which actually is what you are specifically asking the question about and then when you talk about like linguistic minorities or linguistic populations what you are talking about is not just a person is speaking a different language but it also gets a little bit broader in the sense of the choices you make about how you talk may be different than that of the mainstream. So, linguistic becomes broader.

DR. STEINWACHS: It is how you use the language, too.

DR. MAYS: Yes. So, it is like when we get into race and ethnicity and one group says, “What are you talking about? I already gave you my race when I gave you my ethnicity.”

MS. RODGERS: So, are you suggesting that we use only one of those terms or that we define them so that —

DR. CARR: No. They have sort of different meanings. The second definition really is with linguistics and language is more the primary languages, but I mean I suppose the second one —

DR. MAYS: I think if we are saying what question to ask it may be primary language. If you are saying the population it probably is a linguistic population, a population with linguistic needs or something like that but I think we can pay attention to how we have used that.

DR. STEINWACHS: Okay, so, we have sort of two in the table. One is the current collecting data on race,ethnicity and measuring health disparities, period. The other is drawing out the infrastructure to enhance reliability completeness of racial, ethnic and linguistic data in the US.

Any other suggestions. We don’t have to settle this right now, but I thought it was a way to help us sort of set the parameters and the boundary of the report, too. I think that was part of this discussion was you know how broad as well as when we say that we are trying to communicate the title is a good place to start.

MS. MC CALL: Do we take up the issue of geography when we talk about territory?

DR. STEINWACHS: Vicky, Carol was asking about whether we sort of pick up the issue of geography when we talk about the US and territories.

MS. MC CALL: And I am really speaking about you know geographic disparities.

DR. SCANLON: From the report what we are talking about is at the local level and the state level as well but sort of community or state in terms of influence. So, territory doesn’t really capture that.

MS. MC CALL: The question is really the lack of data is really around the area that we have talked about and so that is why they are called out in the heading here. Within the specifics, if we think that we have expressed our desire adequately to understand these at a very local level then we can leave it out of the title.

DR. MAYS: Yes, that was one of the issues I raised is we actually had in the past a hearing, not us but before my time and there is a report on the territories and at first what we were trying to do was go back and capture recommendations that have been made and that they would never address and there is a whole report on, I mean you think the infrastructure needs we have here are an issue. If you read the territories report I mean they don’t even get invited to meetings where we have discussions about some of these issues because there are no resources. You know you are talking about places where the infrastructure sometimes needs a lot of shoring up but I don’t think, I raised it because I just don’t think we have discussed it well enough and that maybe we should consider taking it out.

I mean I hate to leave the territories out but at the same time have we done an adequate job? I don’t think so.

DR. STEINWACHS: So, to take the second version and say, “Health data needs of racial and ethnic minorities in the US,” and just leave it at that?

DR. MAYS: Except we don’t use the word “minorities.” So, it would have to be racial and ethnic supopulations, probably.

MR. LOCALIA: Does the US not include territories/

DR. STEINWACHS: You are being not that specific and I think it is a, I don’t know if by saying the US you are necessarily in people’s minds excluding territories even though technically the US probably —

DR. MAYS: I think you do in most people’s minds in terms of what they understand of geography.

DR. BREEN: As I look as those two titles I think that capacity and infrastructure are going to capture the imagination of this Administration much more than needs.

DR. MAYS: It is the regular title with the bottom off because I think the collecting data we might want to consider that. So, it is collecting data on race and ethnicity and measuring health disparities.

DR. STEINWACHS: Okay, so, we have three options.

You are trying to get Gene’s up. Gene’s was collecting data on race and ethnicity. Gene, you can see I am advocate here. I am supporting you.

MS. MC CALL: Was the recommendation to flip the order of these?

DR. STEUERLE: Carol, I think that one recommendation was to just drop off the in federal population based surveys off the current title. That was what Gene had put forward and then there was the recommendation which you heard that tried to pull words out of the two recommendations in the report and said, “Strengthen the capacity of the health statistics infrastructure to enhance the quality, reliability and completeness of racial, ethnic and linguistic data in the US,” and so that was really, it has all the right words in it; it is just too long probably. We could probably shorten it.

MS. MC CALL: The thing that I am struck by is that as we try to kind of reword the title using words from each of these strategies is that to me both of these are strategies but they are not necessarily the objective. So, if you found that these strategies are followed they are strategies in order to actually accomplish what, and so some of those words actually need to get into the title.

DR. BREEN: I think the third one captures the objectives best.

DR. STEINWACHS: The third one is collecting data on race and ethnicity and measuring health disparities.

MS. MC CALL: Collecting data is a strategy. That is what we want to do but it doesn’t say why we want to do it.

DR. SCANLON: We, also, talk about using. I guess I am wondering if we took the first one and just said, “Strengthening racial, ethnic and linguistic data in the US,” because we deal with collection. We deal with dissemination. We deal with analysis.

DR. STEINWACHS: Okay, so this is strengthening the quality, reliability and completeness.

DR. SCANLON: No, strengthening racial, ethnic and linguistic data, health-related racial and ethnic and linguistic data in the US.

DR. BREEN: If you just cut down to the chase you could get rid of most of it.

MS. MC CALL: Then we say we do all these things in order to strengthen our understanding and insight into those things. That is why we do it so that we can actually change it.

DR. STEINWACHS: Okay, let us see if we captured Carol’s idea. Okay, Carol, why don’t you say it again?

MS. MC CALL: I said that the reason why we do it is really to increase our understanding and insight into the disparities that exist especially along these dimensions but that is why we do it so that we can take action on those and reduce those disparities and kind of raise the tide for everybody. That is why we do it. In order to do it yes we have to collect the data. We have to standardize. We have to analyze it. We have to synchronize it, all those things.

DR. STEINWACHS: So, maybe a strategy here, I don’t know if this works for others is strengthening racial, ethnic and linguistic data in the US and put a colon and then say, “Enhancing our understanding of disparities.” I was trying to get, because I thought Carol’s idea, you know, what is the goal for doing this, and then the strategy is to strengthen.

MS. MC CALL: Right.

DR. STEINWACHS: I am looking at my partners here and looking for the right word so that the goal is, what is a crisp statement of our goal?

DR. MAYS: Eliminate or reduce rather than just understanding. That is the Secretary’s goal.

DR. STEINWACHS: Okay, well, then I think the Secretary’s goal should be our goal.

MS. MC CALL: Absolutely.


MS. RODGERS: How about strengthening racial, ethnic and linguistic health data in the US to reduce health disparities?

DR. STEINWACHS: Or to eliminate health disparities. Okay,love it. This group is good. Carol, you function well on the telephone, better than I by a long shot.

DR. MAYS: I know I just said that we needed a web cam for her to see it.

DR. STEINWACHS: Many times on the other end of the telephone my mind is off wandering someplace. Okay, so the current working one now is strengthening racial, ethnic and linguistic health data in the US to eliminate health disparities.

MS. MC CALL: What is linguistic health data?

DR. MAYS: We can get rid of health now.

DR. STEINWACHS: I think we can get rid of health and it is strengthening racial, ethnic and linguistic data in the United States to eliminate health disparities. Does that talk to goal and strategy?

DR. MAYS: We are asking for income data.

DR. STEINWACHS: It is all health related.

DR. MAYS: We want it all. We think it is all related to health.

DR. STEUERLE: May I ask as an economist whether people are really comfortable with the word “eliminate”? I mean I don’t think you ever eliminate. I would be much more comfortable with a word like “reduce.”

PARTICIPANT: Economists don’t believe we can eliminate poverty.

DR. STEUERLE: It is worse than that. I think actually as costs go up you can actually increase redistribution in society and have greater disparities in health because it just gets harder and harder to make it equal the bigger the system is.

DR. COOPER: But isn’t that the ultimate goal, I mean even if we don’t reach it?

DR. STEINWACHS: It is the vision statement. It is the big vision.

DR. STEUERLE: I guess my fear is that it can lead you to not hit targets of opportunity.

MS. GRANTHON: If I might just share a minute of just a little bit of past experience with this specific issue of eliminating versus reducing health disparities for like 9 years I worked in the Office of Disease Prevention and Health Promotion on the Healthy People Initiative and from 2000 to 2010 we changed that second goal from reducing to eliminate health disparities and we had public comment periods and you are correct in your statement that there were both sides of the coin. There are some objectives that you really are not going to be able to eliminate but the goal was, and sort of what we were reaching out to, and I would be happy to share some of those comments. It is going to the Department, too, in terms of this is going to the Secretary and so I think it makes sense here.

DR. COOPER: Would it be possible to state, “Reduce and eliminate”? Reduce is a marker moving towards elimination.

DR. STEUERLE: I am okay.

DR. CARR: But there is another part to this. Are we assuming that by strengthening racial and ethnic and linguistic data that is all it takes to eliminate disparities? So, I think that is where it is wrong. That component will reduce and the sum of the many different initiatives will eliminate.

DR. BREEN: No, it won’t even reduce. It won’t do anything. It has to be applied which is always the problem with the connector between those two but I mean this is another discussion that I think we have all had in many different venues is how do you link that. I mean we have discussed this in committee quite a lot. How do you link that because our ultimate goal is eliminating health disparities. We are doing one piece which is just really about monitoring.

DR. MAYS: But then can you say —

DR. BREEN: And targeting and identifying.

DR. MAYS: Can you then say, “Racial, ethnic,” and I think we have to deal with linguistic data as a strategy to eliminate health disparities in the US? Could you do that?

DR. COOPER: Or as a tool.

DR. STEINWACHS: We have to take this as far as is useful. I recognize that —

MS. MC CALL: As a tool to identify opportunities to eliminate.

DR. GRANTHON: It is pretty long though.

MR. LOCALIA: I think that strengthening the data is not in itself going to eliminate health disparities if that can be done or should be done.

I think it is fine because it is a necessary but not sufficient condition and I think you assume that this —

DR. SCANLON: You could deal with it by bringing back Don’s colon and changing it to saying, “Eliminating health disparities” is the first part, colon, strengthening racial, ethnic and linguistic data in the US.

DR. STEINWACHS: Do people like that, eliminating health disparities, colon?

Okay, Bill, you are being promoted.

DR. STEUERLE: Not to beat a dead horse but I think Justine is right about linguistics. It just looks funny there.

If you are trying to say that there are linguistic groups or something, but we are not really collecting linguistic data.

DR. STEINWACHS: Okay, we don’t want to lose you just over the title. Let us try to settle does linguistics need to be in the title here as different from I guess because I think what you are saying is that it doesn’t communicate very clearly what we are doing and so it needs more explanation and does show up in the text but —

DR. MAYS: My suggestion is that we seek expert consultation on that in terms of what the best word is.

DR. STEINWACHS: So, whether it is language or linguistics we ought to keep it in the, I am just trying to get a sense of keep it in the title or if we can cut the title further to strengthening racial and ethnic data, is that sufficient?

MS. MC CALL: One option might be to consider using the words “and related data.”

MR. LOCALIA: If you move data after strengthening and put strengthening data on and you would want to say, “Race, ethnicity and language in the US.” We have been using the term “linguistic.” But I think you want to preserve the term “language.”


MR. LOCALIA: There was an awful lot of that in the hearings.

DR. STEINWACHS: Strengthening data on race, ethnicity and —

MR. LOCALIA: Is it language groups you are after?

DR. BREEN: No, we are trying to find out what language people speak because a lot of people are linguistically isolated. They may not be getting quality of care or access to care.

DR. STEUERLE: You are trying to identify their language group.

DR. MAYS: I was going to say in analysis you are also going to want to pull them out because remember some surveys talk about asking not just do you speak, how much you speak English; who else in the household does; do you need translation? So if we say what we are going after it is a little more than just a question of what is your primary language but in the health care encounter we are really asking what do you need.

DR. STEUERLE: It is your ability to communicate effectively around health.

DR. MAYS: Exactly.

DR. STEINWACHS: Unless there are some strong suggestions at this point it seems to me we have got a title that does the things we talked about. One, it states clearly what the goal is. Two, it talks about the key elements of the strategy that come up in the strategy recommendations and I think part of the discussion has been helpful in at least thinking about what the delimiters of the report are or not.

DR. MAYS: Carol had a point but I don’t know, I may have lost where it went about this “and other related data.” What was the point?

DR. STEINWACHS: I think Carol was trying to help us with the linguistics.

MS. MC CALL: There was another thought in there as well which is to say, you know, the things that come to mind are not just the languages people speak but also their educational levels, all of the barriers that are going to prevent this access to care or even comprehending what needs to be done. That includes geography. It includes income. It includes education, race, ethnicity, language. There are lots of things and so we are only really calling out three and so I don’t know if we just want to by calling them out, we really frame the reader on those things and are we comfortable with that?

DR. STEINWACHS: Okay, let me let others respond to that because this really does as Carol says, focus on three elements here.

MR. LOCALIA: Can we enumerate the other elements we want in it? The length of the title is not an issue, socioeconomic position, ethnography.

DR. MAYS: Or geography or place, education.

DR. BREEN: Socioeconomic position —

DR. MAYS: That is right and place. It would almost be if we said, “What are the others?” Oh, there are issues about nativity, country of origin and racial identity. Those are the other things that are in there.

MR. LOCALIA: Language could stand for country of origin. Racial identity —

DR. BREEN: The other whole category is access to care, health insurance status, usual source of care, those sorts of things.

MR. LOCALIA: Wouldn’t socioeconomic position embrace that?

DR. BREEN: Usually they are distinguished.

MS. MADANS: You are changing this from race to anything related to health. I thought the focus was on race and things that would be related to race and their effect on health which is why if you are talking about disparities you have to talk about groups. So, I thought the original focus was on the race, ethnic subgroups and other topics that might help you explain those disparities. I don’t think you want to lose the focus on race, ethnic. Otherwise it is just another report on things related to health.

DR. MAYS: I think what we did in the very first hearing was to ask the question because that is what we were trying not to have people to do is to just rely on race and ethnicity. So, we asked the question at the first hearing what other variables do you think should be collected.

MS. MADANS; But in the context of race. Otherwise race, ethnicity becomes one other variable related to health.

DR. MAYS: Oh, I agree. Yes, that part I agree with, but I think in terms of the reader that is kind of like you need to be pretty well knowledgeable about the issue.

DR. BREEN: We had a discussion about this. I don’t remember exactly when it was but in the last couple of months and I think we decided that we were keen to collect all of this information and to improve it in our data collection set but that we should focus, this Committee should focus on race, ethnicity and language and it really was more an issue of feasibility than anything and again it was to focus and kind of what we were talking about before that if the Secretary sees a bazillion things he is less likely to do any than if he sees what looks like a doable set that everyone had a concern about.

DR. STEINWACHS: Unless there are very strong feelings why don’t we go with this. We can always come back to it again but I think we need to start moving on. So, it is eliminate health disparities, colon, strengthen data on race, ethnicity and language in the US.

DR. MAYS: It is like data collection on race, ethnicity. We can move on.

DR. STEINWACHS: It is very near our scheduled break time. I thought that maybe what would make sense is to let you get up for 10 minutes, stretch a little bit and get some coffee and then let us come back and start into the meat of the report since we have now settled a little bit of the discussion around scope and try also to have a succinct communication of what the report is about.

(Brief recess.)

DR. STEINWACHS: Let me ask your advice on how you want to do it? Shall we just sort of walk through starting with the introduction and then going on there? I think that if the suggestions you have are editorial the easiest thing to do would be to give those to Ann and to Vicky.

What I am looking for are message issues since we now have had quite a bit of discussion around the focus, trying to make sure that we are in the introduction and as we go through the two strategy recommendations that we are defining what the report needs to cover.

So, any comments on the introduction section?

DR. SCANLON: May I ask a question and this is again the cold reader speaking? The secure data centers have bothered me since I read the first draft in the sense of I don’t really understand sort of how they are going to work and they are mentioned on Page 3 but when I get further into the back I have kind of had this concept of people being able to go and use things there but to me it is not just geographic location that can be the barrier.

It is a whole set of rules that you have regarding access to this information that is the real barrier and so again as a cold reader when I see it I don’t quite understand the concept.

When I see it later and I understand it a little bit better I am still concerned that it doesn’t really solve much of the problem. I don’t know where the Census Data Center is.

Maybe it is out in Suitland but they are not going to let me in regardless, and so what it does in terms of accomplishing something I am not sure.

DR. MAYS: Secure data center typically is one in which you get to access the data at even finer levels. You, for example, in the secure data center might be able to put together variables that you wouldn’t outside.

So, you actually do get place or geography and for these small populations what probably would happen is that they are not available for you to use in the public data use but in the secure data center I might be able to go in and do some analysis on all of the American Indians or I could do it on a subpopulation and typically there are rules and those rules are not in the data center such that I can’t get access but it is like the rules typically are things like I can’t walk out with certain printouts.

You usually give them like dummy setups and they are helping you to run those things. There is a person or a process that makes sure that you will not be able to identify because you are now in a place where there is a greater likelihood that you could identify the person.

So you run things but there is someone or a way in which the program prevents it from saying, “Oh, that is Vicky who lives over in Suitland, Maryland” or something like that because you know I was on TV because I had HIV/AIDS or something and then you suddenly could identify me.

So, I think usually it is not a barrier. I thought Russ put in to make sure that there was adequate software and hardware and stuff like that but usually you get an even better access to the data in secured data centers.

DR. SCANLON: I understand the objective. It is the question of how can we achieve this and also I am still wondering whether we can do it.

I mean how much of it is a person looking over your shoulder as opposed to the software could control what a person accesses prevents you printing things out at your local site. I mean it would be so much more feasible for the department to think about doing this if it wasn’t thinking about having to do it across the country as opposed to developing a capacity that would have secure access.

MR. HITCHCOCK: There are levels. I could talk about this a little. I wish Deborah was here but she had a conference call she had to go on but at NCHS where I worked for a while and can speak for a little bit there is a level of access, remote access that you can get. It doesn’t give you everything that being there in person would allow you to do. I don’t know the exact requirements or criteria that is used but I know there are levels. One is remote and one is —

DR. SCANLON; I guess I am thinking we can raise the access level that is feasible remotely because we can develop the software that is going to be effective and we can also in the process widen the potential set of users. What we are talking about, you know, you can go in or anything like that. I mean for me there is this question of who is the you. If I show up at NCHS what are they going to let me do and —

DR. BREEN: I can address that because we do some of this at NCI and have one of these data centers and of course NCHS, I have used their data center before and typically what you have to do is write a proposal and if you have a funded proposal you are in a much stronger position to be allowed to use these data but you have to show —

DR. SCANLON: A federally funded proposal?

DR. BREEN: It doesn’t matter what the funding is so much as if it is funded. In other words it has been peer reviewed. Somebody else has looked at it and said that this is a reasonable piece of research and this is a reasonable researcher because essentially they are giving you, you know they are trusting you and so they want some evidence that they can trust you and that you are not going to reveal confidential data because that is the bottom line here. So, you would have to show that you are a bona fide researcher with a research plan in order to be allowed to use the data and then you are either given the data to use at your own site or you gain access to the data center and so that is typically how that is done.

Oh, one other thing though would you suggest that we recommend that DHHS try to develop some sort of software fire walls kinds of things that would make this available remotely? I don’t think this exists at this point but it is not out of the question.

DR. SCANLON: From the perspective of where technology is going I think it makes some sense. I had a colleague 2 days ago make a phone call to someone three blocks away and it was routed through India and so the corporate sector is clearly taking advantage of certain things and I think we should think about it in this context sort of as well.

DR. STEINWACHS: Who else is there on the phone?

MS. LUCAS: Jackie Lucas.



DR. STEINWACHS: Oh, great.

MS. LUCAS: I was just going to say I don’t know if someone there has a laptop and Internet access? You can pull up the NCHS web site and it has a considerable amount of information on there about the procedures for remote access.

DR. BREEN: But it is as I described it, isn’t it, Jackie? I mean that is what we have always had to do to gain access.

MS. LUCAS: About the submitting proposals and all that, yes.

DR. STEUERLE: I am just wondering there are all sorts of strategies for getting more people access to data using them. I mean I was an adviser to the Statistics of Income Division at IRS for years and every year we had the same discussion about how can we get more people access to the data because it is confidential text data and what seems to make the difference is not the complaint. It is whether somebody internal to the organization decides that either personally or some other way they get some way of figuring out how to get it done and sometimes it is creating remote access. Sometimes it is deciding we are going to double the number of whatever IPAs we have or somebody else deciding that the IPA definition doesn’t work because they have some IPA definitions where if you are a university professor you can get in, but if you are a university professor emeritus you can’t, I mean just all sorts of weird things. So, somebody determined that worries about how — and we could sit around as a Committee. We all perhaps have our war stories here but it seems to me the question is is there someone within the agencies and maybe in some of these agencies there are and in some there aren’t who actually has as a goal as part of the job description and part of the measure of his or her success that they are trying to figure out ways to get more people access and in some cases the data are too confidential. You will never get remote access. So, we all have our different little strategies. I am just worrying at the end of the day and we all I think have had this experience that we think there is not enough of the data being available to people who could really make good use of it. So, it really takes somebody internal to the organization to really care enough and to have a reward structure that says, even if means nothing more than getting from five to 10 people inside the organization sometimes to do research, sometimes that is a major advance even though we may think the whole world should have access. That may be quite measurable and so I don’t want to word craft here but I think the language should be around trying to assure that someone in the organization is charge of trying to figure out how to deal with these issues and show and for whom there is actually some type of internal reward for succeeding.

DR. MAYS: What I was going to say is that there is a difference in levels of access between remote typically and on site and a lot of researchers are more interested in the ability to do on site, particularly if you have a funded proposal, particularly if you have, you know part of what we are saying is this needs to be closer to you.

So, you would really want people to go in I think but I was just going to say and I get a little nervous if we do a big thing about saying that it should be remote because then you are going to get — it doesn’t maybe solve all the problems. I think increasing remote is good but not, I just would freak if it is to the exclusion of the other. So, part of what I think we may want to do is like try to suggest ways.

You know how we have well, here is the issue; here is the strategy and then here is something about it is that maybe that is helpful and I think the things that Gene just said would help us to say you know if we could identify a person for whom it becomes their task to increase this and can we then monitor these increases in the use. In addition the use of technology, increased access, which would mean the development of software if we could like maybe say a few things like that then I think that that might help if that sounds reasonable to people.

MR. LOCALIA: Everything that people have said here I kind of agree with. So, Wednesday I was trying to figure out how to wrestle with this. I am the person who inserted this No. 3 on Page 19. A little bit of history. In early March NCHS had that request to comment on its proposed rules and regulations and I actually wrote extensive comments but I don’t have access to my e-mail. So, I can’t bring them up but in those comments I raised all the possibilities I think that people have talked about here, all of them, different levels of access depending on the purpose because more recent statutes have distinguished between access to data for statistical purposes versus access to data for non-statistical purposes and statistical purposes are defined in more recent legislation as the kind of things that people do when they are looking at health disparities on race, ethnicity, language and that kind of thing rather than going out and trying to find out where is the best place to send your junk mail flyers. So, a commercial purpose is very different from a statistical purpose.

Now, the only reason I inserted paragraph to be as brief as it was is that when I wrote these comments there were a lot of hoots and hollers I understand. So, I am not sure how much you want to wave this red flag other than consider additional options and that could possibly be a starting point for a lot of discussion on a lot of options that have been mentioned here. I am very much in favor of that and I also know a lot of people out and about are very much interested in that. So, there are some technological innovations that may enhance access to data that is national synthetic data but that will not work with local data because those synthetic data will not work with small groupings.

So, I would say let us keep it kind of brief but then that could be a lead in to another set of discussions later and we also have to talk to the Privacy Subcommittee on that and we can open that up.

So, I don’t want to put a whole lot of things here and then dig our graves and not enable us because somebody may come and just say, “Nein.”

MR. HITCHCOCK: It is sort of becoming apparent that the agencies have wildly different policies about what sorts of data they can release and I think at some point I think the data council is going to start to take a look at this but it wouldn’t hurt in this report to make a recommendation that the department sort of moves towards depending on the legislative requirements —

MR. LOCALIA: I think that is a great idea and actually in one of my e-mail comments and I think it was at the end of February I did say that there were inconsistencies across different statutes and agencies that did not seem to be consistent with the underlying purposes. So, if there is an underlying purpose we ought to be consistent but it is not. So, I can get a lot of very precise information that flies under the HIPAA radar from hospital but I can’t get it from other sources, but I don’t know how much you want to say specifically here, given that I already tried to cover that in a previous e-mail.

DR. BREEN: One thing that I liked that paragraph that Russ inserted. I really liked Russ’ comments in general, but the only thing I would recommend as a friendly amendment was instead of providing access to explicit say, “Increasing access,” because that is really our goal and then that gets around the issue that Bill brought up of what exactly are we asking for. Well, we want an improvement.

DR. MAYS: I am trying to figure out how to capture this because I don’t want to walk away and having changed things and then we have a brand new text that people haven’t seen. So, is it possible to think about fixing a few things at some point and the being able to put it up like, you know it would be great to make sure from Dale that we get this issue kind of how we want to say this comment about the department, I started writing, “Moves towards examination of statutes and policies of various agencies.” It is almost like if we could kind of have a time where we could do that and then put them back up it would be great.

DR. STEINWACHS: Why don’t we try to get from here to our lunch break with trying to identify things that may be with more substantial changes and that ought to have a chance to go up and so we can see them so we don’t find ourselves going back to basic reconsideration.

DR. MAYS: Because this is more than word smithing.

DR. STEINWACHS: Yes. Okay, that came out of the introduction. Anything else in the introduction before we go to recommendation 1?

Okay, the first recommendation is HHS is urged to enhance the quality, reliability, completeness of data collection, reporting on racial, ethnic and linguistic subpopulations in the United States and territories. Any modifications of that recommendation? Do we want to drop territories or do you want to leave that?

MR. HITCHCOCK: I would leave it.

DR. COOPER: On the first page in that second paragraph where we say, “Blacks, whites and Latinos,” is right where Nancy was saying, “Increase the access.”

DR. STEINWACHS: This is Page 4, second paragraph, they provide useful information —

DR. COOPER; On blacks, whites and Latinos. We don’t need to be consistent with our language with the OMB. Whenever they say, “Latinos” there is supposed to be “Hispanic/Latinos.” We don’t have to say that?

DR. BREEN: Do we have to say, “Blacks/African-Americans”? Do we have to say, “White/Caucasians”?

DR.COOPER: No, but it specifically lays it out in the directive.

DR. MAYS: Someone here can probably speak to the issue but some Hispanics don’t want to be referred to as Latino and vice versa.

DR. STEINWACHS: So, I think this is an excellent suggestion.

DR. MAYS: So, it is Latino/Hispanic or Hispanic/Latino?

DR. COOPER: I think it is Hispanic/Latino.

DR. MAYS: All right. We will go through and take care of that.

DR. STEINWACHS: Anything else? Okay, let me take you to Page 7, recommendation. Anything on Page 7?

DR. BREEN: We talk about, let me see, in the first paragraph it is an introduction. It is right under the recommendation and after the title, eliminating health disparities it says, “The NRC underscored the importance of collecting data on race, ethnicity, socioeconomic position and acculturation.” I thought it said, “Language.” Did it say, “Acculturation”?

DR. STEINWACHS: Okay, we are going to check. Audrey will hold our feet to the fire.

DR. BREEN: The reason I thought it was kind of important is acculturation is quite difficult to measure. Language is more straightforward, and if we are trying to, we don’t want to be overly simple here but if we can be simple I think we are better off.

DR. MAYS: All of you can check but throughout the book they actually used acculturation. I stayed out of the acculturation because I think that there is work that is coming out that is beginning to say that we need to break down that concept of quote, acculturation into all these other subvariables.

So, I was kind of looking towards where we are going in the future.

DR. STEINWACHS: Justine found at least the title, Page 3, of the report said, “Data on race, ethnicity, socioeconomic position and acculturation and language use.” So, I think we ought to get the language in there and then the acculturation and I think it is consistent with the report.

DR. COOPER: May I back up again?

DR. STEINWACHS: Please, Leslie?

DR. COOPER: Page 5, and I am just saying under the terms of the last sentence where we say, “Subpopulation,” is where my discussion will fit. One of the things that seems to continuously come up is that when we look at ethnicity some people are trying to refer to that as race and we are really supposed to be doing race and ethnicity. So, you could be Hispanic/Latino and belong to any of these other racial groups, but that doesn’t seem to always come out. So, I guess part of this is the need for more training so that I mean luckily at NIH because of our progress records to get NIH dollars there is a page there that you have to complete but it is continuously reminding investigators that Hispanic/Latino is not a race. It is only an ethnicity and I don’t know how we could help make it clear because there obviously is a need for training.

The other issue is that when you look at what is in these racial categories there is diversity within those categories and I don’t know how we communicate that you have to be careful in terms of generalizing to a population and you haven’t included that population in what you are studying. I mean a real good example is sometimes in terms of looking at this Hispanic/Latino population where many times they don’t classify the race then they are not including like the Dominicans. We always know that when it comes to the heavier the pigmentation within that category that is more of the disparate populations in terms of having a focus on that and when you get into for example, I can’t say that it is mixed race, but if you have like the Chinese population and you have the darker Chinese and whether that is because they are mixed Chinese or they are from the southern portion, but the hues are not the same and I am sure that the cultural issues and the disparities are not the same, and I don’t really know how to try to provide some additional clarification or training so that those populations are sure to be captured so that we don’t end up really talking about, a lot of people tend to generalize and they are only talking about a white Hispanic population, and I just threw that out.

DR. MAYS: There is a couple of places, and maybe it is not like jumping out and that is when I talk about the state stuff where I talk about the ways in which even in different places people are categorized differently. There is another —

DR. BREEN: It is on Page 11 and 12, I think. I am using the version that Russ made corrections on.

PARTICIPANT: Also, on Page 7.

DR. COOPER: As long as it is in here. I just wanted to bring that out because I know that continues to be — and I know a big part of it is training.

DR. MAYS: There is a different emphasis than what you are saying. What this emphasizes is the issue of classification and what you are saying is the issue of diversity within a classification. So, it is a little different in terms of what the emphasis is here. I mean I think a sentence or two somewhere to emphasize that, you know, this issue of diversity but I think it goes back to my earlier question of do we need to define what we are talking about in terms of subpopulations so that people really understand that we are not talking about different groups of racial and ethnic minorities but we are talking about within those groups whether that would be helpful to expand that just a little bit.

PARTICIPANT: I think it probably would be helpful because the word “subpopulations” gets interpreted at a lot of levels. I think it ought to be clear in the report what level at which we are trying to focus the issue, so, within the categories.

DR. BREEN: I think where it gets really important is that you want to collect the data so that you are providing a maximum flexibility for analysts. So, if somebody comes along and wants to look at Hispanics with black race or distinguish the racial categories within Hispanics or American Indians or Asians or any other group that they could do that with the data that we collect and that we are advocating collecting.

DR. STEINWACHS: Okay, Page 7, again? Eight?

DR. BREEN: Are we still on 7?

DR. STEINWACHS: Seven sounds good.

DR. BREEN: Sometimes we forget that white isn’t race and so I am wondering in the sentence where it says, “Although both OMB and HHS strongly encourage the collection of data on racial and ethnic subgroups few federal systems report data at the level of subgroups particularly subgroups within black, Asian or Pacific Islander, American Indian or Alaskan,” actually it addresses exactly what you are talking about and poor whites, I never know what to do with them because they don’t really get singled out via racial ethnic categorization. So, I wasn’t sure if there was a way to sort of bring that subgrouping in here or how we might be able to do that. Does anybody have any ideas?

DR. COOPER: I think it will come in when you look in terms of socioeconomic.

DR. BREEN: So, leave it until then? So, it doesn’t need to come in here? Okay.

Then I got confused at the bottom of Page 7 in terms of, and it comes up again in the strategies for this section I think, yes, where it is not always clear when you talk about leadership at the Secretary’s level at DHHS. DHSS obviously can do a lot within the, I mean the Secretary can do a lot within the Department and some of these agencies are within the Department and some aren’t and then at another point we talk about coordination and it kind of seems like maybe we are talking about coordination among the different departments but some of the agencies within DHHS are also listed. So, I think we need to kind of sort that out a little bit more if that was the intent that some of these things would be done within DHHS and some of them among the different departments.

DR. STEINWACHS: We ought to make those distinctions very clear.

DR. BREEN: Because otherwise it is going to look silly to them because they know these distinctions well.

DR. SCANLON: And also because even though the Secretary has changed, and Tommy Thompson was very focused on the one department concept and it was the same issue of coordinating with the VA or with anybody else.

DR. BREEN: And it comes up with recommendations or strategies 1 and 3 where I got a little confused, too.

DR. STEINWACHS: Vicky, do you have a solution?

DR. MAYS: What I was trying to do is when talking about methodological things is it was really across; it is not within HHS but it is like when you are trying to understand how do we ask these questions; what is best practices? Then I was trying to shift us to like talk to the Census people and talk to a larger group of people. I think when it is talking about some of these committees or things like that those are things that are within HHS. So, I can understand exactly now what you are saying and it is because it is kind of mushed up and Dale was the one that I think also pointed out like SAMHSA is nowhere here and he and I were having e-mail interchanges about well, there is a couple other people that have these huge data sets. So, that is kind of where I need help I think because I don’t know the federal system too well.

MR. HITCHCOCK: Even just structuring it a little bit differently, Vicky, where you have got that last sentence on Page 7 that talks about other federal agencies such as the US Bureau of the Census and that is fine, SSA; this is fine and then it talks about CMS. That is not. That is you know within the department sort of thing. So, it is almost an organizational sort of way.

DR. MAYS: So, more to say the other agencies, those within and then list and also others across government.

DR. BREEN: Or don’t give examples.

DR. MAYS: I really want them to know which ones to go to.

DR. STEINWACHS: So, I think just splitting those two categories, those within HHS and those that are reaching across government.

DR. MAYS: And it is like SAMHSA, were there any that other people thought were left out?


DR. MAYS: Oh, yes, FDA.

DR. STEINWACHS: Statistics and Income, IRS Statistics and Income.

DR. MAYS: Oh, I like them.

DR. COOPER: CDC is missing.

DR. MAYS: I don’t know whether we say, “CDC” or we say, “NCHS.”

DR. COOPER: NCHS is part of CDC.

DR. MAYS: I know, but I am just saying, “Which do we say?”

DR. STEINWACHS: I would say CDC.

DR. COOPER: And I want to back track. This is just for consistency throughout. I don’t want to refer to anyplace that is specific but whenever we talk about the racial ethnic groups we should probably just use the categories as listed within the directive.

DR. MAYS: Why?

DR. COOPER: Well, for example, right there on Page 7, well, we were just there for a minute, but let me see, data on the level of subgroups that is in that first paragraph, probably around 11.

DR. MAYS: Right.

DR. COOPER: Okay, and right there like where we say, “American Indian and Alaskan Natives,” of course, I immediately started going and counting you know since you pointed out that whites weren’t there and it is only five categories. You go across here and count it and it looks like it is six, and it is six because you have American Indians and Alaskan Natives and it should be or and then as I look right here and so it is just what the exact words are. I mean if we want to encourage other people to be compliant we should probably set the standard.

MS. LUCAS: Along the same lines instead of Pacific Islander you should say, “Native Hawaiian and other Pacific Islanders.”

DR. MAYS: That is right. Sorry.

DR. STEINWACHS: Leslie, you are good. I mean what can I say? Leslie, you are going to get the next Committee assignment. You know this.

DR. MAYS: What I thought you were about to say which is where I was like oh,no is that I don’t have to put every term that is used like when I say, “Black,” I don’t have to put black/African American/” —

DR. COOPER: If you want to use what the standard says, it says, “Black or African American.”

DR. MAYS: But I am saying, are you saying that I need to do that for each one, so, black or African American?

DR. COOPER: Or at the beginning set up a little standard and define it that when you say, “Black,” you are including whatever.

DR. MAYS: Okay, we will have a little footnote starting at Page 7.

DR. STEINWACHS: We do want to toe the line in this group. We are an example. Anything else on this?

Page 8?

DR. BREEN: This is just minor but around six lines down on Page 8 IOM is part of the National Research Council. So, I think we should just use one of them because it seems like we don’t know that.

DR. MAYS: Yes, who published it was, it gets confusing later when you go to the reference list to look and one will be the author, the National Research Council and then the other will be the Institute of Medicine.

MR. HITCHCOCK: I don’t know how that does fit in the organization, published by the National Research Council.

DR. MAYS: You see that says that the publication for that will say, the National Research Council. The publication for that says, the Institute of Medicine.

DR. STEUERLE: How about this? Drop the word “the” and put “its.”

DR. MAYS: How is that?

DR. STEUERLE: Is IOM in NRC, is that what you are saying?

DR. STEINWACHS: IOM is in the National Academy of Sciences. I am trying to now figure out where NRC sits in that. I think that NRC sits in the National Academy of Sciences.

PARTICIPANT: They have changed to the National Academies now and NRC sits according to this —

DR. STEINWACHS: It is under the National Academies. So, it looks like NRC, the Institute of Medicine, Academy of Engineering, Academy of Sciences are all sort of equal level.

DR. BREEN: I almost look at NRC as being the publishing house in this case.

DR. BREEN: So, it is all right to have both of them?

DR. MAYS: Or recent National Academy reports and then to say it, so that we look like we know.

DR. STEINWACHS: Okay, more on Page 8?

MS. LUCAS: I guess it starts on the previous page, on Page 7, but it also is in the first paragraph on Page 8. It talks a lot about linkages between systems and I am wondering are we assuming that people already know about existing database linkages because we don’t mention any of what it says. I am assuming that we are trying to say that we want to foster more of that but there is already stuff there but it is not mentioned here.


DR. STEINWACHS: We are here. We are cogitating.

DR. MAYS: I guess the issue is somewhere in here, I don’t remember anymore, I say something about going to the department’s gateway and I think if you go to the gateway you will see what is existing. So, maybe wherever we first introduce the gateway to say something like it will allow you to see what exists in many of these areas or something, isn’t all this stuff on the gateway?

MR. HITCHCOCK: Yes, all the department stuff certainly and some from other agencies, some from states. It has got something like 7000 entries.

DR. MAYS: So, I would hate to feel that at each point we have to, unless you are feeling like there is not a balance here, to have to say every time, yes, you do this, but we also need you to do more of this. I don’t know.

DR. BREEN: How about if we did something like you know as the gateway notes because I don’t think we want to mention all the examples or any examples for that matter but as the gateway notes a number of data sets have already been linked or combined or whatever term we want to use. However, more of this needs to be done and then leadership is necessary to foster.

DR. MAYS: Okay.

DR. BREEN: And maybe a little more transition just to indicate that you know there is not enough of it and it needs to be more standardized.

DR. MAYS: The only thing I am going to say is that it may not be right here where it goes. What I am kind of reacting to is we ar taking a tone. Sometimes we try to make a point about something, but I get what Jackie is saying but it is just like if we could just have a little flexibility.

DR. STEINWACHS: Yes, flexibility.

DR. MAYS: It will read well. It may be that it goes somewhere else. Ann, would you agree with that?


DR. STEINWACHS: So, Vicky a recommendation to this group is to maintain and not increase our flexibility.

DR. MAYS: It may be that it fits somewhere else because I think there is another place that I think it will fit better.

DR. STEINWACHS: Okay, my job is to try to keep us moving but yet give due consideration to all things.

So, anything more on Page 8?

DR. BREEN: Yes, one more, the second to the last sentence, it reads, “In the meantime, the department could play an important leadership role in ensuring that sufficient attention is paid to establishing a system that will include subpopulation, race, ethnicities designations and training for those who collect these data and I would add, there is a period there, “And we encourage the department to do that.”

DR. SCANLON: This is actually a point I would like to raise. I think this is unfortunately going in the wrong direction because it is ties sort of to this issue of claims as opposed to getting race and ethnicity data in administrative systems which are completely different. I mean talk about Medicare for a second. If I wanted to study this issue in Medicare the approach I would use is get this information on enrollment forms for the program and be able to match with the claims. We have got a billion claims coming in every year now. When the drug benefit starts it is going to go to 2 billion. If we ask all those providers to give us race and ethnicity information and I try to use it as a researcher all I am going to do is find all these conflicts between this provider reporting it this way and this provider reporting it that way and I am going to have to go through and make some decision rules as to how I am going to clean the data. I would rather get it once, get it right. There is the potential for linkages between claims and enrollment records and that would allow us to do good studies, allow us to use the administrative data and the second problem here is there is going to be incredible resistance from providers.

We talk about not only sort of the issue of training for the collection, we are talking about training the people that are going to give the information to convince them that they should be giving it. I mean they should be willing to provide it and so, there are so many things that are barriers here, and there is a better way I think and a much more efficient way to do what we want to accomplish. We have administrative data to be able to be used for this purpose. These last two sentences I think can be kind of recast a bit so that we are supporting the idea of building into administrative systems the ability to do this work without tying it back to this effort under way with the uniform billing committee which is looking at claims because I don’t know where they are,but ultimately I would say that is the wrong approach.

MS. LUCAS: I guess my initial reactions to your comments are I am in violent agreement with what you just said. It is going to be a mess unless we make specific recommendations that around, you know, interoperability, when this information is collected, where it is collected and how it is shared because you don’t want to collect it every time. It is a bad model. It is a bad data architecture because it is just going to invite tremendous amounts of conflict and then you will have truths but no truth. So I don’t know if we want to make a specific recommendation about that or just kind of call out the fact that that must be done, you know, collect it at a central point and kind of like at the point of enrollment or in a membership file, not on every claim.

DR. STEINWACHS: Let me just make sure I am clear? So, the UBO4 is the hospital discharge billing record.

MS. LUCAS: Is it?

DR. STEINWACHS: I think so. I don’t think it is the routine office, but I don’t know. I was trying to make sure of that because we have many times on the hospitalization been collecting it right or wrong, good or bad but that would be very different than every office visit, every prescription.

So, I just wanted to make sure that we were on the — does anyone know for sure whether the UBO4 is a —

DR. MAYS: Carol, can you go online?

DR. STEINWACHS: Carol, Vicky would like you to go online and check out the UBO4.

DR. SCANLON: I mean I think that is an important point. That reduces the burden but it also would leave us short of what our goal would be which is in the administrative system to be able to look at the entire Medicare population including particularly the younger elderly who don’t use hospitals for their first 5 years in the program.

MS. LUCAS: What are we looking for here?

DR. SCANLON: What the UBO4 claim is.

DR. STEINWACHS: Yes, if we could. Let us try, I guess just to make sure of the conversation. Currently the Medicare essentially from Social Security data I guess is fed race as it was collected at the time people got their Social Security cards and so that information exists in the kind of mechanism but it does not meet our current definition of how we would like to collect it and so I agree with you that not everyone gets hospitalized. If you live long enough I guess the probability gets up pretty high, right? So, there is an issue I guess of how do you get information that fits current classifications for the Medicare population and this doesn’t really address that.

DR. SCANLON: And I think sort of moving forward at a minimum what we should be thinking about is when people come at 65 and enroll we should be asking the questions that we want answered and then secondly we should be considering strategies for how do we deal with those who are currently enrolled and in terms of trying to improve the information because I think those might be more effective than asking every provider to do this and particularly since what I came away with and what I learned from reading this report is how much more complicated these questions are than we would normally think about and so as I said we are talking perhaps 2 million people that will be filling out these forms at one point or another and that is a lot of people to train.

DR. STEINWACHS: Maybe the thing to do and maybe what you are suggesting, Bill, is to use the last half of that paragraph that starts now currently with the NUBC and so on but to use that to identify the Medicare problem that needs too be solved.

We don’t have testimony. We don’t have a strategy for the solution I guess.

MS. MC CALL: Right. What I think that we should not do is A, say that we do or B say that somehow we want this to be on every single claim and therefore captured by the care giver. I think that is inviting a real mess and in terms of what would really happen and I think a lot of people know that practically speaking that that is not going to be a viable solution. So, I think we just need to cull that out. It is going to beg issues around kind of unique person identifiers. That is where that is going to go. There needs to be an identification number or something that identifies me uniquely and then there has to be something that is linked all the time. We capture it once and then it is available and it links to basically what happens to me throughout my course of care.

DR. STEINWACHS: So, Carol, you want one identifying number. I want three personally. So, I can use them as I see fit. I am laughing.

MS. MC CALL: Absolutely, but that is where it is going to go I think.

DR. CARR; So, as I am looking through eliminating health disparities the recommendations included in this are very close to what we are recommending. However, their recommendation six one says, “DHHS should require health insurers, hospitals and private medical groups to collect data on race, ethnicity, socioeconomic position, acculturation and language,” and so one thought I had is since we parallel this so closely would it be helpful for us to say that we support everything here but this is where we differ because I think if you read this and then you read this it is similar and this is already in the public domain and I think that if we had a substantial disagreement with this that makes this report value added. It is something that we didn’t already know.

DR. SCANLON: I think in terms of dealing with this paragraph I would start out with talking about the objective which is that we would like to be able to use the administrative data to examine sort of this issue of disparities and to accomplish that we have to have good information on race and ethnicity within those administrative data and then that leaves open the question of exactly how you accomplish that and they are suggesting one approach and I think we are saying that you need to think about whether that is the right approach.

DR. MAYS: Part of what I was going to say is there are points at which I think we do map well with them but it is a little different in terms of maybe how we are going to get there. So, we sent a letter and in the letter, I can’t remember the exact date, but one of the letters we sent we actually went through and said which of our strategies that we thought were kind of supported by their strategies. So, we have tried to say that some of these things have been talked about by them and then kind of do what I thought we were doing but maybe a little bit more uniquely.

I think that in this section, I mean this is why I sent an e-mail like “I need help in this section,” that we probably should, I don’t know. I think maybe we should rewrite this section but again this is where I said that I have only learned probably half of the quality stuff because I get messages sometimes, not all the time. So I don’t know.

DR. STEINWACHS: Why don’t we try, and I would like to keep this moving. I think Bill’s suggestion was to lead off with a brief statement that this really is about the use of administrative data or —

DR. MAYS: I am going to ask that the actual rewriting be done because I don’t know that I have the nuances well enough for this one.

MS. RODGERS: Could I make a suggestion that Bill rewrite just this paragraph because you clearly know the most about it and I think it would be most efficient for you to do it rather than us all to try to do it.

DR. STEINWACHS: Okay, I put Bill’s name by it. I like the strategy, delegation. Okay, I am on Page 9.

Leslie will take me back to Page 5, I know.

Anything on Page 9?

DR. BREEN: I was wondering. I just wasn’t clear in the question about clarification, there is something about a number of important feasibility studies currently under way that can inform efforts about the feasibility of collecting data, complete and accurate data on race, ethnicity and primary language, and then there is mention of the —

DR. STEINWACHS: This is in the private sector?

DR. BREEN: That is exactly what I am wondering. If there is a number of important studies either they should be cited or maybe it refers to the citation that is there on the bottom. I just wasn’t sure.

DR. STEINWACHS: This is the top of Page 9.

DR. BREEN: Yes, it is in the middle of that first paragraph. Collecting such data is the beginning of the line and then it says, “A number of important feasibility studies are currently under way.”

DR. STEINWACHS: So, why don’t we just make sure that I guess that the citation is there, whether that is a citation at the bottom or a citation —

DR. BREEN: No, that is just for one of them,

DR. MAYS: So,you just want me to put the citations for all of these. Actually I have the references that I can insert..

DR. STEINWACHS: I think that would be good. People may be interested in trying to locate them.

DR. MAYS: I have references.

DR. STEINWACHS: Okay, anything else on Page 9?

Page 10, Justine?

DR. CARR: I was trying to diagram the first sentence. It serves me far better now than in the fifth grade and I am better at it, too. So, if I read the first line, so, we are saying that HHS should encourage federal agencies to hold subgroup-specific conferences to focus on best practice methods for collecting, analyzing and disseminating data on health disparities in racial, ethnic subgroup populations as well as for closing the gap in the disparities.

Is this more broad reaching than what our title suggests? I think this is sort of —

PARTICIPANT: It is a heck of a sentence.

DR. CARR: Yes, I can share the diagram with you, but I think this is first of all hard to read but also I think it is more far reaching than what we were saying. Where it should be broken down, our first thing if we go back to our title is we want to hold subgroup meetings on methods for collecting and analyzing. Do we want methods on collecting, analyzing and disseminating data on health disparities and also how to close the gap? I mean that is bigger than what the title suggests.

DR. MAYS: We can drop the closing the gap part.

DR. CARR: Yes, so, take that out and this is where I sort of thought in part one we were talking about let us get good data; part two, let us educate good researchers to integrate and manipulate the data and this is a little bit of both. You are going to have a meeting to collect it, analyze it and then disseminate it.

DR. STEINWACHS: So you think maybe here we ought to be focusing on the collection?

DR. CARR: I think so. I mean it might be the same meeting that you have but this whole section says that we want to have reliable good data and the next one is we want sophisticated data manipulators.

MS. LUCAS: Right, and I think if insist on collection even that might break down into first its capture. We have got to get it on the hook and then we have got to get in the boat.

DR. COOPER: And then you have got to get it out. So, when you go beyond the collection and the analysis it is also in terms of helping to disseminate.

DR. MAYS: But I think it is the difference between which recommendation we are on. The first one I think is about collection but it is the second where the —

DR. STEINWACHS: So, we want to lose this. We want to go to the second.

DR. CARR: Get your solid building blocks and then the second one is use them intelligently. So the analyzing and dissemination would go to part two or recommendation two.

MS. LUCAS: On the collection I guess what I would love to see is us break that into two pieces which again has to do with its capture and kind of where this is going to happen. It goes back to what we were just talking about. Is it every claim and every point of touch or is it at a couple of key points? The first is capture and the second is linking that and so some of that is about infrastructure and things like that, but I can capture it 100 places, but if they are not the same it does the researchers no good.

I do have a crosswalk in front of me if you want to ask specific questions. I have a crosswalk from a UB92 to a UB04.

DR. STEINWACHS: Okay, I guess on the UB04 this is the hospital discharge bill right?


DR. STEINWACHS: And so on UB04 I assume they are capturing race and ethnicity consistent with the OMB directive.

MS. LUCAS: Actually that would be a bad assumption. What they have on patients they had before. They had patient address. They are now breaking that out into specific fields. They had patient’s birth date. That is still there. They have patient’s sex. They eliminated marital status and, ladies and gentlemen, that is it.

DR. STEINWACHS: Okay, because on some of the state discharge abstracts and I thought Medicare was also getting it on their UB91, I thought they were getting race and sort of a simple category. Okay, so, race, ethnicity is not there and it is not there on 04, either is what you are saying.

MS. LUCAS: No, not on the 04.

DR. BREEN: But this says in the next version of the UB04.

DR. MAYS: We had testimony here. I will remind you because I went back and looked. I did some of the quality stuff but there was a person that came and talked to the full Committee and actually on our web site actually it talks about the fact that that was the statement she made is that this change is going to come in the revision of the current UB04. There had been discussion in other places around the fact that the hospital discharge data varies by states in terms of the extent to which race and ethnicity are captured and then Marjorie commented for us also.

DR. CARR: Right. She said that the NUBC is currently working on including the capacity to collect race and ethnicity in the next version of a uniform bill for hospitals UBO4. This action will help establish the business case for using the OMB 1997 guidance standards.

MS. LUCAS: Right. I mean again in an ideal world what would be here is a unique person code. So, nothing would be captured here. There would be a unique person code and that information, race, ethnicity, etc., on that person would be somewhere else and so it is able to be linked.

DR. STEINWACHS: Okay, Bill is working on that section already.

DR. SCANLON: This may be too radical but here is what I would write if I were a poem, okay?


DR. SCANLON: The NCVHS recognizes the potentially important contribution of administrative data such as health services claims may have in identifying and understanding disparities. This contribution will only be realized if reliable, accurate and complete racial and ethnic information is included in administrative health data systems.

The department can play an important leadership role in identifying the most effective means to achieve the inclusion of those data in both public and private sector administrative systems.

Then just do the last sentence that is already in the paragraph. The department’s leadership will also be critical in helping respondents understand why data on race and ethnicity and language are necessary and how it can be beneficial and useful to their own health care.

DR. STEINWACHS: Hear, hear.

DR. MAYS: So, we are getting rid of the whole thing about data being —


DR. MAYS: Don’t you want to encourage the states to do race, ethnicity and discharge data or are you just saying that it is just right now just have a broad comment?

DR. SCANLON: That might be a good sort of fall back and I guess the question would be whether we should focus on that in a separate document if we decide that that is a good fall back strategy because we know that sort of getting to where I am suggesting which is that administrative systems have this information and we can look at it is it is better but it may also be sort of longer term and so if we want to think about how we can make sort of hospital discharge records useful in the short term maybe highlight it in a separate document because it could get lost in here in terms of the department.

MR. LOCALIA: On question, making such a suggestion do we want to add something about challenges of getting enrollment type information on people who don’t have health insurance? It is relatively easy to do this in managed care systems and in Medicare and Medicaid but there are lots of people who have no health insurance. So, if you don’t get it in UB04 in terms of a claim where you get a large number of people who also may be overly represented in some subpopulations of interest, I don’t know the answer to that. How do we, or do we want to capture that nuance and where do we want to put it and how do we want to put it, if so?

MR. HITCHCOCK: Do you mean if say the community survey that Census is using it will give you at some point population estimates for cities and counties, if they were to add a question on a health insurance question of some sort so that at some point in time you would know whether a person was covered or not?

MR. LOCALIA: Here is my idea. You get information on people’s race, ethnicity, language in an enrollment process which means Medicaid, Medicare on first enrollment or if you are on a private insurance when you enroll but there are lots of people who have encounters with health are who don’t have any health insurance and just go to emergency rooms and offices on a case-by-case basis.

DR. STEINWACHS: So, we don’t capture it on the claim.

MR. LOCALIA: So, we don’t capture it on an enrollment form. It is not captured on a claim. It is not captured but it must be captured because you are going to lose a lot of people, that information on a lot of people. You know, they will say, “Other,” or “Missing,” or “NA,” and that is not going to work.

MR. HITCHCOCK: How do you do that?

MR. LOCALIA: That is a big question.

DR. SCANLON: It is a big question and part of it for me where would the data flow I mean from say that physician’s office where someone came and they were then billed as an uninsured patient, and it may be in a medical record someplace but again I can’t think of how we are going to access to it.

DR. CARR: The other thing though, I think studies have shown that patients who are uninsured end up going to emergency rooms because that is where they can afford the care. So, that would be captured in the hospital data sets.

MR. LOCALIA: Only if it was captured in a hospital administrative encounter form, not a claim form. What I am driving at is there needs to be some — if you don’t do it in a claim and you don’t do it in an insurance enrollment process there needs to be another way.

DR. STEINWACHS: Let me just try this because I think what you had written was cast very broadly which did not exclude this, but I think what you are pointing out very clearly is that there are some very difficult areas and challenging.

So, either there could be another line in here t say that this effort ought to recognize special challenges of collecting information on groups such as the uninsured or something —

DR. SCANLON: That is fine.

DR. STEINWACHS: Why don’t we do that because I don’t think we have testimony and other things to really support a detailed examination of this issue.

DR. MAYS: But I think we can even add more credibility to it by saying unless I am wrong and I don’t think I am that the Latino population has the highest rate of being uninsured. So, again, we make the case that there is something specific to these populations.

DR. STEINWACHS: Insurance is not distributed evenly across racial and ethnic minorities.

DR. COOPER: I assume that HHS leadership —

DR. MAYS: What page are you on?

DR. COOPER: I am looking at two pages but this is just clarification and in one section, the section we were just on we talked about the department’s leadership, the statement you just provided and on Page 9 we talk about HHS leadership. They are the same, right?


DR. COOPER: So, we should really use the same term. Are we really talking about the level of the Secretary and should we say —

DR. STEINWACHS: Secretary’s leadership.

DR. MAYS: We kind of got our hands slapped last time. So, it is a little different about sometimes it is like does the Secretary need it or is it the Department. So, some of those things where it is like, I will just say you know it is kind of like those things where you already have committees and things and all that going to the Secretary you know because the Secretary it is like somebody needs to pay attention to making this happen versus other things where it is just more casual, I kind of said, “The department.”

DR. COOPER: When it is broad it really doesn’t lay responsibility on anybody.

DR. MAYS: No, but this goes to the Secretary, doesn’t it?


DR. MAYS: We had a discussion and I am just trying to respond to that which is that you don’t shoot everything to say that the Secretary ought to tell somebody. So, there were differences and you know we may need to look at that but —

DR. STEINWACHS: So, we can’t tell the Secretary to do more than three things.


DR. COOPER: You are raising department versus HHS.

DR. MAYS: Oh, okay, I am sorry. I was doing the Secretary thing.

DR. STEINWACHS: In my job for which I get a reward, Gene is going to reward me if I keep this agenda moving.

Is there anything more on Page 10?

DR. MAYS: Okay, we will just do HHS.

DR. CARR: Following up on the rest, the second sentence in that paragraph these conferences may also be useful in broadening the discussion to include the importance of and methods for collecting other variables such as nativity and country of origin and so as to identify and eliminate health disparities.

So, do we still want that? I mean that does focus on methods for collection.


DR. CARR: So, that goes with that. I guess it is to identify —

DR. STEINWACHS: Factors related to —

DR. CARR: I am not sure we are eliminating. We are creating tools that will be used to eliminate. I think the part on the end and eliminate, that the conference may also be useful in broadening the discussion so as to —

DR. STEINWACHS: To understand health disparities maybe.

DR. MAYS: No, so as to develop tools?

DR. CARR: I am not changing language but I am just saying again it is sort of we are going to have a meeting and talk about how to collect things and then eliminate health disparities. I mean it seems a little sweeping.

DR. STEINWACHS: So either identify or understand and do away with the eliminate here?

DR. CARR: That is our title but I mean I don’t even know that we have to go there, you know that they would come together to include the importance in methods for collecting other variables such as you know maybe related to health disparities you know, something like that.

DR. BREEN: I am not sure how to resolve this but I do want to sort of go on record here that I think the reason we keep putting eliminate health disparities in here is that advocates outside the Federal Government as well as people within the Federal Government have been around a long time and feel that we have been identifying for 20 years or so and that we are probably ready for some action and so there is a concern that we actually do something as opposed to simply understand better for the umpteenth time because there is a whole lot we already understand and we could implement.

So, I think there is a tension there and there is an attempt to make sure that we mention that there is a goal here that the end goal is not simply to identify the health disparities but also to do something about it. So, I am not sure how to word it but —

DR. CARR: It might belong in the introduction.

DR. MAYS: So, as to identify and build tools useful in the elimination of health disparities. Is that okay?

DR. CARR: I just think if that is what we have just said, is that we have been studying this forever; there are still some sets we have to do because we need to get on to the elimination, if we had that as an over arching thing, it just seems like an afterthought on all these things. We are going to do this, this and this and then we are also going to eliminate and I think it understates the importance of it by just having it as a tag line, and I think it is better to put it right up front that this is what we are doing and this is how we are doing it.

DR. COOPER: What about to develop the tools and strategy, because one of the things, I mean as you go through and you identify you have the critical pieces of data, then it is going to help you to somewhat focus on in terms of which populations you are going to target?

DR. CARR: That is part two, I thought. Part two is with solid building blocks we then integrate, ask the right questions.

DR. STEINWACHS: Leslie jumped ahead this time.

Okay, anything more on this?

Page 10, No. 2?

Page 11?

MR. HITCHCOCK: What is the department’s interagency survey group in No. 3?

DR. MAYS: I don’t know what the name of it was. We talked about it here before but I don’t know what it, honestly that was one of those. When we had brought this up before I think either you or Jennifer said that there is some group that is across the department that does surveys. I may be confused. I thought there was some group that looks, I was told they meet; they look at surveys.

MS. BURWELL: I think Vicky is emphasizing external.

DR. MAYS: I was told that there is one but I think it isn’t the Wolman one, but there is some other one that supposedly —

MR. HITCHCOCK: The data strategy group of the data council.

DR. MAYS: Okay, the data strategy group of the data council. Thank you.

DR. STEINWACHS: Okay, there is internal to the department data council and then four talks about outside the department.

So,Audrey, what you were suggesting is that the Wolman group is —

DR. MAYS: Can you tell me a little bit about it?

MS. MADANS: It is the data council on statistical problems.

DR. MAYS: Can you tell me what it does, so we know whether it should be mentioned someplace?

MS. MADANS: It used to be called the agency heads. It is the heads of all the statistical agencies that meet with the chief statistician to deal with statistical policy issues. That is where the revisions came from. It originated from there. So, they deal with the occupation industry classification, the race standards, anything that is supposedly consistent across the federal statistical agencies, survey methodology, data collection. That is where it starts.

DR. MAYS: May I make a suggestion that we, usually we only cc the data council and that on this they should be cc’d?

MS. MADANS: Go ahead but it won’t make any —

you can, you can send it but they rarely get into department, they will not tell a department what to collect. If they collect it they have to do it in a certain way. So, if you look at the directive the standards, it is very clear it does not say you must —

DR. MAYS: No, it is guidelines.

MS. MADANS; No, it says that if you collect you must collect it this way. It is nothing about whether you have to collect.

DR. MAYS: Okay. Should we be including that group anywhere throughout this document as being included in some of these coordination meetings or is that not the most appropriate?

MS. MADANS: You can say that it can be coordinated through them. That would be fine.

DR. MAYS: No, not through them. We want the Secretary to —

MS. MADANS: The Secretary delegates a lot of work. That is why he doesn’t have to read all —

DR. BREEN: The Secretary is not involved. It is a statistical agency group. It doesn’t go through the Secretary.

DR. STEINWACHS: So, it sounds like we should probably not refer to them.

DR. MAYS: Exactly, but I would suggest a cc. Can we ask for a cc to them, not just data council but at least this report gets cc’d to them also?

DR. STEINWACHS: We will see what we can do.

MS. MADANS: It is very easy to just give it to Kathy.

DR. MAYS: That is all I am asking is that at least they receive it.


DR. BREEN: We probably should have a little 15 minute discussion at the end to see who is going to get a copy, you know talking about dissemination. There are places it should go.

DR. STEINWACHS: Okay, good suggestion.

Okay, as part of my compulsive disorder we have to keep this moving. Otherwise you have to give me pills. I think there are pills for some of these disorders.

Anything more on 11, moving to increased technical assistance to states.

DR. MAYS: You are okay with this, Jennifer?

DR. STEINWACHS: Jennifer why don’t you read this while we move on and you can always move us back, okay? I don’t want anyone to —

DR. MAYS: Actually you are on 13 in terms of what NCHS da, da, da, da, da, so, okay. That is what it ends up with the strategy.

DR. BREEN: We have got NCHS working with the Census Bureau here, on Page 13.

DR. STEINWACHS: So, 12 leads into the specific 1, 2, 3 for state level.

DR. MAYS: May I raise something on 12?

DR. STEINWACHS: Certainly.

DR. MAYS: I put a little note to myself that after the comment about, I am sorry, second paragraph, states also require, that after the Friedman thing is cited beginning in I forget what year, ACS that we should have some comment there about ACS and what we think ACS will be able to provide in terms of estimates for states, counties, cities. You raised that we are not commenting on ACS and so I thought that might be the place for a sentence about ACS.

MS. MADANS: Where are you?

DR. MAYS: I am sorry, the paragraph that says, “States, also, require denominator level data at the substate levels in order to target specific interventions and reduce health disparities. This requires inter-censal population denominators by race,ethnicity, age, gender, county, city and town,” and I thought maybe there we should say that beginning in such and such the American community survey —

MS. MADANS: Don’t confuse the estimates program with the ACS because ACS is a sample and so in order too get estimates you have to weight it. The weights come from the population estimates program which is completely separate from the ACS and they take the census in 2000 and they add in births and they take out deaths and they do something about migration but the ACS doesn’t really tell them whether they are right or wrong because it is a sample and they then have to weight the ACS up to a national total.

DR. MAYS: We should have kept the —

MS. MADANS: No, what you are getting, you are getting much more information in the inter-censal period but you may be completely off on the numbers. The numbers may be completely off because it is a function of the estimates.

DR. MAYS: I am fine now. I thought that ACS had —

MS. MADANS; You are talking here about denominators.

DR. STEINWACHS: I think we got Vicky back on board now.

MS. MADANS: ACS will give you things like poverty rates.

MR. HITCHCOCK: That is that whole paragraph that talks about the study in Massachusetts. That to me is almost extraneous to what we are doing here and we don’t really, I mean it is an interesting example, but does it really contribute that much to the report?

DR. STEINWACHS: This is the first big paragraph on the page?

MR. HITCHCOCK: The study of racial and ethnic classification —

DR. MAYS: I think this is to help people understand what the issues are. I don’t think people understand the differences between like racial classification and the fact that there are inaccuracies because of the different ways that people tend to use this.

MR. HITCHCOCK: We can put little anecdotes throughout this report that do exactly that. I don’t know that we have got the space in the Secretary’s head or other people’s heads to put these sorts of anecdotes in.

DR. BREEN: Actually I think it should stay in but I would make a minor modification because I think this addresses exactly Leslie’s point. If we were to add to Russ’ edit here the racially diverse Latino or Hispanic/Latino populations of Bolivians, Colombians, Cubans because that is the issue that everybody thought would be little confusing when you start combining race and ethnicity.

MS. RODGERS: But we were including people from the Middle East and North Africa.

DR. BREEN: Maybe you want to simplify that then. First of all it makes you feel like the data aren’t good. So,why collect it to begin with? If you are trying to make an argument that these are meaningful data and now you are saying, “But there are so many complexities we don’t really know what we have,” I think that tends to not strengthen your case. These are very well-known examples and there are ways to deal with them and there is error in all data and I don’t know what it buys you unless you are going to put it in a research thing.

MS. MADANS: Would you delete the paragraph then?

DR. BREEN: I don’t think people know this. I think it is well known for experts sitting at this table but I think when you think about who are the groups that we also want to make sure are reading this report and understanding the difficulties I don’t think they know that. I remember when who was it, Suzanne, presented here, Suzanne Roberts, she was defining Creole and some people later came up and said that they didn’t know that.

So, it is almost like it is a case where these things aren’t easy but if you are saying that what we are doing is freaking the public out, that we are saying something that makes them think the data is worse and why do it, I don’t want to do that.

MS. RODGERS: I have heard that in many of these kinds of meeting where race is not biologic. It doesn’t mean anything. Why are you collecting it, and I think if you highlight some of these issues in this part of the document, there is lots, this is I think early on. What is the document? Is it an education document or is it, you know we have talked to people and we understand the issues and as a Committee we have synthesized the issues. We have come up with what you, the department, need to do and I don’t know if this is where you want to be educating the department on this particular issue nor do I think this is the place to educate everyone else because there are other places to do it.

If you wanted to put something like this in a recommendation, and I don’t know if it is still in here about research then I think you can say that there is still stuff to be done in understanding race and understanding how to measure it and we need to worry about it but I think in here we are not going to be able to give guidance to the states about this problem.

DR. MAYS: What I would like to ask then, I don’t want to like throw it away as much as to see if we can move it to the research part because there is just one thing I am concerned about is who our audience really is for the report. Yes, this also goes to the Secretary and that is the primary audience but in contrast to some of the other committees they have these professional organizations where everybody knows kind of what the language is and everything. We have a different constituency that I would like to make sure that they understand.

So, can I move it then to another place.

MS. MADANS: I think you can move it, but I do think that that is an issue for the Committee. I would say your constituency is the department. I mean you are trying to convince the Secretary to change the allocation of resources.

DR. STEINWACHS: So, let me tell you in my compulsive disorder problem here that I am committed to get us to the end of this one recommendation.

DR. COOPER: May I just make a comment on that? As you change that and move that, Vicky you may want to leave it open so that other groups can also be identified because it is not just like Creoles.

DR. MAYS: These were just examples that come from a very specific study that was done.

DR. STEINWACHS: Okay, so we have made a delegation. We have picked this paragraph up, taken it out of there, moved it to research, to the buffer zone and so just in terms of strategy I was hoping to have us finish the first recommendation which is another couple of pages and then take a break for lunch, come back, finish the next recommendation and then we will see what time we have left for discussing the future agenda but the most important part is finishing this report so we are all comfortable with the changes made that we will support it and explain it to the full Committee.

DR. MAYS: Okay, so no ACS in there.


MS. MADANS: There might be a place to put it in here.

DR. STEINWACHS: Okay, there is 1, 2, 3 here. Are there any comments on the specifics on Page 12 and 13?

DR. SCANLON: It seems in No. 2 when we talk about developing intra-censal estimates by race, ethnicity, are we talking about only the methodological work or do we want them to actually do these?

DR. MAYS: Maybe I made an assumption that if you did the methodological work it would give you, but you have a lot more experience than I do.

DR. SCANLON: It takes more money to actually do something as opposed to thinking about how to do it.

DR. STEINWACHS: But the methodological work and the implementation or something?

MS. MADANS; And you might want to say it is post and inter-census. There are post-censal estimates and there are inter-censal estimates. This is a good recommendation because they are doing it.

DR. STEINWACHS: Okay, even better yet. We have been successful already. Count one for us.

MS. MADANS; I am sorry but are we saying —

DR. SCANLON: We are going to say something about and implementation.

DR. STEINWACHS: Yes and implementation. I thought it was methodological work and implementation. I don’t know. You have to maybe craft the sentence a little bit differently.

MS. MADANS: Part 2 of that, the socioeconomic and contextual that is more the ACS which you can leave it there but we don’t do a lot with them on that.

DR. MAYS: So, you are fine with that as a charge to you?


DR. MAYS: Okay.

MS. PAISANO: I guess my only concern in listening and hearing the hearing now of American Indians and Alaskan Natives, I mean things are always stated as national, state level and below but it is very important to tribes to have estimates for reservations or their tribal land areas.

DR. MAYS: So, national, state level, tribal reservation and —

MS. PAISANO: It says, “And below,” but and below nobody ever thinks about the tribal.

DR. MAYS: No, no, I want to put it in. So, could we say —

MS. PAISANO: Yes, I would include.

DR. STEINWACHS: And below, comma, including.

MS. PAISANO: I would say, “And tribal land areas,” because if you look at the geography for tribes they are termed differently, reservations, trust lands. In Alaska all the villages are considered tribes. You also have tribes who don’t really have a land base.

DR. MAYS: Can we say, “Including tribes and tribal land areas”?

MS. PAISANO: I think that would get it.


DR. MAYS: Again, you missed the very first thing where I said that you have to say whether you think you need money.

MS. MADANS: Yes, we need money. You don’t have to ask that question. The answer is yes, we need money.

DR. MAYS: So, do we need to say anything? I am very serious about this. Do you need to say things in here about funding?

MS. MADANS: I think that is something that would go in the introduction because if you are asking anybody to do more than they are doing there has to be some kind of reallocation of funding. We don’t have money we don’t spend.

DR. STEUERLE: I guess it is being that and I don’t know where it goes but on the funding it is what this document doesn’t address again because there is no ranking of priorities or something. It is sort of like if you had 10 more million dollars you don’t necessarily want first to come to the table to spend it. You really want to have some way of really thinking very strategically about the funding issue and some of the funding issues by the way which is always a tough one in departments is in some cases it might be better if HHS would just give a million dollars to the Census instead of figuring out how to do something themselves and I know all these interagency problems and you can’t do it because you get money internally but somewhere we need somebody who is worried about this. So, maybe you need a couple of sentences. I thought writing them up front about —

MS. MADANS; Never ever say, “Give the money to Census.”

DR. STEUERLE: I understand. Census is the wrong place. I can give you better examples.

DR. STEINWACHS: Okay why don’t we come back at the end and talk about if there is something to be said in here? It is probably up front about resources. I think the issue of priorities is probably going to be tough at this point. I think what Gene you are talking about is it would be good for us to try to make sure we showcase what we think is a couple of the most important things.

DR. STEUERLE: I guess at the end of the day the Secretary gets this. What is he tasked to do, say, “Oh, here, go take care of it,” to somebody and that guy says, “What does this mean?” and he sends it out to all the agencies. “Take care of it.”

DR. STEINWACHS: You are right and —

DR. STEUERLE: But you might at the end politely ask for a report back on things that have been done and how resources might be allocated. You might just ask for something back.

MR. HITCHCOCK: I can tell you the longer we deliberate on this report the more actually is getting done.

DR. STEINWACHS: That is why we are going to come to some success before we are done. We do get reports back on what happens generally but that is through the departmental process. I guess to some extent areas where we can identify groups like the data council that probably gives a clear message. When you say, “Ask the Secretary to do it or the department then I think you are right.

DR. MAYS: I think there are things that are high priority and maybe they are not emerging and it is like in the next set it was for example doing the targeted surveys.I think that the issue of training on the collection of race and ethnicity, so, I mean I can pick, I mean in terms of what evolved from the hearing, etc. So, at the end if you want to do that I think that if you ask me what is the biggest bang for the buck that somebody may have a different one but I think if we began to have these targeted surveys and again I was looking at that and the notion of well, we need to specify it can be all of the populations or those who appear in less than 5 percent or you know we need to put some bounds around it, but that would be the biggest bang for the buck, followed by a data center where you could access to the data, followed by teaching people how to do the collection of race and ethnicity.

If you gave me those three i would be happy.

DR. BREEN: There are clear priorities. Why not put them down there?

DR. MAYS: Every time I bring them up that is what happens. It is big dollars. So, we always have little things.

MS. MADANS: When you say, “The biggest bang for the buck,” in order to make that evaluation you have to start up with how many bucks you are spending. So, if you had $10 million maybe that would be the biggest bang. If you don’t have 10, it is not the biggest bang.

MR. HITCHCOCK: We are actually looking right now at the data council what would be the biggest bang for the buck on a lot of these issues and some of that is small increases in sample size would actually do some rather —

DR. STEINWACHS: Okay, we have three big bangs which is good.

I am moving us on to enhance the data collection on race, ethnicity and primary language in federal health programs. Anything on that?

DR. MAYS: Somebody needs to say something because these are kind of radical what I threw in here.

DR. SCANLON: Fourteen my concern is whether we are being realistic about Medicaid, kind of what the state of the world is. It has been a little while since the last GAO report that I worked on on Medicaid programs but states had trouble getting encounter data in many instances and there is that issue. So, the question when we talk about the answer requiring each state to do something you know are they all going to be capable of it and then there is also an issue here that we start off talking about incentives to do something and then we talk about requirements and in the context of Medicaid today we are looking to take $10 billion out and requirements. The unfunded mandate word has come up before today. So, there is a lot about this paragraph that I am not quite certain about whether it is too far out on a limb for us.

DR. MAYS: I had asked the CMS folks to look and I think Nilsa has been out of the country and is back and she said that she was going to be by and I think I also sent it to Judy but I think Judy sent it to Nilsa. So, I really did want to get some consultation. This is kind of where people would like things to go, kind of more radically in the field where they want it to go, kind of more radically in the field where they wanted to go but we, also, are not trying to send just radical things. I mean we are not a community agency sending radical things up to the Secretary. We want it to be feasible. So, I put what is being asked for.

As a matter of fact if you look on Page 13 at the bottom this is actually some of the things that Frist has now sent forth in his, I have forgotten the name of the bill. I am going to be terrible about that. So, this is part of what —

PARTICIPANT: Is it the health care disparities bill?

DR. MAYS: No, it has another name to it. I want to say health care act or something like that. I am not a bill person. I mean I talked to some of the staffers and stuff. So, this is what they actually want to go forward down here.

Now, the one that says, “No. 1,” is what we have had kind of in the report for a while and that comes from things that have happened in hearings and what have you but again maybe we need to have one of the CMS people help with this one.

DR. SCANLON: I think setting objectives and maybe a little bit less specificity in terms of whether it is going to be requirement or it is going to be incentive would be helpful but setting the objective. It would be ideal again, particularly because so many of the people of concern are Medicaid eligible for that. We get information through the Medicaid program and restructuring it that way will probably make it safe.

DR. MAYS: Okay, so you just told me, again, the Committee also recommends that CMS set objectives for state Medicaid agencies or I mean can you —

DR. SCANLON: The Committee recognizes the importance of having information on Medicaid eligibles and their use of services and I don’t have the exact words here now. I will think about it.

DR. MAYS: Okay.

DR. STEINWACHS: Okay, I have got Bill’s name by that. It is wonderful to have volunteers.

Okay, I assume it would be a good idea to break now for lunch and I think Vicky wanted a little time to cloister with, there was something previously you wanted Dale’s help with?

DR. MAYS: He actually wrote back. The other one he wrote I don’t know now in terms of like we are getting the ACS is out but there was another period. Do you remember?

DR. STEINWACHS: You had also some of what is out of the department and what is in the department. The other would be as if Dale or others could, once this goes through and double check that, I don’t know what is most practical, whether to do it now or later, but we need it done soon.

DR. BREEN: Jennifer might be able to contribute to that, too. I mean she would know the department well.

DR. STEINWACHS: Dale knows it. Okay, what is reasonable, to get back here by one? Is that okay? That is 50 minutes. Do you think that is doable? Okay,

(Thereupon, at 12:10 p.m., a recess was taken until 1:05 p.m., the same day.)


DR. STEINWACHS: Okay, I am on Page 15. Recommendation: HHS is urged to increase and strengthen the capacity of its health statistics infrastructure to collect, analyze, report and disseminate data on the various ethnic, racial, linguistic subpopulations in the United States and territories.. Are we all right with that statement?

I have Justine diagramming the sentence next to me. So, I just want to make sure.

DR. MAYS: I have got to get back to that. That is a good idea.

DR. STEINWACHS: I had one or two lectures in fifth grade on diagramming. It was somewhere around there, and it didn’t stick with me that well. I could have used more of it.

MR. LOCALIA: Could we just say, “Increase and strengthen its infrastructure capacity?

DR. STEINWACHS: So, HHS is urged to increase and strengthen its health statistics infrastructure?

MR. LOCALIA: Isn’t that sufficient?

DR. STEINWACHS: Yes. That makes ie clearer.

MS. MC CALL: So, we are not going to have the separate piece to collect, analyze, report?

DR. MAYS: No, I think we just deleted the capacity of. That is all.

DR. STEINWACHS: Yes, so still to collect, analyze, report and disseminate data on the various —

MS. MC CALL: I would like to add a verb there, and it is to integrate.

DR. STEINWACHS: Collect, analyze —

MS. MC CALL: Collect, integrate, analyze, report and disseminate.

DR. STEINWACHS: Okay, we need to make sure that comes up also in the text, too..

MS. MC CALL: Correct.

DR. MAYS: So, when we have this litany here we will make sure that integrate also appears if it makes sense.

DR. STEINWACHS: Okay, any other word smithing on the recommendation or does that sound/

DR. CARR: Are we eliminating or ameliorating or reducing health disparities, Line 1?

DR. STEINWACHS: Oh, you are into the paragraph. In order to improve health care quality and ameliorate and eliminate? Do you think we ought to —

DR. CARR: We ought to keep to your point about continuing to send the same message. We are sort of skating around, back and forth on this and are we improving health care quality or are we eliminating health disparities or reducing or whatever?

DR. STEINWACHS: I think I would drop the health care quality just because we don’t really deal with it later I don’t think, Vicky, in this section, do we?

DR. MAYS: We do a little bit, but it is okay because I think we are broader than that. So, it is almost like this is more where what follows it begins to stuff about art but I think it is the broad recombination. So, let us not.

DR. STEINWACHS: Okay, so in order to eliminate health disparities HHS must ensure that the nation has and go on from there.

Okay, other suggestions and comments on Page 15?

MR. LOCALIA: One as part of the second paragraph. I think it should say, “And revising procedures.”

DR. BREEN: Instead of “To revise” say, “And revising”?

MR. LOCALIA: Yes, is that okay?


DR. MAYS: The last paragraph, the Department is urged to pursue the following strategies that will help strengthen its health statistics infrastructure, increase use of and quality. I mean I think it was Dale that also cautioning we don’t want to just be on the side of increasing quantity, that we also want to make sure we increase the quality of the data.

DR. CARR: But I mean hopefully that was strategy one. So, I mean I think again we have already said it should be high quality collected data and now we are onto the integration of it.

DR. MAYS: Okay, we will leave it out.

DR. CARR: Back to Paragraph 1, the last sentence, the Department must do a better job of or provide incentives? I mean are they providing incentives inadequately or I mean I guess do a better job of is, I was thinking if we could take that out and just say —

DR. STEINWACHS: Where are you?

DR. CARR: First paragraph, last sentence.

DR. BREEN: I think there are some incentives like for example the loan repayment program is an incentive.

DR. CARR: I guess it is like are you implying they are doing a bad job or would you say that the department must provide more incentives?

DR. CARR: Oh, I see, okay.

DR. BREEN: I think it sounds better to say, you know, additional incentives or provide incentives or something because it is sort of implicit that it is not being well done otherwise. Do you want to say, “Provide more incentives” or just incentives?

DR. CARR: More incentives if you think they are already doing it.

DR. BREEN: Yes, they are doing some, but it is very, very limited.

DR. MAYS: I agree. That is why I started out with a better job, but I think, again, no need to be nasty.

DR. STEINWACHS: Vicky, we never think of you as being nasty. I have never seen that side. I don’t think it exists.

Okay, others?

DR. MAYS: So, where was yours?

DR. CARR: No, I was just saying more. It is more of a quantitative judgment then doing a better job is a quantitative judgment.

DR. STEINWACHS: Okay, I am going to Page 16.

MS. MC CALL: Page 14, second paragraph.

PARTICIPANT: There is only one.

DR. STEINWACHS: It depends on how you printed. So, tell us where you are?

MS. MC CALL: Okay, it says, “Researchers must also have access.” In that first sentence it says that subpopulations who most often suffer disparities, you know, disparities from what? We use that term a lot but has it been defined up above more clearly?

DR. BREEN: Disparities in their health status, health outcomes, access to care and health care treatment. Did you mean disparities is unclear or what?

MS. MC CALL: Yes, disparities from what?

DR. BREEN: Health status, health outcomes, access to care and health care treatment.

MS. MC CALL: It is just kind of odd.

DR. CARR: Experience disparities.

DR. STEINWACHS: Is the word suffer versus something like experience or —

MS. MC CALL: No, if you read it many times you know how things start to sound odd, and it is like I am suffering a disparity in health.

DR. STEINWACHS: I don’t like the suffering part.

MS. RODGERS: You are right. People don’t experience disparities. There are disparities across populations from these different things but people don’t individually experience them.

DR. SCANLON: If we were to talk about this early on maybe we could have a comment about the disparities being negative and that subgroups have poorer health, less access to services, those kinds of things.

DR. MAYS: You know in the beginning where you say that blacks have this; Latinos have that, I mean I didn’t do that at all in terms of writing a whole —

DR. SCANLON: I think a single sentence.

DR. STEINWACHS: Yes, just to say when we say, “Disparities, we are thinking of negative and we are —

DR. CARR: And then the sentence that says, “The department is urged to pursue the following strategies that will strengthen its statistics infrastructure and then we have semicolon, use of data collected and improve dissemination. Also, I was going to say should we call them recommendation No. 1 and recommendation No. 2?

DR. STEINWACHS: Yes. Does that sound good.

DR. BREEN: So researchers is one and the department is two.

DR. STEINWACHS: At the top this is recommendation 2.

DR. CARR: And then at the bottom of the page this is what we mean by two, that they will pursue —

DR. STEINWACHS: So, it is pursue the following strategies will strengthen —

MS. MC CALL: Although I will say that this paragraph does not reflect all of the verbs and words that we put in the title recommendation.

DR. MAYS: I was going to go back and say that this last one is like weak now because once you kind of delineated what you wanted in one and what you wanted in two now this paragraph seems really weak to some extent and it doesn’t really, it is not a real, I think, lead in to it. It just says, “Increased use of data,” as opposed to, I don’t know if it increases it as much as —

MS. RODGERS: We could take out those last two phrases and just say that it will strengthen its health statistics infrastructure.

DR. MAYS: No, I think it is good to have what it is as some active words.

DR. CARR: Isn’t it increase the data collected.

DR. STEINWACHS: That was the first one.

DR. MAYS: Here it is the analysis and we don’t have anything in here about the training.

DR. STEINWACHS: There is also an idea I guess of increased use of data for. Is that for monitoring, evaluation, for you know, and maybe if we could put some action words in there that would make it clearer because certainly from one perspective the department is responsible for monitoring progress or potentially failures in eliminating health care disparities.

DR. CARR: How about that the department is urged to pursue strategies that will foster sophisticated you know, management.

DR. MAYS: Greater use of the data at level of complex analyses of small groups or something like that.

DR. CARR: What we are saying is we want them to foster the education of researchers and the facilitation of the availability of access, you know, facilitation of access to these data. So, we want educated researchers, sophisticated question asking and available resources or access too the data, I guess.

DR. STEINWACHS: Let me ask this because there was a little discussion earlier about how many of us said that we have spent years studying sort of that there are disparities and we haven’t experienced much energy studying how to intervene to change and so I was wondering whether or not there was any value in here of trying to put some words in that the increased use of the data is to seek ways to intervene to solutions for health disparities and to evaluate the progress in moving toward those and then the dissemination of this and I think the point, Vicky, you are making here, also, is that one needs to look at the critical subpopulations. So, there is another part that is drilling down. I will put it that way and that is not the right words for it. So, maybe we could do two things, one, try to enrich this sort of in saying that we recognize now is the time to look toward using greater use to try to identify solutions, disparities, monitoring progress and assessing the special circumstances and focusing on, maybe it is and focusing on critical subpopulations. I am looking at Vicky. Help me.

MS. MC CALL: I like the phrase “identifying solutions.” I don’t want to have to define everything a researcher is going to do.


MS. MC CALL: That is just too broad but I think we can identify and opine on kind of why they are doing it and use words that are consistent with the overall objective.

DR. MAYS: Can you do what you just did again?

DR. STEINWACHS: I can never repeat myself.

DR. MAYS: I was thinking and I couldn’t get it all.

DR. STEINWACHS: Bill, save me.

MS. RODGERS: I am working on part of that. Maybe I can help you. Identify solutions to health disparities and track progress toward these solutions and then there is a different kind of a thought about assessing subpopulations.

DR. STEINWACHS: So, identifying solutions, tracking progress toward eliminating disparities, giving special attention to —

MS. RODGERS: To the diversity of the subpopulations of these racial, ethnic and linguistic groups.

DR. STEINWACHS: Okay, let us keep going.

DR. BREEN: I have a question. In that last paragraph that we have just been discussing and revising and stuff I like what we have done but did we get rid of strengthen the health statistics infrastructure? I hate to see that go.

DR. STEINWACHS: No, that is there. It is a semicolon I guess or maybe a new sentence but we are keeping the front part of that because that is the starting point, the infrastructure facilitates this and then we are keeping dissemination also, Nancy. This is for a matter of record. Nancy was going to hold my feet to the fire.

Okay, I am going on to the next page, the section increase the availability of data on diverse subpopulations. Suggestions here, concerns?

DR. MAYS: The second paragraph, I am sorry, it is actually the third paragraph where it says that we are urged to do all this stuff. I kind of said, “All subpopulations,” and I just think that is a little much. That is like mom and apple pie. It would be I mean, you know like all populations. There could be 60 people or some group in the whole United States and we are kind of calling on them to do a survey. Usually there is some guidance about they make up 5 percent of the population and I am looking at Jennifer for help on this. Sometimes when you decide these things you decide them based on trying to put some balance around them.

So, help us as to how to make this more reasonable?

DR. BREEN: Why do you think it is not reasonable?

DR. MAYS: Well, it says, “Whose numbers are so small and geographically dispersed that due to” da, da, da, da, da, they are not available, and I just don’t —

MS. MADANS: Where are you reading?

DR. STEINWACHS: This is No. 1 under A1, the department is urged to develop a plan.

DR. MAYS: If that is enough then it is fine.

MS. MADANS: I see what you mean. That is every population.

DR. MAYS: Right and it is kind of like it could be 50 people who have just come as new immigrants or something.

MS. MADANS: Obviously everybody we don’t have data on it is because they are too small.

DR. MAYS: Right. So, I am trying to be more reasonable.

DR. BREEN: I think that is the issue.

MS. MADANS: We like 10 percent.

DR. MAYS: Okay, I said, “Five.”

DR. BREEN: What percentage?

MS. MADANS: If you want an estimate of a population it is much easier you know if you can get much lower. If you want to say something about that population then you need to get a sample size that is big enough to say something. So, you probably need 1000 people and in order to get that 1000 in a population if you are doing something around the size of an HIS if we go below 5 percent we are not going to get it.

DR. MAYS: I said, “Five percent.” She said, “Ten.”

MS. MADANS: If you want to spend more money you can go —

DR. BREEN: We are eager to spend more money. I don’t know why you think we are not.

DR. MAYS: Money you don’t have, but anyway. So, I think to be reasonable we should something because then if this goes through and everybody starts advocating, well, there are 30 so and so —

MS. MADANS: I wish I had brought, we have a, maybe Dale has it. We did this chart of looking at the percents from the Census and what we get, how big an HIS would have to be to pick up enough and after you go to what we already have it really, the percents drop so dramatically, I mean you are dealing with less than 1 percent on some of the groups. It is almost inconceivable that you are going to spend, that you are going to be able to get that in any kind of reasonable amount of time. So, it would be useful I think if you just look at the Census. Do you have that with you? Look at the Hispanic subgroups. So, just to give you an idea of the Hispanic subpopulations we have now about 900 Puerto Ricans in the sampling if you add 3 years of data. So, we were kind of going for 3 years of data.

DR. STEINWACHS: So, if you had to estimate what proportion of the US population —

MS. MADANS: We would have no problem doing that. You need enough sample to say something about Puerto Ricans.

DR. STEINWACHS: I was just trying to go the other way, you know, Puerto Ricans 1 percent or —

MS. MADANS: One percent. You don’t need a whole lot to estimate that but Census has already done that for you. So, you wouldn’t look to us to do it. You want us to tell you something about it, right?

DR. STEINWACHS: It sounds like this discussion is sort of leading towards saying that the department ought to develop some criteria. So, if part of what you are talking about is criteria then we might say that we could go down to some subgroup and some percentage of the total population.

The other is where the population is geographically concentrated. So if you talk about an indian tribe in a geographic area —

MS. MADANS: And then you are just going to get that one geographic area.

DR. STEINWACHS: Yes, and so it seems to me the strategy is either one that deals with down at a certain level of the population composition or geographically concentrated populations where you go into a survey once every 10 years and then that is it because you know you are going.

MS. MADANS: These slides actually give you the, I have to, I can’t calculate so quick in my head. What these charts needed, what we were doing was giving you the confidence intervals around selected health estimates for all the subpopulations in HIS. So, if you wanted to look at Koreans for example with a certain number of years of data and estimate health insurance in the under 65 it gave you the percent and the confidence intervals. So, you can get an idea of how much data we already have and then as you start multiplying that by two and three in terms of how big the sample size would have to be —

DR. MAYS: Suppose we say that the department is urged to develop a plan and a criteria for collecting data?

MS. MADANS: That would be good.

DR. MAYS: Okay, suppose we just do that and then that is kind of punting it but again it is like if the data council is already working on this I think that is the message. It is keep up the good work.

DR. STEINWACHS: Okay, so, let us go on.

No. 2 here, technical assistance.

DR. MAYS: Where HRSA appears I am going to put — oh, Dale comes in on cue.

DR. STEINWACHS: Try to make sure that Vicky does not confuse the inside with the outside of HHS, not that any of us could do any better except for Dale.

Okay, so, two, three? Survey research methodology investments; four, department agencies which we just talked about.

MS. PAISANO: It is talking about funding studies or funding research and IHS never has received funds for research. Most of the money goes to delivering services.

DR. BREEN: And that came up big time in the hearing.

DR. MAYS: So, do you want to be left in for the possibility of that or out? The reality is you don’t, but do you want to be left in for consideration?

MS. PAISANO: To get research money to come to IHS? Sure.

DR. BREEN: What is confusing is fund through existing mechanisms, but I don’t think you mean it at the same level which is kind of what you think when you first read that.

DR. STEINWACHS: How about leaving out through existing mechanisms?

MR. HITCHCOCK: Before you go too far the 1 percent evaluation fund is an existing mechanism that could be used for this but I don’t know that IHS is eligible. I think you are eligible but maybe you have taken it into the services pot. I don’t really know.

DR. MAYS: The only reason in several places we had, and again this is something we were warned about before, that if you tell them to use what they already have it is easier to kind of almost immediately put funds into something versus creating something new. That was just something that was said.

MR. HITCHCOCK: We could do that. We have this 1 percent evaluation fund that is at all the agencies and mostly NIH because they have the biggest chunk of the budget but it goes to little research projects and more of it could be pressed towards these topics.

DR. STEINWACHS: And, Dale, a lot of us thought 1 percent was there just to fund AHRQ.

MR. HITCHCOCK: Yes, well, NCHS and ASPE.

DR. STEINWACHS: So, I think it sounds like we just leave it as it is. I think that is what I took from what we said.

DR. MAYS: And maybe you could find out, Edna whether or not you qualify. If you don’t then we could say something different, but if you don’t qualify for it it is a little different.

MS. MADANS: When you talk about determine best practice to increase participation are you talking about like response rate?

DR. MAYS: I am sorry, tell me where you are?

MS. MADANS: We are on four, right, and right after that through existing mechanisms, methodologies and empirical studies to determine best practices to increase participation in health research of racial, ethnic and linguistic populations which to me implies that there is some problem with getting people to agree to participate as opposed to designing data collection that would include more of them.

DR. MAYS: This is actually about the barriers to participation and it may be that because then the next part is here are some of the barriers. What happens when you say that you are going to have a specimen collection?

MS. MADANS: And that is unique to these populations as opposed to just in general?

DR. MAYS: I think the issue of biological specimen collection is pretty unique to, unless i don’t know this, but I mean there are subpopulations very unique to like American Indians.

MS. MADANS: It may be but it is not clear that there might be unique things.

DR. MAYS: Oh, okay.

MS. MADANS: As opposed to the generic issue of how do you do specimen collection.

DR. MAYS: Okay. I didn’t know whether to get that specific but biological specimen collection is a big issue for at least American Indians and the notion of collecting a specimen and once one dies can one you know move on to the —

MR. HITCHCOCK: Or worse than that is collecting it on the pretense of using it for a diabetes study but then turning around and using it for a schizophrenia study or something like that.

DR. MAYS: Yes, that is a different issue but that has come up as a historical issue that actually is causing some groups, it is a big historical issue that is being sued over in terms of American Indians I think in Arizona. So, it is spreading I think to people of that particular population not wanting to participate in studies because of the belief that you give for one thing and it is used for something else. Especially it will be worse after the court cases.

MS. MADANS: And payment, innovative payment.

DR. MAYS: Okay, innovative payment is an issue that sometimes if you are on subsidies like you know you are getting welfare, etc., you are supposed to report anything that is over $25 and so we now have to think about what we are going to pay people and particularly if you are in a group where they are making you pay them by check and all that, you know, you are Westat or UCLA or something like that and they are supposed to report it. You put people at risk, if they don’t then report it to the welfare agency of losing their money.

DR. BREEN: How about innovative incentive methods?

DR. MAYS: Okay.

DR. BREEN: I wonder if that first sentence might be two or something. It is an awfully long sentence.

DR. MAYS: I think in Jennifer’s comment it is almost like we should make these two sentences because the first is about population and then the second, so, I mean I think we do need to kind of break that into two sentences.

DR. STEINWACHS: And on four where you say, “Investigations of the barriers,” maybe it is investigations to overcome barriers.

DR. MAYS: Yes.

DR. STEINWACHS: Related to collection of biological specimens, concerns with privacy, confidentiality and ethical issues and I am not sure how to deal with the —

DR. MAYS: I think it would be a unique thing to say to overcome some of the historical incidents and barriers or something like that.

DR. STEINWACHS: And then make up a list, and I think that would flow better because right now you sort of have to think hard about each piece and you have provided the critical background.


DR. SCANLON: I guess I am kind of lost as to the distinctions between four and five. I think there may be some but again I am not quite focused because we talk about increasing participation and we talk about translation and the introduction to five about benefits to these populations seems to be one step removed.

DR. MAYS: This came up specifically in the hearing and that is first of all what surveys have what racial, ethnic subpopulations in them; would it be useful, for example, to think in terms of a survey integration such that say if one survey doesn’t do this well but you have now looked at all the surveys that are being done, say, on diabetes that ask or no, ask an insurance question and you realize that HIS might do it better in terms of reaching I don’t know some other group. So, in this kind of survey integration you would say, “Let us make sure that they then assess this question and you might even ask them to do an over sampling or you might ask them to do something different but it is looking at all of the surveys, figuring out what you are getting and what you are not getting and figuring out if across the surveys there is any way to either share data or ask someone in another survey to do more of something, but the problem is it is not written well enough to understand that. So, that is an issue.

DR. SCANLON: Here is a suggestion, that the department should undertake a study to determine how well racial and ethnic subgroup populations are represented in HHS surveys and how their representation could be improved through and then I guess maybe survey integration would be next two words.

I am not sure about methods for increasing participation because that seems like it might fit better in four and the use of emerging technologies also might fit in four, and then I might be done with this paragraph because the translation issue in part relates to participation, doesn’t it?

DR. MAYS: I think that is a separate issue because that is an expensive very technical issue.

DR.SCANLON: But you do also talk about translation methods in four.

DR. STEINWACHS: Maybe you want to take translation out of four and create a separate one on translation that talks to the fact that many of our surveys cannot be given to all Americans because they are not translated.

DR. MAYS: So you are not saying a separate number like six, but —

DR. STEINWACHS: I was sort of going for a separate number and make translation that —

DR. MAYS: Okay.

DR. STEINWACHS: I think that is very much a separate issue.

DR. MAYS: Okay, and yes, that kid of raises the language thing.

DR. STEINWACHS: And this I think by doing this on five it brings integration as a central point.

MS. MADANS: Survey integration in the department has meant something very specific. It is really like the HIS, that is survey integration where you build a component off of another and I think you are talking more kind of post, integrating results after the fact.

DR. STEINWACHS: But to do that successfully you would ideally do some integration or harmonization or something like that.

MS. MADANS; If you say survey integration people are going to think you are talking about this —


MS. MADANS: Yes, because that was a painful transition.

DR. STEINWACHS: Okay, so, if you are trying to pool afterwards the same sample frame makes it easier but you are saying that there are other things that could be done. What would you call those?

MS. MADANS: We usually call them coordination, but that means so many things it doesn’t, maybe we should actually use the words you just used in order to combine.

DR. MAYS: Integration and, oh, I am sorry.

MS. MADANS: Combine results of various surveys and then that would require some pre-planning so that you have consistency in key items. That is really what you —


MS. MADANS: We tend to have the consistency on the demographics but then once you are looking at the outcomes we have less consistency.

MR. LOCALIA: I think the word to use is combination but I am not quite sure how to handle the pre-planning to facilitate that because I think what you are talking about is that unless up front you figure out how this is going to happen it can’t happen. Is that correct?

DR. STEINWACHS: that is where sort of the integration idea comes in.

MS. MADANS: So, it is integration and, Dale what were we calling it?

MR. HITCHCOCK: Standardization.

MS. MADANS: Yes, standardization of core items.

DR. STEINWACHS: So, exploring integration standardization and coordination to maximize the potential.

MS. MADANS: So, that data can be combined after.


DR. MAYS: The only thing that makes me nervous is that when you say, “Standardization,” again I always think every survey because it has a different purpose should have a flexibility and then instead I always try to push for equivalency to figure out even if you have done something differently from a kind of statistical point of view can you say that these are equivalent and therefore you can still use and combine them but you all don’t like that I see.

MS. MADANS: I think the standardization word is I think what we want but recognizing that surveys need to be flexible, but if you have I don’t know, diabetes questions and some other survey is asking about did you ever have diabetes and there are like these minor wording changes that really are not part of our mandate.

They are usually part of our history. You really can’t compare those because you will have 100 people telling you this slight wording change will make a difference or this will make a difference and you really can’t combine it and you will say, “Yes, that is true.”

So, I think you want to push these kind of more —

DR. MAYS: Okay, in the face of flexibility.

DR. STEINWACHS: Okay, we have five. We are okay on that and we added a six.

Okay, let me take you to B, expand and secure data centers. Bill?

DR. SCANLON: I think you brought me along in terms of this is an option and I guess the idea would be where can we also explore other approaches as well.

DR. STEINWACHS: I was wondering whether or not under the recommendations on the next page —

DR. MAYS: Yes, I was just going to say, let us go to 19.

DR. STEINWACHS: No. 3 there says that the department should consider additional options for providing access to micro data for research and statistical purposes in a manner consistent with existing study regulations.

DR. MAYS: The department should increase remote access as well as the number of and then in the second you should say, “Funding should be allocated to,” what do you want to say, development of software dissemination of methods for remote access as well as? You weren’t here when we were talking about remote access. Jackie was on the phone. Currently in terms of remote access how much of the data can you get from these small groups through remote access?

MS. MADANS: You can access any data through remote access. The question is what can you see and get sent back to you. So, there are certain things that you can see in the data center but you would not be allowed to take out of the data center, but once on the remote access the only thing that will be returned in terms of the output would be things that you would have been allowed to take out of the data center if you had physically been there because we have not control over it.

So things like line listings, it won’t let you look at individual records. It won’t let you look at small samples you know tables that have very few cells. So, it does an evaluation that is stricter than the evaluation we would do at the data center.


MS. MADANS: Actually I don’t think that is true at all. We don’t get a whole lot of complaints about problems with the remote system.

I think the hardest part of the remote system is if you are not familiar with the file at all sometimes you really need to, you really want to get your hands dirty with the data and you can’t do that.

So, then we will come to the data center for a day and do that and then create an analytic one. So the alternatives to that other than more centers and we are talking to the Census Bureau about joining forces and having joint data centers. It is a big breakthrough because up until last year —

DR. BREEN: Commingling the data?

MS. MADANS: They agreed to talk about it. So, we are talking about it. We are talking to them about a second generation remote access system, one that would be web based plus a more analytic one that we would use.

MS. MC CALL: You said something about what it means for somebody who is not familiar with the data to really get intimate with it and understand it and some of that is being physically close to the data but it is also somehow being close to the experts.

MS. MADANS: Yes, that is true.

MS. MC CALL: And so part of this could be and part of the recommendation could be around creating you know you could call it a user community. You could call it a community of practice around how to make the people who are using it aggressively, make some forms for them to talk about it easily.

MS. MADANS: Good idea. Not quite along this line but something that would help is the idea of creating, we used to call it fake data. So, it is not the real data. We create data but actually that is very useful for understanding the structure of the file and what it looks like.

MS. MC CALL: Synthetic data. That is a good term.

MS. MADANS: Synthetic data, right. For some of this the basic outcome variables are going to be the same. So, the researcher can get pretty familiar with how the let us say the HIS file works by not looking at the individual subgroups and you really get fairly far down the line and that makes your job easier.

It is not like you can’t see the file at all. You are seeing a lot of the file in the public use file. You are just not seeing this one variable and that is the one that if we had some synthetic data which showed you how it was distributed that would be useful.

So, those are the things that —

DR. MAYS: Let me just ask because if the population is small enough don’t you not have access to it or that you end up imputing, I mean you kind of — you don’t? Oh, okay.

MS. MADANS: You can have access to it, but if it is so small that if anything you do with it is going to be identifiable you can’t do anything with it, but if you have a population of pick a group?

DR. MAYS: The Hmong.

MS. MADANS: But I have Korean on my list, okay, so we have 121 Koreans under 18, okay, 3 years of data; well, there is not a whole lot you can do with 121 people. Once you break that down by age and sex and something else, and we want geography, you are going to have to import Koreans because there —

DR. STEINWACHS: That we can do, too.

MS. MADANS: If you want to look at Asian Indians we have 241. These are kids. So, you can start. Under 200 there is not a whole lot you can do. If you had 800 you could start making estimates. So, those are kind of the rules of thumb but if you are getting, some of these groups you are going to either have to do something very geographic which is what you are talking about, very targeted geographic or get 6 to 10 years of data and over sample and that is expensive.

DR. MAYS: Let us fix then because I think you have come up with suggestions that would really enhance and increase the usage of the data if there were CD-ROMs or something online where you kind of have a synthetic data set because researchers typically don’t want to do this because you don’t get to play around with the data well enough to know what you are doing. You always feel like I need to see it more. So, what happens is you don’t get complaints because they don’t use the remote unless they know it. So, let us see if we can go to this and capture what you said. MR. LOCALIA: Some of the options you could tell them very much.

DR. MAYS: I think an i.e., or something would help.

MR. LOCALIA: You also have to distinguish between current surveys and the limitations of the overall sample size. Surveys that may be conducted that are focused, if you had a focused survey the identifiability is still going to be an issue even though the numbers might be sufficient. So, I am not sure how to, this can get so difficult so fast, I am not sure how specific you want to be here.

DR. MAYS: Let us do this piece by piece. Do you want to call for an increase in remote access?

MR. LOCALIA: I don’t know whether you want to call for an increase in remote access or whether you want to call for an increase in options. It might include remote access. It might include some other things.

MS. MADANS: What about for example just list them all out?

MR. LOCALIA: We would need some help in listing them all out. Okay, we could say remote access. One is like synthetic data, reference anything there, add some references. There is all that stuff about I mean the literature is —

DR. BREEN: Aren’t those all synthetic?

MS. MADANS: No, synthetic is totally made up and the other is where you just jumble it. Some of the data are real but —

DR. STEINWACHS: Could we do this? Some of these things that we are talking about fit under three here, don’t they, and there are additional options, having access and including or for example, you know. I don’t think we probably have the time or capacity right now to list everything but I think that some of the things that are mentioned here would make people understand what it is we are thinking about and Bill?

DR. SCANLON; I guess the question for me the discussion seems to have raised the idea that having a secure data center doesn’t really add to your capacity that much. I had envisioned when we were talking earlier that you would be going there daily to do all your work as opposed to kind of this idea of getting familiar with data and then maybe you would be able to do a lot.

MS. MADANS: You could do both.

DR. SCANLON: You could, but the question is do you need to. This is real money. First of all we have regional data centers and then we want them really spread out. There may be sort of questions of interest. It seemed like we were talking about a lot and so given how scarce the resources are the question was this investment versus some of the other things that maybe we would like to have which would be bigger samples.

MR. LOCALIA: It is not necessarily that much money though depending on how it is configured. You could take an existing building and an existing location and designate that as a secure area and subject it to all the necessary access locks to make it secure and make sure it is not on a network and that would be your —

DR. MAYS: The hope would be that you could do what Jennifer is saying which is what we actually pushed for earlier is join in with the Census. The Census puts them in universities. It is already set up. It is just a matter of it is more as I understand now the issue of the fire wall. So, it may not be that we are actually, that they need to build buildings or build something.

MS. MADANS: The cost is really less in the physical infrastructure. It is really in the staff. You have to do the disclosure review or you have to do the disclosure review back at home which is sometimes easier actually to do it and then send it out, but there has to be someone there and it has always been the staff, but you are never going to get a data center in every corner. I mean it is never going to be in every university. So, I think there is some kind of balance between a better spread of the physical data centers, better remote access and these other mechanisms but I think the actual data centers and remote access are going to be more important when you do the targeted surveys where you have geography and it is geography that always really screws you up.

MS. MC CALL: May I raise a question as we go through and talk about data centers and things? My question is these are very specific recommendations around how to design a data and an information architecture. It is not all of it but it is certainly a piece of it and yet when I think about all the things that I deal with and run into every day I know that how those things will get architected(?) in the future both the short term and the long term are really going to be different.

So, do they want to hear from us a recommendation like build more data centers or do they want something that is broader than that?

MR. LOCALIA: You mean less specific, more vague but say to do something?

DR. BREEN: So, instead of saying expand secure data centers, increase options for access to data, something like that?

MS. MC CALL: Yes, laboratory technologies that could include data centers but there may be other things as well.

DR. BREEN: But I love that language, leverage new technologies.

DR. MAYS: Remember that part of what you were also thinking about is having things in places where the community, the people, etc., that are complaining about that they don’t have the ability to get; so, it is almost like it is the person also that is there not just the technology is there but the person like you know we heard the Census Center at UCLA where they actually have a community person that provides things, give it away, explains it, does junkets around town and I would suggest that NCHS could benefit and I may be wrong about the more that we say something like you know, data centers that the work that they are trying to do with the Census that it strengthens their ability to be able to get that and possibly spend money on that.

DR. BREEN: Why don’t we talk about in general leverage new technologies to increase access to data and then make the point that they need to be staffed and technical assistance needs to be provided even to the point of doing data analyses where you have an expert on site who can do data analyses for community groups or whoever need the analysis done?

MS. MC CALL: And it could be communities of practice are in there you know. There are brand new and this may seem like a tangent but there are brand new ways you can use very esoteric, they are called data hashing methods to preserve identity yet which could resolve some of the issues that we are talking about now and that would be a new technology for example that could come in. I just don’t want us to kind of set in stone that it has got to be necessarily a new data center. It may be broader than that. It may include other things. That is all.

DR. STEINWACHS: Let me summarize where we are because I also feel this obsessive-compulsive disorder coming on again. I think what this discussion has done is it sort of changed some of our thinking about the centrality of expanding number of data centers as the solution and so that what we are looking for is a recommendation here that talks about expanding access to these data and additional data centers is an important part of that. I think there is also in this the idea that there is the kind of support and assistance that can be provided by people and there is also the technology pieces, technologies that allow protection of privacy in ways that we have not been able to do before.

My suggestion is that I am not sure in the next couple of minutes if we can get all that crafted. Do you think Carol ought to do this? Carol, Vicky has suggested that if you would take some moments right now and write something. Russell has offered to do it and then, Carol you can check Russell out.

Okay, the fourth item here is the department is urged to design and implement a campaign or an educational campaign to inform racial, ethnic and linguistic subpopulations of the benefits of self-report and should partner with the private sector. This is to really expand the collection, right?

DR. MAYS: Because otherwise it is like we could have great methods and then if people don’t understand why then they are going to try to opt out as much as possible.

DR. SCANLON: I agree with that except just under the heading and I thought we were going to change the heading.

DR. STEINWACHS: It is almost now like it went in the first recommendation.

DR. MAYS: This should be in the first?


DR. MAYS: Okay, thanks.

PARTICIPANT: Under the very first recommendation.

DR. MAYS: Yes.

DR. STEINWACHS: And I don’t know whether you want to sort of have some way to talk about and maybe it is part of the first where you give it someplace where it is clear that there are perceived barriers or reluctance to share this. So, to overcome the reluctance to share racial and ethnic identity information, to have an educational campaign and move that up.

Okay, I am moving to C, improve data user training and under that there are recommendations one, two and three. Discussion of these?

DR. MAYS: Two, I included HRSA and SAMHSA

DR. SCANLON: On one I guess I was thinking that we should just simplify it, that when we talk about RFAs and RFPs, etc., and maybe begin with sort of this idea that we are going to have a broad audience that we talk about the department should support or fund initiatives directed to enhancing the capacity but the vehicle they choose to use is kind of their business but we really want the activity to happen.

DR. MAYS: Right, initiatives is a generic.

DR. STEINWACHS: Okay, so, simplify that.

DR. MAYS: Should fund initiatives?

DR. SCANLON: Support or fund.

DR. MAYS: Which?


DR. MAYS: Oh, support and fund, okay. Remember the whole thing at the beginning, if you need money we were told we were supposed to say it.


MS. MC CALL: We need money.


DR. STEINWACHS: We could say, and Carol says.

Okay, the next one, NIH, NCHS, HRSA, da, da, da, should renew and expand existing grant programs to develop the necessary training and expertise of researchers.

DR. BREEN: How do you define necessary?

DR. STEINWACHS: Maybe to improve or to enhance training and expertise of researchers from racial and ethnic and linguistic —

DR. BREEN: Do you want more researchers or is the thinking that the quality is not good enough?

DR. MAYS: It is numbers and then also it is like and we can leave it out here but it is also I don’t think that some people know for example, what are statistical methodologies for doing culturally competent research. We always think of cultural competence in terms of care. People have been talking about competence in terms of statistical methods even but I don’t think we should put it here because I think we would have to write a whole other section.

DR. STEINWACHS: Vicky, is this limited to members of minority groups or is it people who are interested in doing research on?

DR. MAYS: This was actually the minority groups.

DR. STEINWACHS: Because they are under represented in research.

DR. MAYS: But there are also all these mechanisms that are very specific to that but I think that mechanisms now may have gotten broader like you can have a minority supplement but it is not a minority person.

You could have —

DR. STEINWACHS: Is this also the idea that these are under represented minorities in research?

DR. MAYS: Yes.

DR. STEINWACHS: No, NCHS, NIH should renew or expand existing grant programs to increase —

DR. MAYS: You can say after subpopulations and others with interests in health disparities and health statistics because some of the program actually aren’t the people, like K awards. There are K awards that are specific in terms of disparities like what is different is the National Center for Minority Health and Health Disparities has been developed. That is its focus and it also is supposed to train individuals from the particular groups but many of the programs are also interested in others because they need to be trained to work with the population.

DR. CARR: This is one recommendation about two topics. I mean are we trying to increase representation of subpopulations among researchers? That would be one thing and the second thing is to increase research on subpopulations. So, we are saying two things. We are looking for two things.

DR. MAYS: I think we are talking about training two types of people. The first would be training people who are from these under represented groups and the second would be increasing the research expertise of others to also conduct research on this group.

DR. BREEN: It could be to, let me just say what I am thinking as sort of an example, because do you want new researchers that you are training from under represented minorities to be asking questions pertaining to health disparities or do you just on the one hand want to train people from under represented minorities and then also you want to train people to ask health disparities questions?

DR. MAYS: This was designed to have them trained to answer health disparities questions. It is not a matter of just training minorities say to do statistics but it was because we are trying to eliminate health disparities that you are training a cadre of people to participate in that and I think that is what the —

DR. BREEN: I have some language. Do you want to hear it?

DR. MAYS: Okay.

DR. BREEN: NIH and NCHS should increase the number of grant and other programs to train experts from racial, ethnic and linguistic subpopulations to research questions in health disparities and health statistics.

MS. RODGERS; But that only gets at the first part which is training people from those populations but —

DR. BREEN: But that is what Vicky said.

MS. RODGERS: But she also wants to train people from other populations as well.

DR. BREEN: Oh, you do?

DR. MAYS: Yes.

DR. BREEN: Okay.

DR. SCANLON: Maybe if you said something like to train researchers, in particular those above to you know so that you don’t preclude the idea of training people that are not in those groups.

DR. MAYS: No, the only difference is there are these targeted programs. I just want to make sure the targeted programs continue. At NIH you have programs where they specifically are looking for racial and ethnic minorities as opposed to like the other seems like it is generic and what I am saying is just the opposite which is somebody is detailed. They have to go look for those people. That is how the numbers increase,not that the program is there but it is almost like we said earlier there is somebody that knows that they are supposed to look for racial and ethnic minorities in this K award, in this other award.

DR. STEINWACHS: Okay, how about this? NIH and NCHS should renew and expand existing research training grant programs for racial, ethnic and linguistic under represented minorities with interests in health disparities and health statistics.

DR. MAYS: Okay, what about the non-minorities? We won’t worry about them?

DR. SCANLON: I was thinking that if you dealt in the first sentence with the idea of expanding the researchers with the focus on this area and then in the second sentence when you talk about these particular programs you could say that we want to make sure that these programs which are as you said targeted on minorities that they are continued. I mean I think we are talking about more than just renewing the existing program if you want to expand the population of these researchers.

DR. MAYS: Okay, we can fix this. If we have agreement now we are just word smithing. If we have agreement then I think we are okay.

DR. STEINWACHS: Okay, three, is there anything on recommendation 3? Should support existing efforts to develop new programs that train American Indian an Alaskan Natives and community-based organizations to use and analyze American Indian and Alaskan Native health statistics.

MS. MC CALL: That “and community-based organizations” seems like kind of a non sequitur.

DR. MAYS: And their community-based organizations?

MS. MC CALL: Okay.

DR. MAYS: And then I think apropos of —

MS. PAISANO: I think we need to have tribes or tribal government.

DR. STEINWACHS: And their tribal governments or and their —

PARTICIPANT: Communities?

DR. MAYS: So, is it not community-based but instead to talk about it as —

MS. PAISANO: The trust responsibility is to tribal government. Then you also have primarily in urban areas like community-based organizations and you have like intertribal councils. You know you have a different, several universes where there could be opportunities for this.

DR. MAYS: How about then if we say, “And their community-based organizations, tribes and tribal representatives”?

DR. STEINWACHS: Programs that train American Indians, Alaskan Natives and their tribal and community organizations.

DR. BREEN: This group has been separated out from other under served groups. So, there was a purpose in that, right? Was that because you wanted to link them not just as individual researchers but to their communities?

MS. PAISANO: Because their whole setup is very different.

DR. BREEN: Okay, so that is what we are trying to get at here?


DR. SCANLON: You are talking about training researchers.

DR. MAYS: No, that is why we are saying it is not researchers —

DR. SCANLON: But we have got American Indian and —

DR. MAYS: American Indian and Alaskan Natives —

DR. STEINWACHS: To use and analyze but not to do research.

DR. SCANLON: But that sounds like everybody.

DR. MAYS: I think anybody there that wants to learn how to use their data better. I mean they are structured so differently. The reason is based on the hearing. They are structured very differently than other communities. So, therefore it is not always about a researcher. They have people who do research who have to bring it to the tribal council. The tribal council needs to understand the statistics to be able for example to better use the data that they were given.

MS. PAISANO: It is to train American Indian, Alaskan Native, well, to me it is tribal governments, American Indians, Alaskan Natives, tribal governments, intertribal councils and community-based organizations because in my mind community-based organizations really fall into more the urban areas and the need for urban American Indian, Alaskan Native data is very critical.

DR. SCANLON: Why don’t we try to be not specific about who is being trained but that we are training for a purpose, that we develop a new training program for American Indians and Alaskan Natives, the list you just gave, can use and analyze the data, something along those lines?

DR. BREEN: I was thinking something like that, too, because I think the last two sentences actually state the purpose to foster the development of evidence-based practices that can reduce or eliminate health disparities in this population and then to establish a reporting system to track progress in meeting the goal and so who gets trained to do it could be left open because the important things I think are to get the job done.

DR. STEINWACHS: Is that agreeable, Justine?

DR. CARR: Okay, so, I am just going back to first we are going to build good data. Then we are going to utilize data and so this is one of the ways that we are going to utilize, but is this the only population that should be singled out to use evidence-based?

DR. MAYS: No, it is not about evidence based practices.

DR. CARR: But I mean why not Creole or all the others? Is it the structure?

DR. MAYS: It is several things. It is their relationship. For example, we can’t dictate a lot of the data stuff to them the way we can with other groups because they have a different relationship with the government. Too, most of their data actually for the health statistics comes only from service data in terms of Indian Health Service as opposed to other places that we collect some of the data. I mean it is —

DR. CARR: So, we are just sharing expertise on how to use data and their data set is different.

DR. MAYS: Some of their data set. I mean they are in some of the surveys but some of the data that they rely on predominantly is not under our jurisdiction. It is under the tribal jurisdiction and they need more resources. It is like for us a researcher is probably there who would say what to do if somebody was worried about it. It will be a person who is the head of the tribe will make a decision about what to do and they need some trained —

MS. PAISANO: I mean it is to fulfill the treaty obligations from the Federal Government.

DR. STEINWACHS: So, I think if there was a, you know I think what Justine is saying which I think makes sense is that we need to make the reader understand why and maybe this is as Edna is saying it is to fulfill treaty agreements or treaty obligations, fulfill treaty obligations in a recognition of a special relationship of American Indians to the US Government and so have that kind of lead in I think and then the two last sentences there I think certainly are a way to try to reduce disparities but maybe that is less important here than just to say that the reason we are doing this specially and in a different way is because of those two things. Does that make —

DR. MAYS: Yes, I was going to suggest that maybe we drop the such efforts based on —

DR. STEINWACHS: Yes, why don’t we drop that.

DR. MAYS: It is almost like we are promising the world and let us just get rid of that but a reporting system should be established is to be kept.

DR. STEINWACHS: Okay, that is fine with me. Do others agree?


DR. MAYS: We may touch base with you about this.

DR. BREEN: There is only one more thing and that is that the hearing on American Indians made it clear that they have very inadequate funds and until they have a lot more funds they are going to be delivering services. They are not going to be doing data collection or data analysis or really worrying about monitoring much of anything. They would much rather spend the money on service provision.

So, maybe we should put in there something about money, and we could also cite that what is it called, is it the 21st century report? It was a task force on American Indians that was distributed at one of the meetings.

MS. PAISANO: There is a restructuring report.

DR. BREEN: Yes, and it really went into quite a lot of detail of the deficiencies in the Indian Health Service and where funding was needed to beef it up. I don’t know if it talked about research at all.

MS. PAISANO: There are, also, the civil rights reports.

DR. BREEN: It was out of the Office of Civil Rights.

DR. MAYS: Remember we are going to do a whole different letter. So, I think we may want to expand and put that in the letter and here it is just merely about health statistics because these could be health statistics from NCHS.

DR. STEINWACHS: And if you wanted to deal with it you could develop new programs that train and support the use and analysis of. You are asking DHHS to do new programs that train but you could also ask them to support the analysis and use of data and so that might help a little. I am taking you on to improve capacity to collect data on geographic and socioeconomic position, geocoding, comments on these? They lead up to what is now on my Page 23 to the recommendation of the department is urged to convene across federal agencies a group to examine what assessments of socioeconomic conditions are currently available from federal data collection efforts and what linkages are possible between health and other sources of economic data and to disseminate this information through the department’s gateway.

DR. MAYS: On 23, where it says, “Information about variables such as,” I did add discrimination. That has become a very big issue.

DR. BREEN: I don’t want to take too much time here but the socioeconomic position data, did other people find this a little bit confusing? I thought it could potentially be a little bit confusing in the sense that we are moving from something that granted it is pretty primitive but we do know how to measure, that is socioeconomic status into something that from the text we clearly don’t know how to measure socioeconomic position and it is a much more complex concept which is probably going to get at things much better in the long run but I was just concerned. I added a few things here to maybe clarify a little bit and ground it a little bit more, but I mean the socioeconomic position stuff is quite new and I think it is appropriate to bring it in but it must needs to be brought in in a way so that it doesn’t sound like we are clueless about how to measure it which I think it sounds like now.

DR. STEINWACHS: Do you want to read us what you have added?

DR. BREEN: I just put in some references and then I put in up in the paragraph that is titled socioeconomic position on the fourth line down, fifth line down, sorry, this expanded view tries to conceptualize a process and is referred to as socioeconomic position because it is; it is process, and it is going to be hard to measure. I mean it is going to require longitudinal data, panel data. It is going to require looking at a much more complex set of variables than we look at now and like I said, I am supportive of it and I think we should do methodological work on it but I guess I feel like I am not sure, I wasn’t sure where this is going in terms of advocating.

DR. CARR: I wonder, also, we made the comment earlier about is this an educational document or is this a call to action document and a lot of this because it is new needs explanation. We have a lot of education.

DR. STEINWACHS: So, Vicky, tell us what we should be doing?

DR. MAYS: Okay, one, I went to the OBRSSR, is that what it is, web site? They have had a meeting on this. I actually attended the meeting which is why I was looking for the document. They talk about socioeconomic status and the need to move the socioeconomic position but some of the things that you see in the paragraph that says, “Collecting data that include health insurance coverage, employment, status within occupations,” it is like we don’t have a perfect method.

We haven’t decided what a standard measurement of socioeconomic position would be, but we do know something about what kinds of things we want to look at. I think that is the distinction.

DR. CARR: So, is this recommendation 1 or 2? It is a data element.

DR. BREEN: And also I wasn’t looking so much at those details. I was looking at current wealth, historical contextual variables such as socioeconomic conditions during childhood. I mean those are things that we don’t routinely collect and we don’t really have much about..

DR. CARR: I think we have current state and future state and in current state we don’t have this definition fully worked out and we also don’t have access to the elements that will build definitions. So, I think it would be contemporaneous to mention it but to not put it in the call to action until we know what action is called for.

DR. BREEN: Or to say that we are transitioning or something like that rather than sort of saying you know we have this old way and the new way because to me I am reading this and thinking, well, we should be collecting information on socioeconomic position and that is what this document is saying.

DR. MAYS: But I think we have a possibility, and I am getting a little nervous because it is almost like what I don’t want it to be is until we work it out that we then do it. I mean if that is the case let us take SEP out and just call for linkages to give us other data on things like education.

MS. MADANS: Those things we have. You know, I think you are talking about two very different things. I guess my training was this is all socioeconomic status. If you want to call it something different, okay, but the list here of insurance, employment, maybe not worker, manager, supervisor, occupation, education, income, poverty, wealth, we know you do a great job on wealth but it is in almost, these things are in almost every one of these major surveys. So, I think the big difference is this generational perspective.

DR. MAYS: And the historical context.

MS. MADANS: But you have to get that independently almost from the survey. You wouldn’t ask people about historical context. So, I think we can that we should expand and there can be some general statement about expanding what you might include and the kind of covariates but you don’t want to lose the fact that you have to keep collecting the stuff you are collecting and do a better job of it. Income and wealth are notoriously hard to collect, and you don’t have to make this grand thing, you know, we don’t know what is in position; we haven’t made that. We can still say that we need to do a better job of collecting these covariates and move into these newer areas. I think the linkage issue is completely different.

DR. STEINWACHS: I have got Russell and we do need to find a way to bring closure to this.

MR. LOCALIA: What I did was starting on Page 18, I retitled it access to data on subpopulations and then they require that we do title page A on Page 16. So, here is where I start on Page 18. The first paragraph starting although the department collects, is the same down to the last sentence, potential users. Then I stop. The Bureau of Census is out and I will tell you why.

The next sentence starts with the subcommittee urges that any solution should strike a balance between individual privacy and confidentiality. So, most of that paragraph there is now the next paragraph. It ends then with other populations. Then I have three subpoints which I will read the first two of them. The third one is the same as the current fourth. Is everybody following? Based on what you said earlier it does say that the department should continue to work with the Bureau of Census to place or share additional secure data centers in academic and community centers.

MS. MADANS: Take out the continue because you don’t know we are working.

MR. LOCALIA: That is what I needed to know, should work? Okay, funding should ensure that all centers are adequately staffed and supplied with current hardware and all necessary software so that interested users will have ready access to data needed to assess the health of subpopulations.

Two, the department should consider additional options including disseminating and providing technical support for synthetic micro data and promoting research on other forms of protecting identifiable data on subpopulations. These and other emerging methods can increase access to data on subpopulations for research and statistical purposes in a manner consistent with existing statutes and regulations and in alignment with the expectations of privacy and potential response.

Three is the same as current four and that is it.

DR. MAYS: Can I just, not just synthetic but is there any reason if perturbed data is available not —

MR. LOCALIA: I didn’t want to get into the — I just have to say that the synthetic data approach is better than the perturbed data approach to the best of my knowledge. I think the perturbed data most people would say is not as good as the synthetic data approach, but that is why I have these and other emerging methods.

DR. MAYS: Okay, then that is fine.

MS. MADANS: You didn’t mention data centers.

MR. LOCALIA: That is in two. It just talks about to place or share additional secured data centers in academic and community settings. So, I kind of, let me see where the remote data centers shows up? At the end of the first paragraph there is still the sentence of secure data center currently which severely reduces the access to data supplied by potential users. So, I haven’t said anything about remote access but we could put that in someplace or somebody could put that in someplace. So, what I have done is I have cut out a bunch of things because I thought they could be moved around a little bit but I don’t think I really eliminated anything. I think all the pieces are there including the working with the Census for additional locations. What I would suggest is I give this to somebody for additional —

DR. STEINWACHS: Why don’t you give it to Ann.

MR. HITCHCOCK: I had sort of a flash on the only one data center exists. Does AHRQ have a data center?

MS. MADANS: They have a data center for just the —

MR. HITCHCOCK: But I mean does NCHS? I don’t know that we can say that only data center really exists across the department.

MR. LOCALIA: Do you want to say that only limited or —

MS. MADANS: Extremely limited or very limited.

MR. LOCALIA: Only very limited numbers of centers currently exist. Hold on a second, limited numbers of centers exist.

Remote access helps out, that would go in paragraph, okay, paragraph one is should work with the Census to place additional, funding should adequately staff, do you want to add a sentence there about something, options for remote access should be expanded or something? Is that all right/

DR. STEINWACHS: That sounds good.

MR. LOCALIA: Then the second —

MS. RODGERS: Russell, why don’t you e-mail it to me? I will give you my card.

MR. LOCALIA: Yes, I can do that.

Then the second paragraph says, “Consider additional options which would be technical ones,” and then paragraph three is now four which says, excuse me, I didn’t touch that at all, “The department is urged to design a campaign.” I didn’t touch that at all.

DR. STEINWACHS: Campaign was going to move because that really deals with —

MR. LOCALIA: Delete three?

DR. STEINWACHS: Why don’t you delete three from that.

MR. LOCALIA: Is that okay with everybody?

DR. MAYS: Yes, very good.

DR. STEINWACHS: Don’s obsessive-compulsive behaviors are rising again. I promised Vicky that she could leave in 26 minutes and when Vicky leaves, you know, so, I think we have got a problem here about how far we take this discussion of socioeconomic status and socioeconomic position. So, Vicky, help me.

It sort of sounds like I mean Jennifer and some are sort of arguing let us keep it brief, let us recognize the importance of those data and maybe it stops there and so we wouldn’t get into a long discussion of sort of trying to educate people about it but just recognize its importance.

DR. MAYS: We will cut that section to be recognizing what is, where the field needs to go and to encourage — to me what is the bottom line here is one, the linkage with other data sets that have better socioeconomic status data, first of all.

MS. MADANS: You are talking about direct linkage with other data sets?

DR. MAYS: If possible, yes, because we have problems with income.

MS. MADANS: It is very limited, you know.

DR. MAYS: But I think that that is what was talked about is to be able for example to work with Social Security Administration, to be able, Gene is not here but to work with IRS.

I mean those are the kinds of things that are being asked for to explore those things happening.

DR. STEINWACHS: So, what I think you are saying is that one of the major challenges or one of the problems to overcome is the limited availability of information on income.

DR. MAYS: That is one, yes.

DR. STEINWACHS: Is that the major one because these other ones I think Jennifer was trying to make the point that a lot of these are routinely collected. Income is very hard to collect when you try to collect it in the sense of response rate and yet it plays a central role.


DR. SCANLON: I was going to say that the IRS is not an answer either because we are concerned about a population of largely low-income people in many instances and they don’t file or they may not have complete records.

DR. BREEN: If they don’t file then you know they have an income below a certain baseline.

DR. SCANLON: The IRS is a very skewed picture of income because of the tax laws and there are huge issues in terms of access to that information.

At JO we had access to virtually everything on an unlimited basis except for tax records you had to go through a whole separate process to get information on tax records.

DR. STEINWACHS: So, the recommendation that comes out of this section maybe could be sharpened and say that the department is urged to convene across federal agencies a group to examine what assessments are currently available on socioeconomic positions.

Maybe ways to improve information on socioeconomic position, particularly on income, I don’t know. I am trying to find a way to sort of get you from —

DR. MAYS: I think we have to rewrite it.

DR. STEINWACHS: Why don’t you rewrite it.

DR. BREEN: I would be happy to help you with that.

DR. MAYS: Thank you.

DR. STEINWACHS: That allowed Don to go on to Page 24.

MS. RODGERS: Actually could I raise one other issue? We don’t have a strategy about geocoding here. We have only got one about —

DR. STEINWACHS: There was nothing said about geocoding in our recommendation.

MS. RODGERS: Right, but we have got a whole paragraph.

DR. STEINWACHS: So, either we probably ought to say something in a recommendation or —

DR. MAYS: Is that one where it belongs in one rather than here?


DR. MAYS: I think that that is one because we actually talked about it earlier and I think that should go to one.


MR. HITCHCOCK: On Page 20, geocoding data, we need to make sure that we add Alaskan Natives to the American Indians and there is an example of a population group that is going to be very important. We don’t really understand them all that much.

DR. STEINWACHS: Alaskan Natives to American Indians on Page 21?

MR. HITCHCOCK: Page 20 on geocoding.

DR. STEINWACHS: Is it Page 20? Oh, you and I have different page numbers. We are going to move that up into recommendation one category and then I —

DR. MAYS: I think again we need to take care of that.

MS. MADANS: It is not only that you want the information. This said, “Increase the methods for using,” in addition to collect the data.

DR. MAYS: Where are you?

MS. MADANS: I am on the geocoding paragraph, second paragraph on my Page 22 and it says, “Increasing methods for the use of accurate geocoding while at the same time maintaining is necessary.”

So, I thought the reason the reason this was in two was because the focus was on the methods not in addition to the fact that we need to geocode the data and develop methods to use that. It is not straightforward how to use that stuff sometimes.

DR. STEINWACHS: So, maybe it ought to stay here?

MS. MADANS: Maybe you don’t need all this verbiage. It could be shortened. I don’t know but I think if you want a recommendation you could change that into a recommendation.

MS. RODGERS: Maybe move some of the explanatory stuff earlier in recommendation one and then just keep this as the —

DR. MAYS: What is the actual recommendation?

MS. MADANS: To increase methods you know develop methods for the accurate use of geocoding data.

MR. HITCHCOCK: I would say to expand the use of and then what you said.

DR. MAYS: When you say, “Develop methods,” can you just tell me what sentence you are at?

MS. MADANS: I am right above socioeconomic position data. It says, “Increasing methods for the use of.”

DR. STEINWACHS: Okay, so we have a recommendation.

Next one is linking data systems and dissemination methods to bridge old and new data.

DR. MAYS: Jennifer, does that still need to be there?

MS. MADANS: I don’t know. You tell me.

DR. MAYS: I think it is not any longer that they need to know about the methods but what they need to know is if there needs to be greater dissemination about the work that you all have done but I just read the article that came out by Jennifer Parker and you are on it and so then it made me rethink the whole thing. I brought it with me on the plane and so I had a quick read of it but I mean it was like a really great article about helping people to understand how to do the bridging and how to get better information on race and ethnicity. You just opened it up again.

MS. MADANS: We can’t leave it alone. I don’t think it is a problem to leave it there. If you want to cut it, it is not, I mean you could always say that we should provide more technical assistance.

DR. MAYS: We did.

MS. MADANS: We have this humongous bibliography that nobody knows about.

DR. MAYS: Okay, then we will cite it. The article is very good. So, I would like to at least, let us leave it I will cite the article and then if you will look at just that I will ask you to tell me if there is another particular point drawn from the article that you would like here because it is very good.

DR. STEINWACHS: So, the recommendation is to leave this section as it is unless there are other specific questions or comments.

Moving to F, reduce missing and inaccurate data on subpopulations

MR. HITCHCOCK: Telling the Secretary what happens when you move the order of questions around is a little more detailed than he is likely to absorb. Again, we are doing some sort of educational thing here that I don’t know. At this point in time I would say, “Leave it in there,” but if you asked me this morning I wouldn’t have said it.

DR. MAYS: Then maybe it needs to be phrased differently because what this is really about is the Latino choices, what happens in the other, how we end up with missing data. So, I don’t know if this needs to be flipped around so that something else is what really needs this as opposed to it is almost like if people said, “What is the most important?” it may be that we gave you the details about the issue and not really what is so important about it.

DR. SCANLON: But is it also again maybe a part of one?

DR. STEINWACHS: It is part of one.

DR. SCANLON: And something that doesn’t necessarily rise to a caption? It is another one of these things about getting better data.

DR. STEINWACHS: Okay, so if we are going to keep it, it is part of one and the idea is to keep it at this point but move it to one and —

DR. BREEN: Would it be too broad to frame it as improved data quality because for somebody who is not really into this to say, “Reduce missing and inaccurate data on subpopulations,” sounds like a no-brainer.

DR. STEINWACHS: It is data quality and completeness or something like that.

MS. MADANS: You probably don’t need to explain what that is. You can just say, “Improve data quality,” and then you have the —

DR. STEINWACHS: And we skip over my frustrations and you are even better yet. I never like to be frustrated.

Okay, this is Don’s race to the finish and not necessarily the best thing but —

DR. BREEN: Dissemination.

DR. STEINWACHS: That was where I had to get. I know that. Leslie told me I had to get here or else. Increase dissemination of data or health information. I don’t know whether it is the data or the information.

DR. MAYS: Before it was like data, but here it might be the information because it is almost like the data has already been analyzed is what we are talking about here. Before we were trying to get data out to people to use. So, I think it is probably very important here to increase dissemination of —

DR. STEINWACHS: Information on health disparities and —

DR. MAYS: Of health data, I don’t know.

DR. SCANLON: At the end of No. 2 we almost get to talk about information as opposed to the data because we are talking about media.

DR. MAYS: Suppose we say, “Increase dissemination of data and information or data products,” or something like that?

MS. MC CALL: That would be the result of the work, right?

DR. MAYS: Yes.

MS. MC CALL: So, we need to call it something other than data because when you first look at it it just seems like it is what we already talked about. It could be research and findings.

MS. MADANS: The beginning of two sounds the same. It is not until you get to the end.

MR. LOCALIA: If you want to distinguish between data and the results of the analysis it should be increase dissemination of health statistics on whatever you want to talk about not data because data is the individual modes of observations but the statistic is a data summary. So, that is one thing we could do.

DR. MAYS: Increase of what?

MR. LOCALIA: Increase dissemination of health statistics on disparities or subpopulations or, does that help?

MR. HITCHCOCK: I thought Jennifer’s problem was recommendations leading right into data.

MR. LOCALIA: Then you have to do something about the first paragraph. You may want to trim the first paragraph and then make sure that the rest of it is consistent with the summary of the data which would be —

MR. HITCHCOCK: I like just data of information or something like that. We want to disseminate data to certain groups.

MR. LOCALIA: That is covered earlier.

MR. HITCHCOCK: Dissemination earlier?

MR. LOCALIA: No, access.

MS. MC CALL: Drop the word “data,” and I think different terms like information, findings, research, statistics, something.

DR. MAYS: All of those? Can you say that again, Carol?

MS. MC CALL: Sure, information, research, findings, statistics.

MS. RODGERS: How about health statistics and research findings?

DR. MAYS: Then are we moving recommendation one?

DR. STEINWACHS: It might make sense to move recommendation one in with what Russell is working on.

DR. MAYS: This is a release.

MR. LOCALIA: This is public use and so it is different whereas that was data for statistical purposes.

MS. MADANS: Someone before, I forget who said it, that is all part of the data release program and we are accessing various mechanisms to access data some of which are releasing data, some of which aren’t releasing it but providing access, but you would have to rework that whole section.

MR. LOCALIA: This is talking about summary data whereas that is talking about —

MS. MADANS: Except that like the recommendations are all about data tapes. So, if you want to make this about the information, the results of analyses there probably is missing pushing that you need to have more easier access to the actual data, whether it is through the web or CD-ROM and also the conference. I think there was something in here about having user conferences and stuff which would apply to releasable and data that you just access and so you probably have three things you are talking about.

MR. HITCHCOCK: What about release of information? What do we want to call it to different forms of media like in providing that to Hispanic media or —

DR. MAYS: We had that.

MR. HITCHCOCK: Is that in here somewhere?

DR. MAYS: We had that earlier about educational campaigns.

DR. BREEN: You have got it at the end of No. 2 here as well.

MR. HITCHCOCK: No. 2, yes, all right.

DR. BREEN: I don’t think the recommendation is very clear. If I were the Department Secretary and had to wade through that first recommendation I think I would give up.

DR. MAYS: Let us try and see what we can do. I think some of this moves to the earlier part.

MR. LOCALIA: I would say some of it can’t be moved to the earlier part especially the business about CD-ROM. Anybody who has had to deal with Medicare knows that CD-ROM made a great improvement but some of that can be moved to the other section because that is simple ease of access to data, but I would say that this section probably ought to focus on the title which is now health statistics and research findings because then you can carve out the pieces in each paragraph that relate to that and keep them here. So, that will trim this section.

DR. STEINWACHS: Each one of you pat yourself on the back. You have made it through this report and I want to again thank Vicky and Ann because I know it is not over.

Vicky, or Ann could you just say a couple of words about the time frame? The complete draft of this report does go to the Executive Subcommittee for their review and that is sort of the next I think critical time point here.

DR. MAYS: My understanding of what the next stages of this are in terms of what we outline is we now have the dates in my head, except the assignment date as I keep calling it which is we drop dead by the twenty-fourth but we would like it a little bit before then.

We need to work on making the changes that you have indicated. We are going to try to do is early next week we should get the next copy back, the next draft back to you. Maybe what we ought to try to do is you can say whether you want that or not, track changes in so that you can actually see where we have made the changes that you want.

(Thereupon, at 2:55 p.m., the meeting was adjourned.)