[This Transcript is Unedited]
Hubert Humphrey Building
200 Independence Avenue, SW
Washington, D.C.
Agenda Item: Review of the Process, Products and Timetable
DR. MAYS: Good morning, everybody. Thank you for being here. Let's start by introductions around the table, so that everybody knows who is here.
(Whereupon, introductions were performed.)
DR. MAYS: Thank you, everybody who is here, and I've gotten word there are some other people on their way, so I think we should be good to go in terms of the number of people that we will have working with us today.
What I want to do is to start -- if you have the agenda, what I want to do is start by saying what it is we are going to try and accomplish today, why we want to do what we are doing, and by the end of the day where I hope we are going to be, by the time we do the wrapup.
There are two things that we are going to try and accomplish by the September meeting. Of course, the deadline is such that they should be circulated before the September meeting. It was one of the decisions that was made at the executive subcommittee, that when the recommendations that are going to go forth, what we are going to try and do is to make sure that we submit them a little bit earlier to the executive subcommittee. That will allow time for some preliminary discussion, and help us identify any problems or issues that need to be addressed before it goes to the full committee.
A letter is actually what we are going to focus most of our time on today, followed by some time on the report. In the letter, our hope is that we will come up with a set of recommendations that address the issues that we have identified in the population-based surveys.
Important to this letter is that our recommendations are feasible. We don't want to send guidelines to the Secretary, and none of this can be done, and it is all a great idea, but instead, we really want to try and come up with recommendations that can be moved on. It doesn't mean that they may not be difficult and challenging from the perspective of budget and handling the issues, but I do want people to really hone in on things that we think can be done, as opposed to just on the record.
The letter we want to keep short and to the point. That is why it will be backed up by the report. One of the things that is going to drive us is that we have the opportunity to make a contribution, and that contribution is that there are individuals in the wings who really hope to see recommendations come forth that they can use.
For example, there are health plans that address the issue of the collection of data on race and ethnicity and health plans. They might be able to utilize our recommendations to actually encourage their particular health plan to move forward on this.
The other thing that we want to capitalize on is that -- and Harvey can probably speak to this -- is that the Agency for Health Care Research and Quality should be releasing their report on health disparities in September. Do you know when in September, Harvey?
DR. SCHWARTZ: I don't know the specific date, but certainly by the end of the fiscal year.
DR. MAYS: That is until the 30th, right?
DR. SCHWARTZ: Right.
DR. MAYS: So we can count on that report. Both the disparities report as well as the quality report are going to be released by that date, right?
DR. SCHWARTZ: Yes.
DR. MAYS: What we can count on is that the disparities report at least will get the attention, as I understand it, and we talked a little bit about this on one of our conference calls, -- that DHHS is quite interested in the issue of disparity at this point in time. So it would behoove us to try and when this activity is moving along, to have our recommendations also to be on the table, so that they will also be considered in whatever actions are planned.
So that is part of why we want to make sure that we get these two things finished and hopefully approved. That part we can never predict, but hopefully improved by the end of September meeting.
Marjorie?
DR. GREENBERG: So you are thinking in terms of two letters?
DR. MAYS: One letter and one report.
DR. GREENBERG: One letter and two recommendations then? One relating to the targeted surveys and one related to collection of race-ethnicity by health plans?
DR. MAYS: I think that is the discussion today. I prefer to see if we can try and get them into one letter, unless it doesn't seem feasible. There is a way in which to talk about some of the health plan issues as larger population-based information, as opposed to population-based surveys.
DR. GREENBERG: Okay, because I have been having this e-mal exchange with John Lumpkin, who feels that the latter, on data collection by health plans of race and ethnicity, that the committee really needs to try to get a recommendation before it on that for the September meeting.
DR. MAYS: Right, that is what is spurring us on. John and I had the conversation, so what I want to do is wrap this into -- there is a way which I think we can do it. If it doesn't seem that it works smoothly enough in the letter on population-based surveys, then we would do a separate letter. But I would like to try and see if we can make it one letter.
DR. GREENBERG: With the general problem and the solution is.
DR. MAYS: Exactly.
DR. GREENBERG: Is the report also supposed to be ready by September?
DR. MAYS: We are going to try and have the report ready. I think there is a priority on getting the letters out, because the letters are clearly -- I have been told that letters get focused, as opposed to reports. Reports sometimes go, thank you and we'll read this later because it has a lot in it, but I would like to think we can get the report.
We have a working draft on that, but what I think we don't have is a better sense of how to deliver the information in a way in which it is going to have impact.
DR. GREENBERG: I just would like to request that by the end of the day, we have guidance for the other writer. I don't know whether what she was working on was supposed to be part of this report or a different report, but we need to come to closure with her.
DR. MAYS: Okay. She would in the report do editing.
DR. GREENBERG: She has also written some stuff on history of the past, and that is kind of in limbo at the moment.
DR. MAYS: Right, because Olivia actually did part of that, so it is like it would be editing in terms of integrating some material in.
PARTICIPANT: I already sent that information to her, so I think the latest version that she has incorporates what I had sent to her. We had talked several months ago.
DR. GREENBERG: I had gotten a letter from her which -- we need to come to closure with her. So we can discuss this outline or whatever, but it is just one of my objectives for the day. I told her I would try to get it resolved at this meeting.
DR. MAYS: Okay. Well, since I don't have what you have, it would be helpful to have what you have, so that as I plan, I can have a better sense of how to achieve it.
Questions, any other questions? Good morning, Nancy.
DR. BREEN: Good morning. I apologize for being late. I am Nancy Breen, I'm with the National Cancer Institute. I got off at the wrong subway. Pretty bad for a native.
DR. MAYS: Anyway, I think then we are getting some clarity about what we are hoping we are going to achieve by the end of the day, which would be a draft of the letter. We have an outline for the letter. The other would be a sense of where we want to go in terms of the report. The third thing would be a sense of how we are going to accomplish this, and what the time frame is.
Did someone join us on the phone? That might have been Jennifer getting off.
So let's turn to the agenda, and talk briefly about what some of the issues are for the letter. Then we are going to move to talking about the letter and the report in terms of its scope.
The first issue is, how comprehensive will the recommendations be. We have struggled with this. As you know, in the population-based report we have -- is it 61 we are to? I think we took like maybe four or five of those out, so we have a number of recommendations. In the letter we don't intend to try and address all of those, but I think we do today need to struggle with exactly what are the probably three or four or maximum five most important issues that we want to put before the Secretary's attention.
Two, -- and Marjorie started us on this discussion, and that is whether or not we are just going to talk about population-based surveys in the letter and then do something different in terms of talking about the collection of data in health plans, or whether we are going to try and wrap them into the same letter. I'd like to see if we can't wrap it into the same letter. The reason for that is that I think that it bring us back to talking about population health, and just talking about different methods and ways of collecting data in terms of population health, some of which are under the auspices of the federal government, and some of which would be under the auspices of private entities like health plans. The government has health plans also, so I don't think it is a far stretch, but let's keep that as one of the things we want to do.
We also need to come to some decision about the characterization of how we are referring to these surveys. I have given you various things that have come up; thank you to several people who have given us some titles to use. We've got targeted population, priority populations, family of community surveys, geographical area surveys. It isn't choosing a title, but it is actually choosing how we are characterizing what we think is important, so that the Secretary's office understands it, but also more importantly, that agencies can see and understand where we are going with it. So if this is a part of their agenda in the strategic plan, that they also then can see a role for participating in a movement towards doing these surveys.
The hard thing that we also are going to struggle with is what subpopulations -- or how far down, I should say, not what subpopulations, because I think our hearings have helped us with that, but how far down are we drilling into these populations. Are we making suggestions to the Secretary as to which groups? Are we making any suggestions as to any ordering of what needs to be done? Some of that, I am of the opinion, can be political. What I hope is that we are going to start by talking about the science of it, and then letting someone else get in the politics of it, unless there is some reason that we need to talk about politics.
When we talk about the subpopulations, it really raises a lot of issues. It raises issues about feasibility, and that is feasibility not just in terms of cost, but feasibility in terms of whether it can be done. To some extent, things aren't done because it is so difficult to do. So we really need to think about whether or not what we are suggesting is doable.
We also have to think about this issue of the post-OMB classification of race and ethnicity in terms of multiple race. We have to think about, when we make these recommendations, what that means. We are no longer just dealing with subgroups as having maybe one racial ethnic group, but individuals -- if we look at the Census data -- have identified with more than one race. In making the recommendation to the Secretary, I think we have to think about how that issue is handled.
So we have some tough issues, and we have to grapple with these, else we send forth recommendations that I think are just mom and apple pie. If anything sticks with me, it was Jeff's comment about, these need to be sophisticated recommendations. So we do need to grapple with this today.
Questions, comments? Any issues you want to add for us to be thinking about as we go through the day?
DR. SCHWARTZ: Maybe I am coming into this in the middle of things, so what I am going to comment on now might be a little off base because of coming into the middle.
One of the things, if it hasn't already been discussed, that I would appreciate understanding a bit better would be the management of any additional survey endeavor, that is, what are the range of resources, inputs, what kind of potential outputs would any type of survey endeavor cover, and to what extent would the recommendations cover issues such as where this type of initiative would be managed and who would be responsible.
DR. BREEN: Are we just putting that down as something to discuss?
DR. SCHWARTZ: That is what I'm not clear about. In particular with respect to what we might call special surveys, I am assuming, and correct me if I'm wrong, that in one sense, if there were special surveys, that in the sense they would be ad hoc, or there would be some type of specific character of a project to be carried out, and that the surveys currently are not produced, but there is some type of special interest. If there is, that needs to be expressed. Then in addition, I would think if it is feasible, there needs to be some type of entity or organizational structure to produce it.
Now, if I am off base, please correct me, but these are the types of things I'm not clear about.
DR. MAYS: Harvey, let me make sure I understand what you are asking. Do you want us to discuss in terms of putting forth recommendations who we think should do the survey, how frequently we think the survey should be done?
DR. SCHWARTZ: That might be part of it, but my suggestion would be, if this is beyond the scope of the nature of the recommendations that the group should be making, my suggestion would be that we consider making the recommendation to the Secretary that would suggest some other mechanism or body whereby such type of decisions might take place, because that would make it more feasible in my view. That is just my opinion.
DR. GREENBERG: Well, generally I think the -- I don't know that I understood the first part of your question, but just going to the second part, I think generally, the committee tries to avoid telling the Department or suggesting exactly where something should be done. Since all of the recommendations go to the HHS Data Council, that would probably be the venue for sorting that out. Although, depending on what one recommends, there are some obvious candidates. Like, if you are talking about community HANES, people who already do HANES probably are the logical ones. If you are talking about a special MEPS type survey, then that would be likely to be done by -- because I think just in the outline, it does mention that whatever information that is collected should be compatible with and consistent with national data, so that comparisons can be made, et cetera.
If it were a survey that is unlike anything anybody has done in the past, but I don't think that is probably what you are talking about --
DR. SCHWARTZ: That is what I was asking. I'm not clear if it is necessarily the case that our group is going to recommend that any special survey be conducted that is tied into a currently designed and shielded survey.
Now, if we already made a decision that that is the case, then I'll retract the first part. But I thought the issues were sort of open. So I guess another way of raising my question would be, are there particular surveys that there has been some type of systematic or comprehensive review, such that we are at a point where we can suggest that there be special surveys tied into one or more of those, or are we talking about something new?
DR. MAYS: I think what we want to do is put your question in reserve. We have for example several people who are visiting us between the 10:30 to noon period, who have done some thinking about this and who will share how this might happen.
DR. SCHWARTZ: With their particular surveys?
DR. MAYS: Yes. Then I think it is up to us to move forward with the recommendation.
DR. GREENBERG: Just one other thought, and this may not have been where Harvey was going, but recognizing the limited opportunity or new resources, I guess one thing the subcommittee should think about is whether there is any -- based on testimony received or analysis done, may want to suggest that perhaps some of these surveys, because of their importance, maybe should be done in lieu of some national survey in any given year. I don't know whether that -- I'm not recommending this, although it has been recommended in the past. But the fact is that if you are talking about Jim saying this is not just apple pie, we need recommendations that have some chance of being adopted and a zero sum game, that is clearly one option that needs to be considered.
It has been considered in past years, I know, going back to the Hispanic HANES, when the Hispanic HANES was done, it was not done at the same time the national HANES wasn't done.
So this is just something to think about. I don't know whether you feel you have the information or the basis to make those kinds of recommendations, or even want to. As I said, I am not recommending them. But it is just something to think about, that we are dealing in an environment where we are pretty much at a flat subject.
DR. MAYS: That is hard. Thank you both for putting it on the table.
DR. SCHWARTZ: I'll comment very briefly then. I'm not recommending one or the other. One of the things that I am focusing on is similar to what Marjorie just mentioned about the zero sum game, and particularly the letter of May 30 from Secretary Thompson to Dr. Lumpkin, in which the Secretary mentions that there are many urgent national priorities that must be dealt with in the budget process, among them the need to explain overall spending.
So I think that this issue of constraining overall spending is very important with respect to the feasibility or the actionability of any recommendations.
DR. MAYS: Can you see if you can get Jackie Lucas to join us? She was talking about, NCHS had actually looked at some of the feasibility and cost issues. So it might be nice to actually know what ball park we are in in some things. So let's see if we can get her. We'll have Jennifer back later, so we can also raise it with Jennifer.
Thank you, Harvey.
DR. SCHWARTZ: You're welcome.
DR. GREENBERG: Don't tell John that I was suggesting that we --
DR. MAYS: No, we are not making those adjustments, but we actually need to hear from NCHS about cost. I actually raised that in an e-mail. Audrey, there was a person that Jackie had suggested. Did you get that person?
DR. BURWELL: Linda Tompkins. No, I didn't.
DR. MAYS: Linda Tompkins, that is who it is. Apparently that was the person who Jackie Lucas thought might have some sense of cost and feasibility.
DR. BREEN: Vickie, what are you thinking about in cost and feasibility? I think another issue that is kind of implicit in this conversation is, there really are two agencies within DHHS that are collecting health survey data. That is AHRQ and NCHS.
So I think all the big surveys are pretty much housed in those two agencies.
PARTICIPANT: Except for maybe the Medicare beneficiaries.
(Simultaneous discussion.)
DR. BREEN: I don't call that a national survey, but maybe we can.
DR. GREENBERG: It is a state-based survey, but I think all the states do it.
DR. BREEN: They do. They are all involved now, but they don't all ask the same questions all the time.
So I was wondering, do we need cost data from more than just NCHS? Or what would be the purpose of the cost data?
DR. MAYS: Just so that we are in a ball park, to get some sense of, as we make these recommendations. They have actually looked at doing the special targeted surveys, is what I think Jackie Lucas was talking about.
For instance, they are the ones that have done the Hispanic HANES. My bet would be that the costs are probably higher to get in, in terms of the number of call-backs. My bet is that there is extra work in terms of sampling frames. My bet is that there are all these added costs that we haven't thought about.
Like, we kind of have an idea of what it costs to do CHIS, the California Health Interview Survey. That has a large minority component to it, and a lot of that minority component is done outside of even the RDD. So you have a sense of all these extra costs.
DR. GREENBERG: We certainly have estimated the cost of a community HANES. I don't have that, because that is something that NCHS has been trying to do. But it is somewhat different.
DR. MAYS: I think it is to sober us up a little bit and get some sense of what it takes to do this, because again, I think Harvey has probably brought us to a very important point, which you also were echoing, Marjorie. The budget is -- it is an interesting issue. I think everybody sitting around the table, and agencies; we don't want to make recommendations and then agencies find that the only way to get this done is that they are all committed as part of their budgets. So there is a lot to think about here.
It is important that we do this, but I just want everybody to be thoughtful about it. I think for it to be done will have implications.
MR. HITCHCOCK: I think we need to think in terms of options too, not just a way to go about this. Marjorie, what you were saying about either replacing the survey or -- sure, you could for instance run several more HANES sample location stands, they call them, within New York City. The survey would already be there, would be in operation. You could cut down on a lot of costs, and you could look at various subgroups in New York City. It is not a new survey; it is a way of economizing on the infrastructure of the existing survey.
DR. CARTER-POKRAS: If I could just remind, in case people didn't recall this, Ed Sondik had testified to the Deputy Secretary a few years back on the cost of increasing the sample size of the national health interview survey for various subpopulation. So that cost memo may be helpful, in addition to the estimates for the cost for NHANES. It did look at going beyond the subgroups that we currently have.
Ed Sondik had sent it to the Deputy Secretary under the Clinton Administration. You may find -- because of your time frame that you are trying to get this done before September, you may find that you just want a couple of good examples, rather than doing an exhaustive survey.
DR. MAYS: Oh, yes, there is no way that we are going to do costs for everything. But I just want us to have some sense, even if it is the Hispanic HANES or something else, what it is we are trying to get at. I like Dale's point. I think we should have some flexibility.
DR. BREEN: The other thing I was noticing, I was reading these two books, Toward a National Health Care Survey and Guidance for the National Health Care Disparities Report, because they are both about data collection in this country. Actually, the National Health Care Survey one is really old now.
But it has in here the legislation for the National Center for Health Statistics, which was founded I think to collect data in what ended up being the national health interview survey, because we had no data in this country, regular data source, that was up to date on health and use of health services.
The legislation was interesting to read for two reasons. I bring it up because of these two things, in terms of helping us think about how we might structure things. First off, in a lot of cases, they will say, this needs to be done, and the Secretary ought to have this agency do it, or bring a task force. So we give it to the Secretary and make a suggestion about how the Secretary might get it done. In some cases, it is sent off to NCHS. AHRQ didn't exist at the time, but whatever.
Then the other thing was, racial and ethnic populations are not mentioned as part of the mandate for routine data collection in this legislation. Now, maybe it has been updated since then, but the way it is in the legislation here is that special studies should be done from time to time on racial ethnic populations. They call them special studies, because we were wondering about language.
We might want to suggest that this legislation -- find out if it is has been updated, and if it hasn't, to say that racial ethnic different should definitely be part of the mandate for what NCHS and the surveys do, and probably all six OMB racial ethnic categories should be included. That would bring us to the need for these special surveys in order to get at populations that are fairly small relative to the total U.S. population, but important locally.
DR. BURWELL: The legislation has been updated. As a matter of fact, in the Office of Minority Health legislation, we are charged with seeing that NCHS collects racial and ethnic data.
DR. BREEN: So you are the watchdogs for that?
DR. BURWELL: Well, we try.
DR. BREEN: But what enforcement power do you have? Do you have a little police force over there, Audrey?
DR. BURWELL: I wish we did. No, but we are charged with seeing that racial and ethnic data is collected. The Data Council is working with a group on race ethnicity. Olivia knows more about this than I do. They developed an inclusion policy. The Department has adopted that. The OMB Directive 15 revised the standards. By January 1, 2003, all surveys are to collect the minimum standards, so that has been significantly updated.
DR. BREEN: So you think that we are okay with that now? We don't need to make that kind of recommendation?
DR. BURWELL: No.
DR. BREEN: About the enforcement, do you have a friendly relationship with the Data Council and NCHS? Is that how all this stuff gets done?
DR. BURWELL: I'm on the Data Council's work group, yes.
DR. BREEN: I'm just trying to understand how it gets done, and if it is institutionalized well enough to our satisfaction. If it is, that's fine.
DR. MAYS: But Nancy, I think the issue probably is that the Department collects the data, but it is not in significant enough numbers for there to be an analysis, so it is hard to -- they collect it, but then what happens is that for privacy reasons, we can't really use it.
Then I think the other may be, if it is the minimum, which I am assuming is often what is done in population-based surveys, it doesn't get us down to the smaller populations.
DR. BREEN: That are big in some localities.
DR. MAYS: Exactly. I think those are the two issues that in terms of trying to have the Department think about this. They collect it, but we need them probably to do targeted -- or as Dale was saying, another option is, in some places even over sampling, so we know what some of the over sampling issues are.
But the other thing is, there needs to be a plan by the Department to produce usable data for all subpopulations.
DR. GREENBERG: Well, I don't know about all subpopulations.
DR. MAYS: But see, I think that is the issue. That is the issue that we are going to grapple with. Will Guatemanians ever have data? Or will the Tongans ever have data?
DR. BREEN: Even the Mainland Hawaiians.
DR. MAYS: Yes. When they come to the hearings, what they say is they aren't even in the game, because to write a grant for some of the federal RFAs, they don't have basic data.
DR. GREENBERG: If you have a plan, then --
MR. HITCHCOCK: The plan is consistent with the earlier Data Council recommendations.
DR. GREENBERG: I don't know whether you might want to suggest any priorities. That was how I reacted. When you think of all, the plan could be, in 2035, we will collect information on Tongans, because they are such a small group.
DR. MAYS: But it is interesting, because I think this is where we are really getting into the heart of this, is the outline that Dale did. At first I was like, I wonder why he said that? Then I really got it. The question was, where are we going to start, are there some priorities.
DR. GREENBERG: Yes.
DR. MAYS: The priorities can be derived from either the hearings that we have had or the priorities that derive from scientific data that says the disparity is just horrible. But that is going to be hard to do, to suddenly now pick from.
But we can't do everything. So I think we need to find a way to comment about this. Let me take a couple of other comments, and then I'm going to move us along.
DR. GREENBERG: If we are going to try to think outside of the box, another approach could be -- although frankly this was one of the approaches in the minority statistics grant program, but it was never adequately funded, but is to give grants or cooperative agreements to communities or to organizations that represent particular populations, to do the studies themselves, perhaps based on a standardized survey mechanism or something like that, so that we might get a broader coverage then and have more chance of actually involving the community and having the communities themselves identify problems and all that.
So that is another approach to having all the surveys done by one of the agencies. Probably it wouldn't be either/or. But when I think in terms of the really broad scope of groups, as you said, we really need the information, that might be more feasible than thinking in terms of this all being done at the national level. So just adding that to the armamentarium.
DR. MAYS: That is somewhat what I think NIH does. It is almost like, if there were more -- right now, NIH will tell you, they don't want huge studies, because Congress wants them to have more R0-1s. So the notion of launching big studies -- your budget hits a half a million dollars, you've got to get permission now to come in with that.
But there are instances in which I think it has been NIH that has responded. Like for example, right now there is a series of studies that they are doing in mental health, where they are making sure each racial ethnic group -- I won't say each, several racial ethnic groups, for example, they will have data for those groups. They are using the SEDI, so they are going to get diagnostic data. There are other instances of that.
So I guess it doesn't have to be either-or, but we are going against the grain in terms of budgets right now, of what is being recommended. NIH will probably have a fit if we send this forth, too, in the sense that it is against what they are being told to do.
DR. HUERTIN-ROBERTS: It is true, we want more rather than bigger, and the budget is flat. But what can be done is, we can go in with say CDC or a number of agencies, so that the amount coming out of each of the middle agencies' pool would be relatively small, but there would be enough money to cover the cost of the study.
DR. MAYS: Can I just ask, has NCHS ever worked with NIH in doing these studies?
(Simultaneous discussion.)
DR. BREEN: There is a history of collaboration, but it tends to be sort of scattershot. I think Suzanne's suggestion is very much with the grain of the current Administration.
DR. BURWELL: Which suggestion?
DR. BREEN: Her suggestion of agencies working together. I say that because the current Administration wants the government to be more efficient, more lean and mean, as the economists would say. Efficiencies usually mean that you take a couple of things and make them into one, or you have groups working together so that you reduce the bureaucracy required to do it, that sort of thing.
When I was looking for what are our new things about Healthy People 2000, our strategic plan, I Google'd to find it, and I happened upon the Texas state strategic plan for their Department of Health and Human Services. It is pretty amazing; they have been working on it since 1991, and it is amazingly well orchestrated.
DR. MAYS: I did the same thing. I saw Texas and I saw Minnesota, and there was one other that came up that I just looked at.
DR. BREEN: And the values are very consistent with the current federal Administration. They are really trying to move things together and to orchestrate things, to have agencies work together, to have regions work together to save money, to provide better services.
I think the more we can make recommendations that will allow the federal government to work together, which I think is a good thing, I think there is not enough of that, anyway. Not to reduce money, that wouldn't be my purpose, but I really think we need to work together better. We compete right now.
So if we could incorporate that in our recommendations by suggesting that collaboration occur in order to get some particular outcome, I think that would be a good thing.
DR. MAYS: What we are going to do is just list these things so that we don't use them, so that we can stay focused when we get to the afternoon, when we actually start doing this.
Let me ask Racy to put this stuff up on the post-it board. I think first of all what we want to think about is this inter-agency collaboration. That is one that we want to put up there. We want to look at -- I'm sorry, I'm doing this too fast. Collaboration. You can just say inter-agency collab, and we can come back to it.
We want to look at flexible and several methods for achieving the goal.
MR. HITCHCOCK: Approaches.
DR. MAYS: Approaches, thank you.
DR. BREEN: Is that flexible approaches to surveillance, or what?
DR. CARTER-POKRAS: Obtaining data on smaller populations?
DR. MAYS: I think what we are talking about is -- I don't think it is just surveillance data.
DR. CARTER-POKRAS: It is really obtaining data on smaller populations.
MR. HITCHCOCK: And comparable data, too.
DR. MAYS: Yes, we'll make sure that we say -- so if you will also make sure that you look at what Susan Weems sent us about the special population surveys. I think she has some language in there that is very helpful, so we will return to her language.
What else? We need to have some plan about how to do this, so put up there a plan.
MR. HITCHCOCK: A long-range plan.
DR. MAYS: Yes, a long-range plan, because that will help us address this notion of priorities, so that everybody feels that they will get a shot at this.
Anything else that came up that we want to have as reminders up there to ourselves?
DR. CARTER-POKRAS: We had language that had been discussed several months ago. Do you want me to read that recommendation?
MS. LAYMAN: Language?
DR. CARTER-POKRAS: A draft of ovararching recommendations.
DR. MAYS: In the report?
DR. CARTER-POKRAS: In the draft report about this. I don't know whether this would be helpful.
DR. GREENBERG: Was this included at all in that letter that Pam sent the November meeting?
DR. MAYS: They called attention to it, but I think the only thing that is in there is probably the --
DR. GREENBERG: It seems to me that the issue of having a plan --
DR. MAYS: The meat is in there, but not the plan.
DR. CARTER-POKRAS: It says, support the development of studies targeted by the population or geographic area. That is what is in the letter. But in the draft report of several months ago --
DR. MAYS: Wait, stop there. Say it again?
DR. CARTER-POKRAS: Support the development of studies targeted by population or geographic area.
DR. MAYS: Yes, I think we are saying that there is the need to do this, but not a plan about how to do it or what the priorities would be. So I just see that as like, there is a need here. That is what went to Data Council, right?
DR. BURWELL: And the Secretary.
DR. MAYS: But I'm just saying, the Secretary sent it to you at the Data Council.
DR. GREENBERG: He referred it to the Data Council.
DR. MAYS: Right.
DR. BURWELL: And we are working on that.
DR. CARTER-POKRAS: And the draft recommendation of several months ago, implement a multi-approach strategy based on documented needs and the results for methodological research for developing data on racial and ethnic minority populations and subpopulations, including one, improving biostatistics, two, over sampling in national surveys where feasible, and that came out of the subcommittee on American Indians hearing. Three, conducting followup and dual frame type sampling approaches for special surveys and four, sponsoring targeted studies for developing data in subpopulations where over sampling is not feasible or cost effective.
Appropriate resources should be available to accomplish these goals. This was based on testimony and discussion by the committee in the followup meeting.
DR. MAYS: We can put those up as part of what we use.
DR. CARTER-POKRAS: I can actually leave this with Gracie.
DR. MAYS: No, no, I have this. We want to make sure that as we work, we have it there where we can see them. I have two or three copies of the report here, so that is just reminding us.
DR. CARTER-POKRAS: Also, because I am a bit concerned about the time, we did have very specific recommendations in regards to improving health care quality. These are the ones that could possibly address requests for recommendations for health plans. So there are nine of them. Whenever you are ready, I can read them out loud or however you want to share them.
DR. MAYS: Anything else? We are going to move a little bit to talking about the report. Anything else that we want to get up there before we lose what we said, so that when we come back this afternoon to start this process, that we actually have it.
DR. BREEN: When I see interagency collaborations, flexible approaches, several methods and long range plans, I guess I feel the need for short-term priorities or something to be said about the short term.
DR. BURWELL: The report on achieving a vision for the 21st century, those statistics, one of the recommendations is to develop mechanisms to rapidly modify the samples, data collection strategies and data collection instruments of ongoing data collection systems in response to emerging issues. So this is a recommendation that has already been made, that can tie in with it.
There are a couple of others. I'm not going to read them all, but --
DR. MAYS: If we could pick those out, and then what we can do is display them later. I think if we take those, some of the other ones we have, and then look at what we have there, then I think what we want to do is to figure out from those -- because it is just a few too many, but to figure out from those what we actually wanted to put in the letter, I think that would be very helpful.
So I think we are on track with what we want to say. What we are not probably on track with is, once we recommend it, can we do it, and how to do it. That is a little bit what we need to talk about. I think that is what Harvey was getting a little nervous about. See, now you made me nervous.
DR. SCHWARTZ: No, I'm not nervous, I'm very supportive. But let me just add one thing, unless we are going to have a fuller discussion on health plans, then I'll hold that off. Are we going to talk about health plans?
DR. MAYS: Yes, we are going to have a fuller discussion of that, but that I think is going to be a little bit more near the afternoon than the morning.
DR. SCHWARTZ: I'm not going to be able to be participating during the afternoon. I'll have to stop after lunch. Let me make one plea, because I can't see the butcher-block paper. That is, maybe consider public-private partnerships instead of just federal partnerships.
I think a number of folks would say that health plans can make a valuable contribution toward eliminating racial and ethnic disparities in health by collecting and using data. I think they already do, and perhaps there is a role that they can play with respect to the initiative that we are currently discussing.
DR. MAYS: Good point, I agree. There are foundations even that have done some of the --
DR. SCHWARTZ: And that might be a way of addressing the zero sum game.
(Simultaneous discussion.)
DR. GREENBERG: You might want to put under that health plans and foundations, just to trigger.
DR. MAYS: Yes.
MR. LOCALIO: Don't you still want everybody to use the same survey?
DR. BREEN: Yes, and some similar sample frames, too.
MR. LOCALIO: You don't want Commonwealth using one survey on one subpopulation and Kaiser doing another survey in another population, and then not being able to compare the two because they define their questions differently, and we lose all of the benefit of having different entities doing different surveys.
So it seems to me a prerequisite to all of this is an agreement that there is one type of survey or subset of a survey that needs to be done, and it needs to be done the same way each time by everyone, over time.
One of the things that can be done for example is, even if in any one year, BRFFS doesn't have enough people, if you combine five years, you can get some fairly decent numbers for not only particular states, but for subsections of the state. But that is dependent on that survey being the same every year. Otherwise, you're out.
DR. MAYS: One of the things that I was thinking about, and I didn't know if it was appropriate to do, was almost like having some kind of consensus meeting or something, where -- and NIH does this all the time, to bring people together, and it would help create that public-private partnership, and have an agreement over X amount of time that this would be the goal. I think what you would see is, Kaiser would say, here is what we want to do because we are more interested in this. Commonwealth would say they are more interested in that.
So it is almost like in recommending this, there has to be an organizing factor. I don't know whether we can recommend -- again, Marjorie, I don't know how detailed you get with the Secretary, because I'm afraid if you don't, what happens is that it is a very small group that gets together, and it doesn't get together to make an agreement that becomes a consensus. I think that what you are saying is absolutely critical.
MR. LOCALIO: I'm not saying that Kaiser must do one survey and nothing else. But certainly if you are going to go out in the field and collect data, and you have a special need, you could say, everybody is going to conduct a common set of questions. Then once they are there and they have the responses, then you can ask whatever additional questions may be of special interest to that particular foundation or study.
But so long as you have a common set of questions in there, they can be used across various surveys and entities. But you would have to have the same set of questions.
DR. BREEN: To build on that, I think maybe a common set of questions that everybody uses all the time. I think it would be hard to get buy-in into that. Also, I think that there needs to be room for flexibility in evolution and development.
My impression from looking at survey and data collection in the census for a long time -- I haven't studied this as such, but my impression is that the government data sources all build to some extent or another on the census. Inasmuch as they have questions pertaining to the same information that the Census collects, they tend to use questions that are the same or comparable to the Census questions, where you see more variation in the health surveys. And of course, the NHIS and the HANES and others tend to build on the Census by having those Census type socioeconomic and demographic questions, then they build on the additional subject matter. They build questions to get information on that additional subject matter that they are mandated to collect for.
Then what the private companies tend to do are more detailed surveys on some aspect. If we could get them to agree to not only have a division of labor, where they do certain things in depth, but also where they might take a population group and look at what is commonly looked at in that population group or in that area, I think it would make more sense to pick areas where these groups are, rather than trying to over sample them in areas where they are not concentrated.
But to come up with a division of labor, where there is again orchestration rather than consistency and sameness throughout, I think that we need to think about or have a task force think about what would be a division of labor that would get us where we want to go, where we are getting general information on the small populations that we need, which we are not getting now, and in some cases perhaps more depth.
In doing the general surveys on those small populations, we are likely to find that there are areas of concern in those populations that need more depth, in terms of the surveys that we do, but we can't know that until we have the general information to start with.
So more of a division of labor and orchestration than sameness, I think, just as a friendly amendment to what you were saying, Russ.
DR. MAYS: That is something that we probably want to discuss at the level of a report as opposed to at the level of a letter. The letter is where we would make the recommendation. I think in a report, we might -- and we do have a little thing in there about public-private partnership.
I think that what you do want is -- and it is kind of what is done now, the same question is repeated throughout, and then you're right, this is like a states' issue. When the states want it in greater detail, they need to know more about the population. I think we want to go there. But at the public-private partnership level, if we could get an agreement on a base set of questions that everybody repeats, and then the private groups maybe can do something different, I think that would be excellent itself.
Let's turn to the report, because I know Olivia has to do a presentation for the cancer fellows this morning.
DR. CARTER-POKRAS: Actually, the clinical fellows today.
DR. MAYS: I'm sorry. Do I dare try to keep up with this schedule? As busy as she was when she was with the federal government, let's just put it that way.
In the report, we have a fairly good draft. But what I think we need to do is to rethink the structure. One, we have lots of recommendations, and I think what we want to think about is how we want to reorganize these recommendations. Two, in the report I think we want to think about whether or not the categories as we currently have them are really ones that we want to have.
I think the report is too long now.
DR. GREENBERG: Do we have a copy of the report?
DR. MAYS: We have sent them several times.
DR. CARTER-POKRAS: They may not have all of them.
DR. GREENBERG: This is the report that is going to support the letter and all our activities.
DR. CARTER-POKRAS: This is what the whole thing looks like so far.
DR. MAYS: But past recommendations as they apply to population-based surveys.
DR. CARTER-POKRAS: Population-based data, not necessarily surveys.
DR. MAYS: I have to stop saying surveys; population-based data.
DR. GREENBERG: But is that work not currently in this?
MR. HITCHCOCK: Let me try. I think this report was supposed to be based on our first meeting, more or less.
DR. GREENBERG: The meeting two years ago?
MR. HITCHCOCK: Yes, but now it sounds like we are moving more to having one comprehensive report that would reflect not only that meeting, but our hearings since then.
DR. MAYS: Well, I wasn't thinking that we were going to put all the racial and ethnic groups in, no. They get in only as a function of -- if there is material on population-based data, not surveys.
DR. CARTER-POKRAS: Right. This is something that we had discussed at length, because just the American Indian and Alaska Native is a significant number of recommendations. So there was a discussion about having a stand-alone report with American Indian and Alaska Native data.
DR. GREENBERG: We are still doing that, aren't we?
DR. CARTER-POKRAS: So that is one of the decisions, how many reports do you want to have come out of the committee. When I was asked to assist, it was only with that first hearing.
DR. MAYS: What is being added is this notion -- because there is this impending need and a timing factor, about health plans.
DR. CARTER-POKRAS: Right, but there are already nine recommendations that were drafted, based on comments that people had made during their testimony, as well as, I looked also at the transcripts for the followup meeting of the subcommittee that was scheduled I think in February, or it was immediately following the hearings, the first population-based hearings.
So I put those together, and also reviewed the reports and letters and recommendations that had been made by the National Committee, not necessarily just the subcommittee on populations, but also the broader committee, in regards to population-based data. So that put all that together into some draft recommendations for your consideration.
But even with the cuts that Vickie had initially made back in January, even with the cuts, there are something over 60 of them, and the discussion was, do you really want that many. There are about five overarching recommendations, so you can just keep those five overarching recommendations, and keep the other ones as back burners for future letters, et cetera, if you want to do that. If you have been asked specifically to look at the health plans, then you may want to look at those nine recommendations in addition to those overarching recommendations.
DR. MAYS: No, it is not looking at the health plans, it is making a recommendation.
DR. CARTER-POKRAS: Right, about the health plans.
DR. MAYS: Looking at the health plans is very different, so it is not looking at them, but it is taking the approach that if we are talking about gathering population-based data, that this is another area in which the gathering of that data could take place, and making a recommendation about it, so that we are not just focused on the surveys that the federal government does, but that instead we also then add this additional recommendation.
DR. CARTER-POKRAS: But there are nine in there. I think there is language that you could use for your letter. For instance, it has language in there that, we applaud the Department of Health and Human Services for its efforts to improve racial and ethnic data by health plans, including the Medicaid managed care regulation and the state children's health insurance program regulation. We encourage you to go further, and here are recommendations on how you can go further.
Actually, there was a previous subcommittee member who was very active in this area. I sat down and I talked with her, and she gave me some ideas of areas that you may want to consider.
DR. MAYS: That was Kathy Coltin.
DR. CARTER-POKRAS: Right, Kathy Coltin. So it is a synthesis of what was discussed at the hearing, your followup, and plus Kathy Coltin's suggestions.
One of the suggestions for instance is, one of the barriers that people perceive is that the public would not want the health plans to be collecting racial and ethnic data. So one of the recommendations was to do a public education campaign on how this information can be used to improve health. This is actually a recommendation that is consistent with recommendations that were made by the Institute of Medicine in their report on equal treatment.
DR. GREENBERG: This goes back to, I don't know many years ago it was, but when we have the NCVHS testimony after HIPAA, but about collection of race and ethnicity by health plans, and we had the woman from Minneapolis, where they had actually done --
DR. SCHWARTZ: Sheila Letherman?
DR. GREENBERG: No, Sheila was on the committee, but she was pretty well known, but anyway I'm blanking on her name now, but they had done education in the health plan. It wasn't just an education of the enrollees, but of the staff, the people who worked for the health plan, because they were very reluctant to ask people. So they just eyeballed them, which certainly if we are getting into smaller populations is worthless.
So you have to ask people to self identify. An education of the intake staff, et cetera, as to why this is being done, why it is important. So that also goes back to testimony as I said in -- I don't remember when it was, but I can find it.
DR. MAYS: Is that the full committee?
DR. GREENBERG: That was the full committee. We had a panel, we brought in a number of people.
DR. CARTER-POKRAS: I think it is referenced in the introduction piece, so the report that they are xeroxing right now.
DR. MAYS: The date?
DR. CARTER-POKRAS: Yes. It is in there. P.O. and Parsons had done a summary of activities, I think it was in 1994 or something.
DR. GREENBERG: No, this was after HIPAA had passed. This was after we had established the Public Health Data Standards Consortium. We were trying to bring this before the committee, the whole issue of whether race and ethnicity should be collected in the encounter.
DR. CARTER-POKRAS: Right. It was all the discussion about HIPAA. In fact, things have changed. I have gotten an update from the subcommittee, and I forwarded it to you from Susie about the HIPAA. Evidently now there is successful advocacy work that has been done by the Department of Health and Human Services to place it in a different implementation guide. So it is not going to be where people were hoping it was going to be; it is going to be in the public health research implementation guide.
DR. GREENBERG: This is a guide by the Public Health Data Standards Consortium, but because there was so much opposition in the designated standards maintenance organizations to actually collect it as part of HIPAA. But this may be reopened if there is now more receptivity.
But there was also the meeting that I think you probably participated in several years ago, with health plans.
DR. CARTER-POKRAS: Right, we had meetings with health plans, and Aetna this spring announced that they are collecting race and ethnicity data, so things have certainly progressed.
DR. MAYS: It is really also turning the health plans around, is Aetna's decision that they are going to go forth and do this. Other plans have been, as I understand at least from my discussions with people in California, they have been waiting to see what is going to happen. Aetna is moving forth. They want to be ready to go if it looks good, in terms of what had been done.
DR. GREENBERG: They said they were going to collect it at enrollment, or on their encounter data? Do you know where they are going to collect it?
DR. MAYS: I don't remember at what point.
DR. GREENBERG: Do we know what brought Aetna to that decision?
DR. CARTER-POKRAS: These were lengthy discussions that Micky, Lori and I had participated in way back when.
DR. GREENBERG: She was involved in those meetings.
DR. CARTER-POKRAS: Right.
DR. MAYS: One thing I know is that Aetna -- and I can't say about the other plans, but Aetna has had a major community orientation to it. I don't know if that has something to do with why they are the first ones to step out. They have been funding -- they have a foundation where they have been funding some work in this area. They give community grants.
DR. CARTER-POKRAS: It is also balancing negative press. They actually insured SLAITS. All this was coming out at the --
DR. MAYS: Part of where all this money is coming from is why they have this whole community. That is what I was leading to, that part of this whole community foundation part has spun off because of the basis of their original funds.
The other groups have -- if Aetna doesn't take a big hit, there was discussions that they would do this. So far it looks like there hasn't been major flack. So that may be between the Department and this corporate image that they now are ready to do this, but they want to be ready I think in the sense of, somebody says that this is the thing to do.
DR. CARTER-POKRAS: And Aetna is not the first. There are others. Denver Health for instance has been funded by the Office of Minority Health and the Agency for Health Care Research and Quality to test this. David Nerrens had worked for Henry Ford, and he came and testified several years ago that they had been doing this for a portion of their enrollees.
DR. MAYS: Yes, I think that is one other plan that -- I think we even heard some testimony that they have been under development doing this. So there is precedent for it. But the major ones I think still want someone to say, this is something that you should do, and here is why.
There are reasons for it. They also have to answer to their stockholders. Aetna is dealing with the basis of its money, but the others are not necessarily at that same point. But stockholders want to know, is there going to be a liability for example for doing this, is there a business case for doing this.
So right now, I think what they are trying to do is hope that if their recommendations -- what will happen is that you say, this is probably around the corner, to be suggested and mandated, so we are just gearing up to do this.
DR. CARTER-POKRAS: And just to let you know about this public education campaign, the state of Maryland is one of four states that prohibits the collection of reporting of race and ethnicity data by health plans. Only four of the states, and Maryland is one of the four.
Claudia Bechet was involved three years ago, working with some legislators who drafted some mucosal epithelium to see whether they could get that law changed, to allow health plans to collect the information. They got a lot of flack from the minority communities. So that never went anywhere.
DR. MAYS: Those laws were actually done because of protection. People thought that if they had the data by race and ethnicity, that there would be greater discrimination.
DR. CARTER-POKRAS: Same idea as in California, right? Did you read the Washington Post front page? The reason why he is doing that is because he has a thing about Mexican-Americans, okay? That is the reason why he is doing it.
DR. MAYS: No, the basis of why it is being done in California isn't discrimination. It has much more to do with feeling like that it is privacy.
DR. CARTER-POKRAS: That is not the official title. It is not the privacy initiative.
DR. MAYS: No, I'm saying that is what he tried to sell it on, is that it really is about privacy, because that is the same thing in terms of, we make decisions sometimes on race and ethnicity, and that information should be private. But anyway.
DR. CARTER-POKRAS: One of the questions is, do you want one report or do you want more than one report? If you want more than one report, --
DR. MAYS: We are talking about one report here. We are talking about one report here.
DR. CARTER-POKRAS: And what pieces do you want to remain in here?
MR. LOCALIO: What do you mean by here?
DR. MAYS: Here today, what we are talking about is one report in terms of population-based data.
DR. BREEN: But to decide what is in that report, maybe we should -- because I still thought we were doing one on American Indians, Alaska Natives and Asian Pacific Islanders.
DR. MAYS: I hope so, too.
DR. GREENBERG: So those recommendations would not go in here?
DR. CARTER-POKRAS: We could take them out and then save them for the other ones.
DR. BREEN: Or briefly reiterate them, but the more detail would go in the other one.
DR. MAYS: No, we are getting confused here. The other populations have very specific reports. They have nothing to do with population-based data. It has to do with a wide range of data collection issues.
In this report that we are looking at today, it really is about population-based data. The extent to which there are issues about other racial and ethnic groups, they are in there. So reports on other groups are not confined to population-based data. As a matter of fact, they are actually quite different, because those groups don't have much in the way of population-based data.
If you go to American Indians and Alaska Natives, what you are going to find is, that is a problem. Instead, what we focused on is a lot of this service data from the Indian Health Service and a lot of the problems that exist there. We focused on some of the infrastructure issues from the collection of data. We focus on comparability across systems.
So it is a little different in terms of the issues that have come up there, which is why when you are saying take it out, what you are talking about is population-based data.
DR. GREENBERG: So if there were recommendations related to the population-based surveys that were related to American Indians or Native Americans, those would go into the report that Olivia is working on?
DR. MAYS: Yes.
DR. CARTER-POKRAS: That was the question that I had. There were close to ten of those, and if you want to cut your number of recommendations, that would be an easy way of cutting six of them.
DR. MAYS: I think what we need to do is reorganize it. I think that sometimes what we have is a lot of detail in a recommendation. I think that as we have been talking other times, part of what we probably need to do is to look at what the broader umbrella is, and maybe use some of these as examples, and then narrow it down. We are on very specific things on some of the recommendations, and I think they could be lifted up to a broader level.
So I don't think what we are going to do is get rid of them as much as I think we are going to change the format of how we do this.
MR. HITCHCOCK: So the letter we were talking about earlier, that letter to this report is what?
DR. MAYS: The letter that we are writing should be about the broad overarching population-based data. The report would be almost a litany of what a lot of those things would be. But we have eons of them; we won't put them in the letter.
DR. GREENBERG: So the report supports the letter.
DR. MAYS: The report would be like details of the kinds of issues that --
DR. GREENBERG: The letter would reference the report, either that it is also being provided or that it will follow.
MR. HITCHCOCK: There were two main points in the letter. One was the targeted population, and the other one was bringing in health care providers somehow.
DR. MAYS: That is what is on the agenda, is when I asked how narrow or how comprehensive are the recommendations. That is what is on the table. It could be that the complexity of just the small targeted population surveys are such that that is all that should be there, with reference to a report that details more.
Or what you can do is take -- remember, there are five or six overarching; all of those can go in. But it is like, what is the most critical thing, what should the committee be trying to accomplish. I think it is the small area, targeted population, whatever you want to call it, surveys.
So if that in and of itself is complex enough that that is all we should have in the letter, coupled with the health plans, then we should let that be.
DR. GREENBERG: But put the health plans in the same letter.
DR. MAYS: I think we can put the health plans in the same letter, but I just wanted to hear how complex we are going to get. For example, if we have to talk about some of the feasibility and sampling issues here, then maybe we end up with one letter.
DR. GREENBERG: That type of discussion could really be in the report, getting into more detail, or the complexities. It could be referenced in the letter, but it could be dealt with more in the report. We don't want the letter to be more than a few pages long.
DR. MAYS: I think max like three pages of single spaces.
DR. BREEN: I'd like a one-page letter, but that may be too short for these purposes.
DR. GREENBERG: There is something to be said -- although I can see how you are tying these together, there could be something to be said for two separate letters, one on the targeted surveys, and one on the health plans. Each one would then require a response that would really focus on that, whereas otherwise, you put them both in the letter, and one maybe doesn't get quite --
DR. CARTER-POKRAS: Plus it could be triaged to different places, too.
DR. GREENBERG: Right. Probably the targeted survey one is going to go to the Data Council, or it is going to go to the racial and ethnic group. The health plan one maybe should go somewhere else.
DR. MAYS: That is what I think today is about, as far as these decisions. As we go through discussing these issues, I think we see more whether or not it can be done. My hope was that it can, but it sounds like strategically, once those go, they might have a better chance for full attention if they go separately, is what you are saying.
DR. CARTER-POKRAS: Well, you can discuss, and I will touch base with Audrey.
DR. GREENBERG: So you're not coming back?
DR. CARTER-POKRAS: I have to go out to Baltimore.
DR. MAYS: Thank you.
DR. GREENBERG: We need to take one little break before noon. So do you want to do it now, take a five-minute break?
DR. MAYS: Okay, let's take a five-minute break, and then we could come back and do introductions and make sure with the new people that are here -- I think we have name tags for them all, so we can get them to the table.
(Brief recess.)
Agenda Item: Consultation on Small Area/Geographic Area Targeted Population Studies
DR. MAYS: Thank you for agreeing to spend some time with us today. Let me just give you a quick background of what we are trying to accomplish and why we are happy that you could take some time out and be with us. I apologize for running late.
The Subcommittee on Populations and the National Committee on Vital and Health Statistics has been holding a series of hearings. The hearings have been designed to actually look at the issue of the collection of data on race and ethnicity, whether or not post the OMB guidance it is working. At the same time, we have been asking questions about also whether or not we are correcting data that is needed to determine if there are health disparities, thinking about Healthy People 2010.
In the hearings, what keeps emerging is that as we visit with specific racial and ethnic groups, they feel like they don't have data. It is not the issue of, it's not being collected, to some extent, it is collected, but the numbers are too small for them to be able to have data that is analyzed, or in the instance of some of the subpopulations, that data really doesn't for them exist in sufficient numbers for them to be able to respond to some of the Department's disparity applications. So they are desperately asking for data.
One of the things that we thought is probably the most significant recommendation that is emerging is the need for targeted population surveys. As you know, we have asked you even the question of what to call this.
So that is what brings us to ask you to join us today. We are writing a letter that will go to the Secretary, and we are also working on a report. But I think primarily, we want to get advice from you because we are writing this letter. It is one of the recommendations we want to make.
So in our previous discussions of phone calls, et cetera, we have come up with a list of questions which I hope each of you have gotten.
DR. GREENBERG: They probably didn't get them, because I didn't get them until yesterday afternoon.
DR. MAYS: So this is informal. We didn't need Power Point or any of that kind of stuff as much as we would really like to hear your thinking about these issues that we are struggling with. So how would you like to start?
PARTICIPANT: I was going to give a little history of activities and research that --
DR. MAYS: Let me just do one thing. Susan, are you on the phone?
DR. QUEEN: Yes, I am.
DR. MAYS: Jackie? Let me just introduce you. I realize they can't see your name tags, so they don't realize who is here. Susan Queen, Jackie Lucas and Harvey Schwartz is on the telephone. So if when you talk the first time you would introduce yourselves, so that they know who you are.
DR. BURT: This is Vickie Burt. I was going to do like a really quick review of the history, and you can ask the questions that are relevant. Kathryn was going to talk about some of our current activities, a little something on the burner now. I thought I'd go over it quickly as more important questions you want answers to. So do you want me to start?
DR. MAYS: Yes.
DR. BURT: Starting back in -- this is like research and proposals that the HANES program has made in conducting exams and studies outside our current mode of our mobile exam centers, which are really big and expensive, and/or sampling plans for racial and ethnic minorities.
Back in 1975, we did the first of two studies, where we did examinations at fixed sites in Baltimore and Chicago, using evaluation funds to try to look at doing exams outside the exam center. That was not motivated -- at least, if it was, it was never put in writing by any desire to do racial and ethnic groups. It was more a desire to see if it was cheaper, that was one thing. Basically, they had lower response rates, lower data quality, and similar costs for examination at the fixed site.
In 1978, they did another fixed site study in Manhattan, at the urging of the Office of Management and Budget, to try to find cheaper ways to do things.
MR. HITCHCOCK: I was at that site.
DR. BURT: It was 1978, so you can calculate Dale's age. You were 12 when you went there, right?
MR. HITCHCOCK: I was there as a child.
DR. BURT: It was basically a similar thing. The response rates were lower and the data quality was lower, and the exam cost was higher.
Having not been here then and having read the reports, I would say that they did not make an intensive effort to get the exam to the same quality and the same response rate as what we get in the mobile exam center. I think in this day and age we could do similar, but you would have to put the same intensity of effort into it as a big site study to do that.
Then in terms of racial and ethnic minorities, we do have estimates for African-Americans for HANES-1 and HANES-2, but I think we over sampled African-Americans. Maybe Dale knows more than I do.
MR. HITCHCOCK: It was by income back in those days.
DR. BURT: But we did have enough African-Americans to do estimates in HANES-1 and HANES-2. HANES-1 was '71 to '75, HANES-2 was '76 to '80. HANES time moves slowly.
1982 to '84, NCHS and our division did the Hispanic HANES survey, which was a survey where we sample Cubans in South Florida, Puerto Ricans in the New York City vicinity, and surrounding states, and Mexican-Americans in four Southwestern states.
After that, we did HANES-3, which was from '88 to '94, where we over sampled African-Americans and Mexican-Americans. So at the end of the six years of the HANES-3 data collection, 30 percent of the sample persons were Mexican-Americans, 30 percent were African-American, 40 percent were everybody else.
Around the beginning of that survey, there had been some design research to look at other ways to collect data. We didn't do any pilot studies. We talked a lot about home exams and other studies that had done home exams, and we did have a home exam in HANES-3 to try to get some of the older people who couldn't come into the exam center. HANES-3 was the first time we examined people 75 and older, so we hadn't just neglected minority groups, but older people, too.
Then from 1994 to '96, we did a really extensive research into our ability to add other minority groups to the survey, or do special studies in other minority groups. We targeted what at that time were some of the bigger groups other than Mexican-Americans and African-Americans. We looked at Cuban-Americans, Puerto Ricans, Chinese-Americans, Filipino-Americans, Japanese-Americans, Asian Indian-Americans, Korean-Americans, Vietnamese. This was part one of the report. There was a part two which we actually went into more detail. I know we looked at American Indians to some extent at that point.
What we found out was that we really couldn't include these groups in terms of over sampling them in the survey. One of the reasons why we can sample the Mexican-Americans and the African-Americans in HANES is because the populations tend to be somewhat segregated. For example, 51 percent of the African-Americans lived in block groups, where more than 60 percent of the population was black, and 27 percent of the Mexican-American population lived in block groups where the population was Mexican-Americans. So we could pick block groups as part of our sample design, where we knew we would screen in a lot of those two minority groups.
The other minority groups tend not to have very many areas where they have that kind of density.
DR. BREEN: What about Indian reservations?
DR. BURT: We didn't particularly address Indian reservations at that time.
DR. BREEN: Or historical areas? There are Indian historical areas in Oklahoma.
DR. BURT: No, we didn't explicitly look at that. We looked at the portion of the American Indian population, which 50 percent or more were not living on reservations. They tend to somewhat live in the Mexican-Americans a lot, so we would get more than we would by random chance, because we over sampled Mexican-Americans and they just happened to be in the same segments for screening. They would be hard to get to.
Basically, our data collection contractor recommended that when you get less than 70 percent of a minority group or block group, it is so expensive to start screening if you are trying to do a representative sample, saying that this sample is representative of the U.S. So I think from that point on, we tended to focus on what we referred to in that period from '94 to '98 as special studies.
Every person we talked to about the planning of the current HANES survey, we also said we wanted to offer an option to try to have special studies at the same time, where we might look at other minority groups where they would have some of the same content as the survey, but it would be a non-national survey.
One of the reasons we did that was because one of the big criticisms that we heard at NCHS about the Hispanic HANES was that it was not done at the same time as the national survey; therefore there were no national estimates at that same point in time for comparability purposes.
Nothing really came of that effort, although we talked to many groups of people, and many people sounded interested. Then I think around -- this is what to call the groups. We were calling it special studies then, then we started referring to them as defined population HANES from 1998 to 2000, when Lenora Kington was our division director. At all times we were talking about minority groups or other reasons to sample small areas, environmental reasons or geographic reasons; it wasn't just minority groups.
We did have some interest from a group called the Delta Nutrition Initiative Research something, I don't have it quite right. It is a consortium of universities on the Mississippi Delta, and they have some nutrition intervention studies. They were interested in doing some kind of HANES type of examinations related to nutrition. We met and worked with them a lot, but no funding was ever forthcoming for anything to come from that. So we had been quite hopeful that something would come of that, and nothing did.
In recent years, from 2000 on, we renamed these smaller versions of HANES as the community HANES. I don't know if that was our intention, but certainly in the last couple of years we seem to have more of an emphasis on geographic areas, communities, than minority groups.
So that is where we are now. We have had talks with various groups in Washington, D.C. and California and New York City and a town near Washington called Brentwood. But I think we are now going to have a New York City HANES, which Kathryn is going to talk about.
DR. MAYS: Let me just ask you one question. When Reynard was there, he had some ideas about these surveys. Were there discussions about potentially doing them?
DR. BURT: Oh, yes. We have been talking about it really actively for almost ten years now, under these various names. Our division is not in the position to initiate the survey on our own. We are doing the national survey now. We were looking for partners and interested parties to have some of those come about.
DR. MAYS: So from your perspective, in order to make it happen, what you need are -- you think that you can do sampling, you think that you could actually do it?
DR. BURT: Oh, yes. It would not be inexpensive.
DR. MAYS: Can you do it in all groups? It is a different challenge. The African-American group is a different challenge, I think the Latino population, particularly if you talk about doing Mexican-Americans. But in terms of for example the Asian, Native Hawaiian or other Pacific Islanders, do you think that you would be able --
DR. BURT: It depends.
DR. MAYS: See, this is where it gets difficult.
DR. BURT: Do you want to say, here we have done a study where we have some of the same or all of the same content as the national HANES, and it is representative of say the Chinese-American population. That would be so expensive that it is prohibitive.
If you wanted to say, we want to go some places where there is a high number of Chinese-Americans and have a special study of them, and always present the data in terms of, it is not nationally representative, that is I think possible.
Here is an extreme example that I can highlight if you want to make it representative. If we wanted to have 200 examinations with three to five year old Chinese-Americans, which that was one of our domains in HANES-3, three to five year olds, with a response rate of 85 percent, we would have to screen 113,000 households to find those 200 children.
MR. HITCHCOCK: That is nationally though, right?
DR. MAYS: Nationally, right? Geographically, you come to San Francisco.
DR. BURT: That would include all block groups where Chinese-American account for one percent or more of the population. There are lots of block groups where they are zero percent of the population.
I don't have it with me, but we looked at a lot of the groups, and what counties they were most dense in. Some of you who live in the Washington area, Montgomery County is one of the top density counties for Asian-Indians. I don't think if you lived there that you would think that that was a large proportion of the population, if that is one of the highest density populations in the U.S.
DR. MAYS: In the U.S.?
MR. HITCHCOCK: It ranks about number ten. We just had a presentation on that, so it is not a real biggie.
DR. BURT: Maybe it has come down. It was about three back in '90. But back in 1990, I wouldn't have said, gee, look at all the -- I know a lot of Asian-Americans, but I wouldn't have said that was something striking about Montgomery County. So it is really, the screening costs are just prohibitive. So you have to do hundreds and hundreds of households to find a couple thousand households where Chinese-Americans live, especially if you want specific age groups, which is what people want.
MR. HITCHCOCK: So if it was going to be community-based, for instance, if you were going to do a survey that would look at Koreans in Los Angeles, for instance. Have you done similar work that would address the --
DR. BURT: I think that was the group we looked at. Maybe not. Yes, we did look at Koreans. I think we thought that we would have to think of maybe alternative mechanisms for sampling those groups, that we may not want to have a national contractor, but maybe a subcontractor that was local, who knew more about the local population, that we may have to have a different mode for who the contractor was. At least our current data collection contractor has done other studies in Chinese-Americans, and they have a lot more trouble getting them to participate than some other groups. So I think that when we -- and we find in the national HANES, we go into various cities and communities and by random chance we may pick a neighborhood that has a particular ethnic group, and then maybe a large proportion of people don't speak English in the group, and that raises a lot of special problems for us. There may be groups we sample who for religious reasons don't want to participate in the survey and we have to get the religious community involved, to try to get them to come to the examination.
So I think the more specialized the study, the more anybody would have to think about getting people involved, scientists and researchers involved who were already involved with that community.
DR. BREEN: But in the case of populations where they are very concentrated, like Asians on the West Coast or New York City or something like that, you have considered the possibility and might be willing to move forward with a survey that would be of those areas, rather than national in scope?
DR. BURT: It certainly is something that our division has talked about a lot over the last ten years. But I would say in the last few years it has been more geographic than minority groups that we have talked about. I don't know why. I think we had a lot more interest from local areas who wanted to pursue the issue.
DR. MAYS: I think what we will move on. You were going to present -- you just have to make sure you introduce yourself.
DR. PORTER: Hi, Jackie and Susan. I work in the National Center for Health Statistics in the Division of Health and Nutrition Examination Surveys; we have a new name. I am in the operations branch, and had been working with another partner in my division on the community HANES effort.
I think if I can help you understand, we have this very solid protocol for the national survey, where we do an interview, and then we do exams and multiple different components. So we have all these established protocols that we feel really delivery high quality data.
The concept is to do that on a smaller scale, in a community, be it for a defined population or just a geographic area, where we would actually almost like a Chinese menu, say I want some income information, some education information, some race ethnic information. Then I want body measures, I want to draw blood, I want to do height and weight, blood pressure, and maybe do a mental health interview, and package it into a smaller survey, have actual smaller examination centers.
These are indeed already built. We have community HANES trailers, if you will. Our national survey exam is a center consisting of four trailers; this community concept would be a single trailer and a self-propelled vehicle, so we could drive to harder to get areas.
So the concept would be to run the survey for three to six months, and maybe interview and examine 2,000 people. This is just throwing numbers out. We certainly have to do some statistical tests to determine how many people we would need to examine for statistical power. But then, with our established protocols and our automated system, be able to then produce results soon. It could be out certainly in six months.
So that is the concept from beginning to end. We are thinking it would be something like 18 months to plan it, to field it, to produce the data and disseminate it.
Now, we have had a lot of inquiries from communities on this. Vickie mentioned a few. I'm just going to reiterate. The District of Columbia was interested, Seat Pleasant, Maryland was interested, Washington State. Vickie alluded to the Mississippi Delta Nutrition Intervention Study. We have talked to people in Marin County.
At the federal level, we have had discussions with ATSDR, as well as NIH's National Center on Minority Health and Health Disparities, and they are interested in doing an Appalachian HANES to look at health disparities.
But our strongest partner to date if you will has been New York City. The health commissioner there, Tom Frieden, intends on doing a New York health examination survey. They want to do the survey. As it is turning out, they are using our technical expertise to do the survey. They are giving us some money to develop the sample design for them for all five boroughs, as well as the information technology, so they can use the same HANES protocols and the same information technology, so they can have data right at the end of their study and they can produce it and disseminate it.
They plan on launching their survey in April or May of 2005. They will be interviewing and examining 2,000 residents in all five boroughs of New York City. They are going to do a limited interview and a limited examination, and that is going to be their community health and nutrition examination survey. So that should be a reality, and we are working closely with them on it.
So yes, it can be done. These surveys cost anywhere between three and five million dollars geographically. If you were starting to screen for subpopulations like we have mentioned, the Asian population in Southern California, it is probably going to be more, just because of the screening effort involved. Screening effort means more interviewers, and more interviewers means more labor, and the labor category drives the cost of the survey.
MR. HITCHCOCK: How many people do you get for three to five million dollars?
DR. PORTER: We have been costing this out with examining 2000.
DR. BREEN: Wouldn't the costs go down if you added to the sample? Aren't they fixed costs, or are they completely interview cost driven?
DR. PORTER: Labor is a lot, but we are really hoping now that we have been working with New York, we will have fully adaptable modules for the interview and for the exam, but that sort of information technology work has been done. Also, some of the trailers will be outfitted then.
Let me just also mention, for New York they are going to do fixed sites. They are not going to use any mobile exam centers. They have opted to use their clinics, because they have clinics in suitable places all around New York.
But nevertheless, as we get sponsors or funders to do this, we will be able to outfit the trailers with all the freezers and the laboratory equipment that we need, and the state-eometers, et cetera. Once those are in the trailers, then we could just do community HANES non-stop for whatever communities need data.
MR. HITCHCOCK: You haven't yet, but would you ever combine the samples from regular HANES with these augmented --
DR. PORTER: That sounds like a great idea, but that really gets tricky with human subjects side. We in our consent forms now promise our national survey respondents, if they are part of this national survey and we promise that their data will be produced for national estimates, we would have to rethink the consents so that they would understand that their data would be used at the community level as well as the national level.
MR. HITCHCOCK: I would think that would be an incentive.
DR. PORTER: Yes. It is something we need to think well in advance. Plus, we have to assure that the community then is going to use the exact same laboratories as the national study. If you are using the segments that we have in one area and we have all the lipid analysis done by Johns Hopkins, then you would want the community HANES that is going on at the same time to send their lipids to Johns Hopkins. It would just be the coordination effort.
DR. MAYS: What is New York going to do? They have a lot of their own labs.
DR. PORTER: They are doing their own labs.
DR. MAYS: So for example, if someone sends you a paper to review or something like that, would you feel that because the labs aren't the same, that there is a problem?
DR. PORTER: No, because we actually are encouraging them to use our same laboratory protocols. That is, if they do the lipid analysis, they will be doing it the same way with the same machines. So they are going to have all that.
MR. HITCHCOCK: The same quality control specimens and all that sort of stuff, too?
DR. PORTER: Yes.
DR. MAYS: Then that works.
MR. LOCALIO: Just more generally, what is the comparability of the information coming from these surveys? Can I compare the Mexican-Americans from Southern California with the Puerto Ricans in New York City?
DR. PORTER: You're saying that if we do two community surveys?
MR. LOCALIO: Yes.
DR. PORTER: If the protocols are the same, then that would be our intention. You would be able to compare the data from the California group and the New York City group, as well as compare them with the national survey data.
DR. BREEN: Which we think is the most important part, is to compare to the national data.
DR. PORTER: Right, because we would have the examiners do the tests in the exact same way, the height and weight the exact same way, the blood pressure the exact same say, the hearing, the audio, whatever you would want to -- whatever assessments you would want.
MR. LOCALIO: One more clarification. You talked about sponsors. Who are your sponsors now?
DR. PORTER: Right now, we have none. Right now, we don't have any.
DR. BREEN: For the national HANES or the community HANES?
DR. PORTER: Oh, I'm sorry. That is what we have been looking for for ten years.
DR. BREEN: Well, NCI has tried to sponsor the NHANES unsuccessfully, which is why you don't have enough money.
DR. PORTER: The community gets funding from dozens of collaborators.
DR. BREEN: So it is the federal government that hasn't come up with it?
DR. PORTER: Yes, that's right. We said that we have very limited resources. We were actually going to partner with them to do this, but when we heard how much it was going to cost, they felt that they would do it themselves. But they really were committed to using all of the NHANES protocols and NHANES technology. So they are giving us a million dollars to do the sample design work for them, as well as adapt the NHANES protocol and build their information technology with them.
MR. LOCALIO: If a private foundation came to you, you would cooperate with the private foundation, the health plan?
DR. PORTER: Yes. We've got the expertise, we've got the trailers, we've got the protocols.
DR. BREEN: How many trailers do you have now?
DR. BURT: Forty.
DR. PORTER: That's right, 40.
DR. BREEN: So you could provide technical assistance, infrastructure.
DR. MAYS: Let me ask Marjorie, and then we're going to move on so we can make sure we get the other two. Marjorie?
DR. GREENBERG: I just wonder if the New York study, whether they are interested in looking at racial and ethnic groups beyond just how they will show up in the --
DR. PORTER: No, they want broad-brush estimates for the whole city, just because of the nature of what they are looking at. For example, diabetes, they have five percent of their population diabetic, and they only have 2,000 respondents. Once you start cutting by race and ethnic groups, they are not going to be able to say much, it gets to be such small cells.
DR. MAYS: Okay, can you -- because I want to make sure we get --
DR. DAVIS: I'm Bill Davis with the National Cancer Institute. I support Nancy Breen in a lot of endeavors. We are in the same division. She can probably talk more knowledgeably about -- I'm not exactly sure what I am supposed to be talking about.
I am in the surveillance research program, which includes the SEER program, and Brenda Edwards is on the committee, and she has been involved with the SEER program for a long time. You probably all know, it is population-based in a subset of the U.S. and collects cancer data there.
DR. MAYS: Can you talk about how you are able to do that? A little bit about how it began? Apparently for some of the racial and ethnic groups, it has been very successful. It always has data on cancer. They have data on a lot of other things, but they have data on cancer. So it is like one of those possible models that we would like to consider.
DR. DAVIS: I'm not too knowledgeable about that. It's too bad Brenda isn't here, because she has been involved for 25 years. I am more a sampling statistician that works across the division. There is a little branch that does the SEER collection.
DR. BREEN: One thing you might want to talk about though is, when areas are chosen for SEER, or when new areas are brought in to SEER, it is done with an attempt to have adequate numbers of each of the racial and ethnic groups in the country to get estimates from.
DR. DAVIS: And now all states are collecting their own. They formed basically through CDC -- and if Reynard was here, I think he could talk more knowledgeably on it than I could, but all the states are forming their own population-based registries and are cooperating, along with Canada and others.
The SEER was there first. It had mostly from university settings, like Connecticut. New Mexico was there, Utah and then they kept building on that. But Connecticut was the prototype which did it for the whole -- so they got all the hospitals in the state to send data to them, and got people to QC the data.
But my impression is, what I have seen published is, you said the statistics are good from SEER on the race and ethnicity, but all I've seen is estimates for blacks and whites, and not broken down by smaller ethnicities than that, maybe because they think the estimates aren't quite as accurate.
DR. MAYS: My understanding is that the American Indians and Alaskan Natives -- because when we were at a hearing, that is the data that they were talking about.
DR. DAVIS: The SEER data, they used it, and it was good enough to be -- maybe that is true.
DR. MAYS: I think given the paucity of what they have, they took it.
DR. DAVIS: The other thing that I am familiar with as a statistician, and maybe you want to do something new, but there is data already out there. It is hard to combine. I'm familiar with this Westat report, ASPE from '99, an assessment of major federal data sets for analysis of Hispanic and Asian or Pacific Islander subgroups, and Native Americans. Joseph Wachsburg and others from 1999. It looked at all the different surveys. Westat does a lot of these surveys including the Census, and figured out how many people you would have of certain types from samples from HANES, from HIS, from all these things.
Then if you want to get specified accuracy just for prevalences, then what happens if you combine over years, if you are willing to assume things are stationary over time, and so on. So a lot of this work has been done before. I think Vickie was basically referring to similar things for the NHANES.
But if questions are asked, obviously there are confidentiality problems when you get into race ethnicity things, and Title 13 and all that stuff, and it is really complicated.
We do little surveys, and Nancy could talk about CHIS. I don't know if she has talked about CHIS, the California Health Interview Survey.
DR. BREEN: We have actually had people come and talk about CHIS.
DR. DAVIS: Okay, so I won't say anything more about that. But --
DR. BREEN: You know what, Bill, I think might be helpful for you to talk about?
DR. DAVIS: Yes?
DR. BREEN: Related to the report you referred to that Westat did, aren't you working on overseeing a project now that is being done to use the BERFIS and the national health interview survey to improve state based and national estimates?
DR. DAVIS: And even smaller --
DR. BREEN: Could you talk about that a little bit? Because that might be a strategy that we could use.
DR. DAVIS: It is still in the research phase. We are working with Matt Shenker of the National Center for Health Statistics, and we have two professors that we are working with, mostly at the University of Michigan, Professor Reganoff has done a lot of the work. And we are working with the people at BERFIS. BERFIS has problems with -- all telephone surveys are having problems with response rates, whereas in-person surveys such as the national health interview survey is done more in person; they do have better response rates. It has a problem with small sample sizes, so we are trying to look at using analytical techniques to make estimates of small area levels using the data, combining the data for both.
Another problem with the telephone surveys is four percent or something of the population doesn't have telephones, and obviously in Indian reservations it can go up to almost 50 percent. Apache County is very high. If there are differences between people who have phones and not phones like in smoking, then your telephone estimate is going to be quite biased. We feel we can try to correct that using more in-person non-telephone surveys.
So we are working on that. I think it is still in the research phase. We are doing the computations at the National Center for Health Statistics, because we are not using the public use -- as you well know, the public use data doesn't release small area identifiers when they release it, and that is a huge problem for all of us. So we are trying to make estimates, and we are using the information that we are doing the stuff on outside of the National Center for Health Statistics. So there are research data centers posted at the Census Bureau and National Center for Health Statistics, and you can go there and make estimates, but it is pretty difficult for a lot of the agencies to do that.
DR. CYNAMON: Hi, I'm Marcie Cynamon from the National Center for Health Statistics. I am the acting branch chief of the newly created Special Populations Survey Branch, so that is my vote for what to call it.
We call it that mostly because we didn't want to be too narrow. We wanted to be able to include special populations that were special because of age or condition or an event such as having an immunization, or income, race, ethnicity, something that would encompass all of those different special populations.
There is so much to tell you about SLAITS. Having listened to your conversation in the earlier section of the meeting, it is very exciting for me, because I think it can address a lot of the issues that were discussed.
A few sentences of background. SLAITS is the newest data mechanism at NCHS. It exists in its current incarnation as a spinoff of the ongoing national immunization survey, which I also direct. The national immunization survey targets a very small population, 19 to 35 month old children. In order to identify sufficient numbers of these children, a huge sample is selected. Over one million households are screened annually to identify state level and local level populations.
The current NIS includes all states, D.C. and 27 other metropolitan areas. SLAITS uses --
DR. BREEN: Excuse me, Marcie. So it is not a national survey.
DR. CYNAMON: It is a national survey and it is a state based survey, and it is a local survey.
DR. MAYS: I got confused, too. Wait a minute, all states, D.C. and 27 metropolitan areas.
DR. CYNAMON: Have their own samples. They can be combined to make national estimates, or you can look at state estimates, or you can also augment the sample to look at other areas that aren't emphasized.
MR. HITCHCOCK: Public use tapes?
DR. CYNAMON: Public use tapes.
MR. HITCHCOCK: How far down can you go with public use tapes?
DR. CYNAMON: Are we talking about SLAITS?
MR. HITCHCOCK: I'll say SLAITS.
DR. CYNAMON: Let me back up, and then I'll address that. SLAITS is completely funded by outside sources, meaning federal sponsors, we have had foundations, the Gerber Foundation through the American Academy of Pediatrics supported one of our surveys.
The mechanism constantly exists, but the surveys don't constantly exist. They only exist when somebody funds them. Our biggest funder today has been HRSA, and the Maternal and Child Health Bureau has funded two, and they are in the process of starting to fund a third national survey, where you can get estimates for all states.
One of the surveys was about children with special health care needs and the other survey is a general children's health survey, and they alternate every two years.
DR. MAYS: Can I ask a question? The mechanisms exist, but it is like the surveys change, right?
DR. CYNAMON: The surveys change. There is not an ongoing thing like HIS.
DR. MAYS: But do you have a bank of questions that you always ask?
DR. CYNAMON: No, it is completely targeted to whoever is paying for it once.
DR. BREEN: Marcie, if you could just clarify again, when you say the mechanism exists, the sample frame exists and you can build anything on it that you want, right?
DR. CYNAMON: Right.
DR. BREEN: Age or gender or population specific ethnic groups.
DR. CYNAMON: Yes. This tells you a little bit about some of them that we have done, the larger ones. We do a lot of methodological research using Department evaluation funding. One we were involved with Aetna for awhile, looking at the possibility -- the strengths and weaknesses of telephone surveys on the American Indian and Alaskan Native population. Not great.
MR. HITCHCOCK: Does it exist independently of the national immunization survey?
DR. CYNAMON: Not now it doesn't, because we rely on their sampling frame. Let's say for a 50-state survey, it saves about two million dollars of survey costs to piggyback using the NIS sampling frame. The national immunization program pays for all of those costs.
MR. HITCHCOCK: That is ongoing into the future?
DR. CYNAMON: That is ongoing into the future until immunization registries can really be depended on for detailed data.
DR. MAYS: Where does NIS get its sample from? Does it get it from Census?
DR. CYNAMON: No, it is a random digit dial survey. It is completely independent of the Census. What isn't independent is something I haven't heard mentioned, and this is jumping around a little, is that when you do focus on small areas or small populations, you have a serious challenge with finding population control totals to do your weighting. That should be always in the back of your minds when you are making recommendations about the statistical soundness of what you are recommending.
i'm not a sampling statistician. Bill can probably talk a lot more about these concerns than I can.
DR. BREEN: He kind of left us with an ellipse. So does SLAITS have that associated with it, or it doesn't, the population control figures?
DR. CYNAMON: This is interesting, this is a tangent. When we did the children's with special health care needs survey, we were ready to produce a public use field right before the state level 2000 census was released.
So we used the 1990 projections, and as everyone knows the Hispanic population was grossly underestimated. So we had to recall our public use file and reweight the whole thing, based on the new figures. So there are problems associated with -- depending on those population control totals.
DR. BREEN: I would say the HANES experience over sampling Mexican-Americans, which is a huge group that is not as straightforward as you think it might be, just to get those totals.
DR. CYNAMON: And another issue that I also wanted to bring up, when you ar focusing on race and ethnicity or income, especially those two kinds of variables. When you over sample, that mensa at the earliest part of the interview, you are going to be asking those very personal questions. They can be considered an effrontery, and it reduces your response rates.
The point I would like to make here is that it is good to use over sampling as part of a broader survey, rather than just target populations. A lot of populations don't like being targeted. We don't have the luxury that HANES has, of being in a community and doing a lot of promotion, and having the in-person outreach. That makes a big difference. People understand what your targeting is about. When you are calling on the telephone, it is not as easy.
One of the other things that we did that I think was very interesting in our big children with special health care needs survey is that we conducted interviews in 11 languages in addition to English and Spanish, which was a first time for us. It was a very interesting and convoluted process. It is difficult to do, and the interviews cost I think five times more than an English or Spanish interview would cost.
DR. BREEN: Were those translation costs, or what were the costs attributable to?
DR. CYNAMON: It was a combination of translation and hiring and training and keeping on board experienced interviewers who were proficient in that language.
MR. HITCHCOCK: In which languages?
DR. CYNAMON: I don't remember all 11, but it was Tagalog and Korean and Mandarin and a couple of other Chinese dialects, several. It was the top 11. Portuguese, Italian.
DR. BREEN: So we know what the top 11 languages are.
MR. HITCHCOCK: In the U.S.?
DR. MAYS: Can I ask a question?
DR. CYNAMON: Yes.
DR. MAYS: Do you have those translated interviews available? Are they on a website? Can other people access them?
DR. CYNAMON: I have them. I can provide them.
MR. HITCHCOCK: What do they look like?
DR. CYNAMON: Well, let's see. We gave them to the IRB, so they are in readable form.
MR. LOCALIO: Are those surveys in the public domain?
DR. CYNAMON: Yes. Anything we do is in the public domain. But addressing the challenge of the small area issues related to confidentiality, my goal is to have something akin to a mini research data center within my branch, where we would have one or two people who can provide data runs at the request of researchers, given that a lot of the focus of these special populations is pushing the limits on confidentiality.
I have kind of gone all over the place, because I have so many points that I wanted to make.
The other issue that I wanted to discuss was the phone/non-telephone adjustments. We funded a lot of research in this area, and we developed a new technique for doing non-telephone adjustments.
Originally, we were looking at health interview survey distributions and comparing the phone/non-phone population from that. But now we know from additional research that people who have interruptions in telephone service of a certain period or longer are more like people who never have phone service than people who always have phone service. We can ask those questions on our telephone interviews and make adjustments based on that.
So we are a little more secure about being able to represent that population. However, I would caution against using a telephone survey for populations that have very low telephone coverage.
DR. BREEN: Which are?
DR. CYNAMON: Like the American Indians and Alaskan Natives. That we know for certain based on research that we have done.
DR. MAYS: American Indians and Alaskan Natives that are on reservations or in the general population?
DR. CYNAMON: Mostly on reservations. Rarely do people only want to study American Indians and Alaskan Natives who only live off of reservations. Usually it is reservations -- I am forgetting my terminology, near reservations, and then in the general communities. Those three are usually -- people want to compare those three.
DR. HUERTIN-ROBERTS: I think in Alaska, I think the issue is just the nature of the state, and the people living in small villages that are very isolated, very difficult to travel through Alaska. So I think just for all Alaska Natives, it would be problematic, as a population.
DR. CYNAMON: And very migratory there, too.
DR. HUERTIN-ROBERTS: Yes.
DR. COOPER: Just a very minor question, but given the changes you see in use of telephones today, with many people switching from having a telephone in their house to using a cell phone, --
(Simultaneous discussion.)
DR. CYNAMON: At this point, we only select exchanges that are landlines. I commissioned some questions on the health interview survey about abandonment of landlines and switching to sole use of cell phones.
Right now, it is still right around two percent of the population that has done that. I just started analyzing that data, so we can make adjustments to that too, just like we do with the non-telephone.
Eventually, telephony I think will change, and we will be able to do some sort of interview with the cell phone populations as well. Right now, it is hard to reimburse people. If for example households have their landline advanced to their cell, we ask if we can make an appointment to call them back when they are home, so we aren't incurring the cost.
DR. COOPER: A lot of low income individuals tend to use these cell phones. I don't even know if they have permanent telephone numbers, or if they are just going by a certain number of minutes.
DR. HUERTIN-ROBERTS: For certain landline plans also, isn't there a certain plan where you have a limited amount of use, and anything over, you pay more? So you are running into that problem even with some persons with landlines.
DR. CYNAMON: I haven't encountered that yet.
DR. BREEN: You mean message units?
DR. HUERTIN-ROBERTS: You pay ten dollars and something for --
DR. CYNAMON: For incoming calls?
MR. HUTCHINSON: It shows up on the Caller ID when somebody has called to participate in the SLAITS.
DR. CYNAMON: I know. We have tried to get them to change, but the data collection contractor does lots of other surveys. The way the telephone center is set up, it is not possible to do that. But we send advance letters to 80 percent of the households, so that helps.
DR. COOPER: And the letters look like junk mail letters?
DR. CYNAMON: They look like Kansas letters and everybody else's letters.
DR. BURT: Ours are beautiful. We raised the standard. It took years to get them to not look like junk mail. It took years of effort.
DR. COOPER: Isn't it in October of this year, there is this number you can call in, so any call that comes into your house without a name, it automatically will block out --
DR. BREEN: No, that is the telemarketing registry. Research is exempt.
DR. CYNAMON: A lot of people ask us how we are different from BRFSS. BRFSS is an ongoing survey. It is administered by the state, so there is variability about how it is conducted. It is targeted to adults, and it tends to cover a lot of topics briefly.
SLAITS is different, in that we design it to meet the sponsor's needs. So if it is children, we do children. Mostly we have been doing children, it has just worked out that way. It can go into great depth on one topic, or cover many topics. And it can be repeated periodically; it doesn't have to be every year. It hasn't been every year.
It can be in one calendar quarter, it can take a whole year. The first time we did children with special health care needs, it took us a year and a half to amass sufficient numbers at the state level.
So there is a lot of variability. It is totally designed to meet whatever the needs are. One thing I would like in the long run is to have some appropriated money so that we can create our own survey and not be based on the national immunization survey, and be more responsive to more sponsors simultaneously.
Right now, we can do two or three surveys at a time. We have in the field currently the general children's health survey and an asthma survey in four states, and a national sample, and New York has paid to add New York to that.
MR. HITCHCOCK: Can you talk a little bit about the amount of time you actually interview a person in the SLAITS? There is a lot of upfront screening that goes on, at least in my experience, with SLAITS, even before you get to the content part of it.
DR. CYNAMON: The upfront screening for the general children's health survey for example is, do you have any children under 18 living in this household, how many, and then they take one. So that takes about a minute. But if you have children with special health care needs, for example, that took five minutes. If you want prevalence estimates, then you need to collect some information about households that don't screen in.
So it depends on what the goal of the survey is.
MR. HITCHCOCK: Probably you don't want to exceed 30 minutes or 15 minutes?
DR. CYNAMON: Right, 30 is about right, max.
DR. MAYS: Can we open this up also to broader discussion, so that if there are questions before we run out of time, for anyone?
DR. BREEN: This is mainly for Marcie, but other people might have ideas, too, suggestions.
You said that you had worked with Aetna in trying to figure out a way to use the telephone survey mechanism to get at small populations in the national population that have a small proportion where everybody is not covered with telephones, and that didn't work out.
Is there anything that it would be helpful for us to know?
DR. CYNAMON: Yes.
DR. BREEN: We are kind of desperate here, and we --
DR. CYNAMON: Dual frame.
DR. BREEN: Dual frame?
DR. CYNAMON: Dual frame. You would have to design something that is partly in person and partly RDD or targeted RDD.
DR. BREEN: And how would you end up with a single sample frame for your denominators, your population control totals?
DR. CYNAMON: You would have to rely on Census estimates or projections from some other source, but really, Census estimates are the best, the decennial estimates. The further away you get from the decennial, the more unstable they become.
DR. BREEN: If you had a dual frame, can you give a little sense of how you would do that? You'd use SLAITS for the telephone.
DR. CYNAMON: And HIS for the in-person, or many of these contractors that are capable of doing large-scale telephone surveys also have the capability of doing in-person interviews.
DR. BREEN: And can you combine them and get standard errors and confidence --
DR. CYNAMON: Yes. You would work very carefully and script it out very well. We have some great sampling statisticians that we work with who love that kind of thing.
DR. BREEN: So did you do it for Aetna?
DR. CYNAMON: Aetna doesn't have any money. Aetna has al to of interest and no money.
MR. HITCHCOCK: The pilot was ASPE money, right?
DR. CYNAMON: Yes, the pilot was ASPE. Actually, it was NIMIE's data when we were first starting up. We said, sure, we can do a survey for that amount, and we really couldn't.
DR. MAYS: But what you are suggesting is for national?
DR. CYNAMON: It could be targeted for geographic area. It could be a particular characteristic. It really depends.
DR. MAYS: If you were to do this in California, you would have to do a dual frame, necessarily. You have the largest urban Indian population, for example, in the U.S., is in L.A. County.
DR. CYNAMON: It is the phone coverage issue, not the distance issue. You don't need a dual frame. It is only when you are talking about really high levels of non-telephone coverage.
DR. GREENBERG: I realize that currently these are customized to the sponsors, but is one of the goals to have a bank of standardized questions?
DR. CYNAMON: Originally SLAITS was developed to be the state level health interview survey, but nobody bought that. There was a lot of interest, but there was no check writing. The money came from these agencies that wanted very highly specialized surveys.
DR. GREENBERG: They don't want to compare it to national data?
DR. CYNAMON: Not as much as they want what they want.
DR. GREENBERG: But if there was interest in -- just like the idea of community HANES, but of doing surveys on a number of different populations, but with comparable data, obviously that is something that could be done.
DR. MAYS: If we move beyond CDC, like you were talking about all the different people that visited you, who visits you saying they want -- we know Aetna does, but who else is saying they want this? Like, do you have customers, do you have interest at HRSA?
DR. CYNAMON: I'm talking to National Center for Disease Prevention and Health Promotion.
DR. MAYS: That is CDC?
DR. CYNAMON: Yes, the office of diabetes, juvenile diabetes and childhood obesity. I'm talking to them next week, and someone at NCI who is interested in parents who have survived cancer who have young children. Who else just called me? Urban Institute just called me about people between 100 and 200 percent of poverty to do a health insurance affordability survey.
DR. MAYS: They just got ten million dollars.
DR. CYNAMON: Oh, thanks, that is helpful to know. Now we know what to put in the proposal. Who else has called me lately? We are repeating the children with special health care needs in a year and a half. Lots of people call. They don't always shave enough money.
DR. BURWELL: I'd be surprised if racial and ethnic groups would do that because of the methodological issues beyond telephone coverage.
DR. CYNAMON: For the American Academy of Pediatrics study, we did screen for Hispanic ethnicity and African-American, to all of our first 2,000 interviews. So there was some level of screening and some level of over sampling to strengthen the estimates.
DR. MAYS: Are there any questions?
MR. HITCHCOCK: Yes, I have a question for Vickie. Just to put things in context here, you were saying as I wrote it down here that to do a community HANES study of around 2,000, it could be anywhere from three to five million?
DR. BURT: Yes, I think that is what we would consider to be --
MR. HITCHCOCK: The question is, what does the regular HANES cost?
DR. BURT: It depends what you include.
MR. HITCHCOCK: What is the economy here? That is what I'm getting at.
DR. BURT: Our survey last year was $33 million for the national survey, for the whole 12 months.
MR. HITCHCOCK: $33 million for 12 months.
DR. CYNAMON: But how many people are in the interview and how many people are examined?
DR. BURT: Five thousand exams, maybe 6,000 interviews.
DR. BREEN: How many interviews?
DR. BURT: About 6,000, maybe not that many, maybe less. Five thousand exams is the bottom line.
DR. BREEN: Can you tell us approximately what proportion of the $33 million goes to interviews and what to exams?
DR. BURT: The interview in part of HANES is very expensive, because we do over sample. Say over a period of six years, we may have to go to 100,000 households to find 18,000 that we sample people for. So that is what it was in HANES-3. I think we went to about 4,000 households and sampled people from 18,000. It is very expensive to go to peoples' doorsteps and ask the screening questions, so that is really the big cost.
Our interviews also have to get informed consent for the interview, the exam, stored blood specimen, stored urine specimens and a dust sample we take in the house. So informed consent might take up to an hour in a household. That is an hour of interviewer time. It is very expensive.
The HANES interviewers travel full time, so they cost a lot of money, because we pay for their room and board and their rental car. So the HANES interviewer is more expensive than say a Census interviewer who is living locally. So HANES interviewer costs are very high. They also of course have to during the informed consent process persuade the person to leave their house and drive somewhere for this four-hour examination.
DR. GREENBERG: We don't use the HIS sample at all?
DR. BURT: No, not anymore.
DR. HUERTIN-ROBERTS: I have a question about coverage. Most of what we have talked about is limited to either national or state focused. But what about the territories?
The reason I'm bringing this up is, I spent an awful week trying to find data on cancer in Puerto Rico. It is really not there. It is really not there. SEER doesn't do it. Everything is collected by state. So you just have these large groups of people who are supposedly covered in part of the U.S. that we don't have information on.
DR. BURT: We made a monumental effort to have a special study of Puerto Ricans. We did the sample design, where we would have a sample in New York and a sample in the Island. It is part of our data collection contract as an option, should it ever be funded, and we even put it in as a CDC initiative. But it is like anything else.
DR. HUERTIN-ROBERTS: Right. I think what happens is, the territories just keep getting forgotten, and there are these large populations that we have no data on their health.
DR. CYNAMON: It's not forgotten so much as -- we too have done methodological research, to try and determine how best to do a survey there. We really don't believe that a telephone survey alone will provide decent estimates, because most of the phone lines are in San Juan, and even then, the coverage is not good. Even though some telephone surveys are being conducted there, and they do --
DR. BREEN: SEER used to have a registry in Puerto Rico, and it didn't work out. The quality wasn't up to snuff. So I think it is also considerably more expensive to do them because you need to build a lot of infrastructure.
My guess is it is similar to some of what we heard in the hearings related to American Indians, where the priorities for research aren't necessarily there. Also, Alabama and Mississippi.
DR. HUERTIN-ROBERTS: I don't doubt that, although data is collected now in Alabama and Mississippi. Still, we have a problem. This is quite a significant population that has no information for their health, in particular with the controversy about what is going on at the Diegas navy base. There is nothing to draw on to make a case one way or the other.
DR. BURT: Over the 15 years I've been at NCHS, Puerto Rico -- there are recurring inquiries about HANES. Once every five years, somebody in the division goes down and talks about it.
DR. HUERTIN-ROBERTS: Puerto Rico is just a case in point. It is not that I'm --
DR. BREEN: Maybe we need to think about how much more money would need to be added to our data collection budget in order to collect adequate information, not only on populations that don't have good coverage in this country, but also the territories, taking into consideration that there may be differential costs, depending on where you are collecting and who you are collecting on and all that.
MR. HITCHCOCK: Don't forget, we've got a whole report that addresses this, two years back, that talks about specifically each of the ancillary areas, Puerto Rico being one of them, and Virgin Islands and the Pacific ones, that addresses their data needs, both in terms of what surveys are already conducted there, and there are a few, and specifically what their data needs are.
DR. BREEN: Did you recommend funding for that? So often we end up with unfunded mandates.
MR. HITCHCOCK: There is basically inclusion.
DR. BREEN: Did you ask to make the additional cost?
DR. MAYS: When we send stuff to the Secretary, we usually don't do that.
DR. BREEN: Because otherwise, this comes out of our flat funded budget, and you just move money from Peter to pay Paul.
MR. HITCHCOCK: That is why you need a long range plan.
DR. BREEN: Yes, I think you have to have a long range plan that you can sell.
DR. MAYS: We are going to bring this to a close.
DR. PAISANO: Marcie is right, IHS really doesn't have money that is for data collection, because legislatively we are to provide services to our constituents.
The American Indian population, Alaska Native population, you have all different kinds of universes. As Marcie has mentioned, you have on and near reservation, you have the urban population, you have Alaska, the Alaska Native villages and incorporations, which is basically all of Alaska except the (word lost) reserve, that is only one reservation in Alaska.
Then you also have Puma, which is really a very large American Indian population, with only one reservation. So you have, as someone referred to, the historic areas of Oklahoma. It is an area where the former reservations were for the statehood of Oklahoma.
Then you also have tribes who don't have a land base. So this is all historical. One of the questions -- for national surveys to address American Indians and Alaskan Native populations, to have data would be great, any kinds of data would be great, in general. I don't know if agencies have thought about adding years of data to be able to produce something, maybe three years of data to produce something.
Again, you have all these different universes and issues, again pointing out what Marcie found out. On reservations, maybe not all of them, you have no telephones, and that is especially true in Alaska, where you may have one phone or one radio for the village.
So all this presents all kind of challenges in collecting data. But also, the government to government relationship, in order to be successful in any country, you have to work with the tribal government.
DR. BURT: I know that there was actually a Navaho HANES, and NCHS didn't do it, but they used the HANES model to some degree, and we left them some trailers. I think it was NIDDK and NHII that conducted that in the early '90s. So that was an instance where that data collection did work. I don't know what logistics they went through to make that happen, but it has happened.
DR. MAYS: So you have an Hispanic HANES, a Navaho HANES. What else? Oh, you didn't do it.
DR. BURT: It was funded and conducted by others.
DR. BREEN: The other thing we haven't talked about is the SIAN study, which I think people would not say was a success, but it was a study of American Indians and Alaskan Natives by the MEPS basically, by AHRQ. AHRQ did that.
DR. SCHWARTZ: But we couldn't continue to get funding for it.
DR. BREEN: Right.
DR. MAYS: What was the name of it?
DR. BREEN: SIAN.
DR. MAYS: Oh, I get it.
DR. BURT: You hear anecdotes. I don't know whether they are true or not. I know someone at some point in my career said maybe half of the cost of the survey was just getting on the reservations. That is a huge cost, just to get in the front door. That scares other people away, hearing that, I have to say.
DR. BREEN: I think it was expensive, but the other thing is that people in the communities felt that the questions were poorly designed, poorly administered and poorly analyzed, because there was no consultation with people from the communities who had an awareness of the cultures and what they were doing. So that is a model, but it would need some revamping if it were to be revived.
DR. MAYS: I want to thank the four of you. I think that what you have put on the table, there is the side of, everybody is looking for donors, sponsors, funder, whatever. But I am happy to hear that if something were to come down the pike and there were something there, that you actually have considered these things quite a bit, and that you are actually in a situation where you would be ready to go.
So I think that to some extent, maybe this should be noted, that this has been thought about at these levels, and that it exists, planning efforts have been engaged in. So it is not like you would actually start from scratch.
DR. BREEN: I think we got the high end estimates of surveys, including exams with considerable screening and all of that. I think maybe SLAITS could provide the low end. Do you have a ball park estimate?
(Simultaneous discussion.)
DR. CYNAMON: A 25-minute interview of 2,000 cases per state, households with children, costs about $12 million.
DR. BURT: But we still haven't worked out licking the phone for the biologic specimens.
DR. BREEN: How much was the total for that?
DR. CYNAMON: Under $12 million.
DR. BREEN: So roughly $12 million.
DR. MAYS: So that is for all 51 states.
DR. CYNAMON: Yes, and a 25-minute interview.
DR. GREENBERG: What was the total number interviewed?
DR. CYNAMON: 102,000 in two weeks. It is completely yours, the sponsors. They are not sharing it with other agencies that want to know other things, although we have put together surveys for several sponsors.
DR. BREEN: So you are saying you're not producing public use tapes?
DR. CYNAMON: Yes, we are.
DR. BREEN: It can just be focused on one topic, not multi-purpose.
MR. HITCHCOCK: My earlier question about the geographic level that somebody could access in terms of data tapes. Is it like a city or a state?
DR. CYNAMON: No, not a city. It would be definitely state. Right now, SLAITS samples have been designed for state estimates, so that is what we have.
DR. COOPER: So there is no information on the District of Columbia?
DR. CYNAMON: Yes, it is considered a state. We consider it a state.
DR. MAYS: Can I just ask a final question? You won't own the data in New York, right?
DR. BURT: No. It would not be a public use data file.
(Simultaneous discussion.)
DR. MAYS: Thank you very much. This has been very useful. I would love to be able to promise you something, but we can't.
For those of you who are sticking around, you know that the eighth floor has a cafeteria.
(The meeting recessed for lunch at 12:10 p.m, to reconvene at 12:55 p.m.)
DR. MAYS: I want to do decision number one, one letter, two letters. Based on what I heard today, I think that we would probably be better off with two letters, one letter on the health plan and one letter on these surveys, because I think this is complicated.
Then, I also would feel like -- unfortunately, Harvey couldn't be with us in the afternoon, so we are going to need to figure out a strategy for a discussion with him about the health plan stuff.
Here is what I'd like to propose. Let's deal with the health plan stuff now, in terms of how will we get it done. Here is my proposal, which is, Olivia is right, we do have a few recommendations that we can look to the larger report for. So I want to pull those.
Marjorie, can you work with me for the second letter, which is the health plan report? What we are going to do is pull some of the recommendations that appear in our populations report.
DR. GREENBERG: This?
DR. MAYS: I don't know what that is. This is what I know our report to be. The report looks like this. Why don't we send it to everybody via e-mail? You can use that today, but I don't know what that is. This has the summary, it has everything, it has all the recommendations.
We will set up a call specifically with Harvey, because he has some suggestions. We will come up with -- like Dale did wit the other letter, we will come up with a draft, and then we'll work from there.
DR. GREENBERG: Draw on the recommendations that are already in this report I'm going to get from --
DR. MAYS: Yes, and we will work with Harvey and Kay. What we will do is, we will have a conference call with lead staff, Harvey, Kay and Stan.
DR. GREENBERG: Who was the last one?
DR. MAYS: Stan Ettinger.
DR. GREENBERG: I know Stan. He has been involved with this also.
DR. MAYS: He was for awhile. He was the one that was sitting in before Harvey became involved. So it is up to them. If he doesn't want to sit in, that's fine.
DR. GREENBERG: Do we want to the extent possible to include all the subcommittee members?
DR. MAYS: Not yet. What I want to do is get the draft together. Like, in the first one we started in a small group to get the draft together for people to work with. Then once we got a draft together, then we got everybody, we sent it to everybody, we got everybody involved.
DR. GREENBERG: I'm just wondering though, if you haven't had discussion with the subcommittee members about the need for this letter --
DR. MAYS: No, we did. We talked to them in our July 17 phone call.
DR. GREENBERG: Okay, and other members around that call?
DR. MAYS: Yes. We didn't have it as a separate letter then, but it was an issue that we said we wanted to talk about.
DR. GREENBERG: I just want to make sure we get the subcommittee members on board, because the last thing that I like to happen is for any members of a subcommittee to be questioning something when it comes to the full committee.
DR. MAYS: One person that we will not be able to get for awhile is Don Steinweck.
DR. GREENBERG: Where is he?
DR. MAYS: He is on a boat somewhere.
DR. GREENBERG: Really?
DR. MAYS: Yes.
DR. GREENBERG: He's on an extended vacation or something?
DR. MAYS: He's on a cruise. That is my understanding. He hasn't been at any of this. He couldn't come tomorrow, so there is nobody from Quality, that tomorrow is going to be --
DR. GREENBERG: At the AHRQ meeting.
DR. MAYS: Yes.
DR. GREENBERG: And Bob Hungate wasn't able to come either?
DR. MAYS: No. He had asked not, and Doug can't do it. So Don, I don't even know when he is back. Gracie can probably tell us when he is back. She has what conference call he is able to join us for. So Eugene has been on, so Don is the only one. Eugene sent us an e-mail, and he wants to talk to us after this is over to know what's what.
So that would be what I would suggest. I'll follow your suggestion that we will talk to Eugene, because we are going to update them, anyway.
DR. GREENBERG: I'm just trying to get a little bit of understanding of what the letter on health plans, generally what it is going to be addressing.
DR. MAYS: She has my report, but generally what we would be saying in that is -- I'll back up, because I could probably in my head somewhat say it. Throughout the course of hearings in the subcommittee, it has become very clear to us that there is a need for data that can improve the quality of health of racial and ethnic minorities. In addition, there has been several scientific studies that have demonstrated that there is a difference in not only health status, but in terms of access to care and quality of care.
In our hearings, we have heard from individuals, and we'll say something about what we heard. Therefore, one of the recommendations that we would make is that at the level of health plans, that the collection of data on race and ethnicity was served to be beneficial for us to be able to eliminate and reduce health disparities.
DR. GREENBERG: To monitor that, I guess.
DR. MAYS: Okay, monitor. Then we would probably go to the information in the hearing that you talked about. It seems to be the thought of some that while this cannot be done. Then we would talk about what Edna has moved ahead and done. We would talk about our earlier recommendations for the collection of this data in the Medicaid program.
We would talk about our letter in terms of SCHIP, that we have suggested it there, and then we would list that it is also important for private plans to do this.
That is the letter. I can kind of in my head generate what I think the letter would be about.
DR. GREENBERG: Okay, fine.
DR. BREEN: It is very separate then from the other letter.
DR. MAYS: If it was going to be with the other, I would do it differently. But once you said you wanted it separate, I could give you a different rationale for what to do.
DR. GREENBERG: And you're going to work on that?
PARTICIPANT: I'll work with --
DR. MAYS: So I'll work with Audrey on the health plan, and then Dale has started the survey one. So that is how we will split the two letters, okay?
DR. GREENBERG: Thank you.
DR. MAYS: So let's now move to talking about the letter on population-based data. Here is what I'm going to suggest that we do. Let's put up the recommendations. Dale, if you could put those recommendations up on the screen, or you could just take them off the wall. Then what we are going to do is, when we get the report, point you to the places in the report where there is another set of recommendations.
But while we are waiting for her to bring the report, what I'd like to do is open up the floor for discussion, because I know that for several of the recommendations, there are some issues. Who wants to start?
DR. BREEN: Which of the recommendations are you talking about? These?
DR. MAYS: Yes. Russell, do you want to start?
MR. LOCALIO: Well, it seems to me I learn more every minute, including at lunch. But you must understand what my perspective is. My perspective is, people should be able to get as much information as possible from whatever is out there.
So I think a recommendation should be, just as we are talking about standards, we ought to really be thinking about standards, how can we help to promote standards, in terms of this information, whether it is surveys or how data are collected, so that people who are using these data can use them consistently.
One of the things that I learned at lunch was that unlike some other areas where standards are particularly easy to implement, we now have a very diverse population where you cannot ask the same type of question for example in different populations, because the populations are so very different.
However, it seems to me you still ought to be able to develop a series of standards that would be common across populations, to the extent you want to compare populations or ethnic groups or geographic areas. If you don't have standards, you are never going to be able to compare groups.
I think from the discussions I have heard in the last few meetings that that is one of the primary goals, is to have people be able to say, this is a special population or a special group, and our needs are different, for worse or better, than somebody else or the general population. Unless you have some type of standard, you can't do that.
DR. HUERTIN-ROBERTS: Let me chime in. I don't disagree with you at all. I think that beyond just having a certain set of questions that you would use for everybody, I do think that you can have some core questions that get at the same ideas, the same meaning, but they will have to be written and expressed in different ways for different populations.
MR. LOCALIO: That's for sure. For your daughter, she is going to have to know that something that is called a hoagie in this part is called a sub up in Boston. That is kind of a simple example, but we have to understand that a person down here doesn't know what a sub is, and a person in Boston has no idea what a hoagie is.
MR. HITCHCOCK: Or what is a grinder.
MR. LOCALIO: So we have to make sure that things are translated. But so long as we are talking about something that is this long and has bread on the outside and various stuff on the inside and everybody is talking about the same concept --
DR. HUERTIN-ROBERTS: And we may not be, I don't know. I guess these kind of issues can be addressed, that they need to be addressed. What we have done in the past is not pay attention to them, to think that we can use standardized survey questionnaires for many populations and that will work.
I think probably in the last ten, maybe 20 years, we have moved away from that. But I think there is still a lot of work to be done.
For the surveys that we are talking about, I think that we shouldn't go on assuming that this is an insoluble problem, but we should go in and we should make recommendations that this problem be addressed, that this is something that needs to be considered and assessed when we are talking about standardization as far as questionnaires.
What does standardization mean? Does standardization mean the same terminology, the same language, or does standardization mean different items getting at the same information, but phrased differently, perhaps using somewhat different concepts, asking whether different concepts are commensurate, whether you use one concept in one group and another concept in another group, are they comparable.
My backgrounds is in mental health. I think that that is a real problem when you are talking about mental health cross culturally, things like depression and schizophrenia. Where does schizophrenia exist, except in the mind of psychologists? So I think that these things need to be addressed, but that doesn't mean that they can't be addressed. I think that is the task ahead of us.
MR. HITCHCOCK: They have been addressed in cross-cultural research over the years. You can look at the plan of operations of Hispanic HANES and read the part about translation, and the conceptual and literal sorts of equivalences that they strove for back then.
DR. HUERTIN-ROBERTS: Yes, I think it has been done. It has not been done consistently.
DR. MAYS: Let me try and address this with a suggestion. Let's maybe in the letter talk a little bit about guidance. If you are going to do that, it is also important that this data be collected in a way that there is -- you have to help me remember what the terms are, which is horrible, because California is so good on this, but that there is concept equivalence.
The standardization is like, for example, you decide that you are going to ask about income, but in asking about income, you then try and determine how to ask about it so that it is equivalent across. So for example, if a population is an immigrant population, and what you are really trying to find out is the extent to which they have income available to them.
You remember for example, in the immigrant population it asks how much money they sent home. Again, there is a very consistent contract of why the person has come over here, which is to quite often sometimes to make money, with the understanding that they send X amount home. Then we look and say, you make $6,000 a year, and they are like they are living off of $2,000 a year. So we would know that for immigrants, we would have to ask that in order to be able to compare them to another population about monies available. So that is like this conceptual equivalency we sometimes talk about.
Then we would also make sure that the language is there. Then we would also make sure that would be things that you would specify, because we wouldn't take for granted that they would do it. I think we often think, if we do language we are okay.
I think the prevailing notion, and I don't think we should get into it with them though, is that there are even questions about back translation. I don't think back translation is the way to go anymore. In fact, translation, all you are doing is making the person fit into the box, and the box may not even be the right box. You find a word for it, but the concept really isn't their concept.
DR. HUERTIN-ROBERTS: Right. In the context of our lunch discussion when you were talking about that section six, I think one way around this, and one reason for doing it is community involvement. To the extent that you can have people from the community, stakeholders, whatever you want to call them, to the extent that you can have them involved in design and translation and development of instruments and so on, you get much closer, you have a much better chance of reaching cultural appropriateness, of having concept equivalence. Instead of having researchers sitting and trying to second guess a group of people, that you work with the people, you have a much better chance of succeeding and making fewer mistakes.
DR. MAYS: I'd like to ask in six about something, and whether or not you would want it as a recommendation. In the hearing that we had in L.A. for the Asian, Native Hawaiians and other Pacific Islanders, what I thought was absolutely fascinating is the extent to which the Census has now developed these community data sites. That really seems to be helping to meet the need.
It is kind of twofold. The Census uses that when they get ready to go into the community, and at the same time as soon as the data comes in, they do training.
For those of you who weren't there, what happens is that the Census has put not just Census site for researchers, but they have Census site for community groups, and they train the community groups on how to use the particular files for their area. So what those community groups then do is serve as a resource in the community to constantly give away the data. So if people are needing to write grants, they go to them. When they want to understand about the population, they go to them.
DR. BREEN: How many of these do they have?
DR. MAYS: When Stuart went on a community with me, he has no more than like a bachelor's degree, and another one had an MPH. But they have really gotten to be very familiar with the data. So they turned out to be great for the community, because they go and tell them what the Census says.
I don't know how many of them there are.
DR. BREEN: That is throughout the country, or just in California?
DR. MAYS: No, it is throughout the country. We have two in L.A.
DR. BREEN: We should build on those. The Census has very little on health, but you need Census information in the context of population data.
DR. MAYS: So what I was going to suggest in six is, if we could also wrap in two things, and it may have been someplace else, about the training of individuals from these particular racial and ethnic groups, who can utilize the data. Then also making use of the placing of data sites back into these communities, so that the survey data can be used.
DR. BREEN: Marjorie and I were just talking, that on top of that infrastructure that the Census Bureau has established or created and the training that they do would be useful to bring in the NHANES, the HIS, the health surveys and that data as well. That would give a lot of bang for the buck, too.
DR. MAYS: Is the Census within HHS?
DR. BREEN: No, that is in Department of Commerce.
DR. PAISANO: They are called Census information centers.
DR. MAYS: Yes, CIC.
DR. PAISANO: And basically, each state has a state data center. Then the Census Bureau supplemented them with the Census information centers, and the focus is looking in minority populations.
The way they were selected was, the Census sent out an RFP or something like that, and they were selected based on competition with a bunch of other groups. So you have them as well for the African-American population, Hispanic, American Indians and Alaskan Natives, Asian and Hawaiian-Pacific Islander. Then additional groups that are probably community-based.
But their list is up on the Census Bureau website for the Census community information centers, as well as state data centers. The program for the state data centers has been going on for a number of censuses. The Census information centers really started after Census 2000, although there was an effort after 1990 to do the same thing. I'm not sure if those that were selected after 1990 are still functioning.
DR. MAYS: The one other thing that they do is, I think some of these individuals were involved -- where this also came from as I understood it from talking to someone afterwards, is that there is a lot of involvement by some of these community groups in helping to get people to participate in the census.
So again, this would be something that the health surveys would also benefit from, is having this community kind of involvement.
MR. LOCALIO: But is the suggestion that you get Commerce Census to start working with health data?
DR. BREEN: They already are. They collect a lot of it. DR. GREENBERG: They already collect some. The Census Bureau collects the health interview survey.
MR. LOCALIO: I understand, but they have not traditionally as I understand it -- the state data centers, at least the one I dealt with in Pennsylvania, had a lot of -- the big customers were businesses. They weren't looking for health information, they were looking for demographics and SES stuff for business.
DR. GREENBERG: And other economic stuff.
MR. LOCALIO: And that is consistent with what Commerce would do. But could these same entities be interested in providing health data?
DR. GREENBERG: Maybe if they were supplemented.
DR. BREEN: It might make sense to have the Department of Health and Human Services supplement what is already going on with Census through Commerce.
MR. LOCALIO: The same people in the same data centers?
DR. BREEN: These community information centers, at that level it would make sense to add in the health data. I'm not sure about the state data centers, because I don't really know what their function is, and if it is largely a business clientele that they have, I don't know to what extent they are interested in health. Though there is health as a business, too.
DR. MAYS: I guess I was making the recommendation more in terms of the CIC. Right now, it doesn't make a lot of sense for the health surveys to be given to them because the numbers are too small for some groups. Now, it would make sense where there are African-Americans, it may make sense in terms of those that focus on Latinos.
Now we are beginning to talk about targeted surveys, population defined surveys, whatever the term is. It would make sense for the CICs to work. You want to make sure that when you go out to do it -- I think it was Marcie that was saying people don't like to be targeted, so you would want to make sure that the community actually did the outreach. Then you would want to make sure that the community got access to the data when it comes back in. It couldn't go down to -- again, it would be the confidentiality issues; it is not going to go down to block level, but it would go down to some reasonable level. There could be a public data use center.
So I don't think it is at the state level. I think it is really that we want this recommendation to be on the CIC level.
DR. COOPER: It is really a nice concept. Also, it empowers the community to be more engaged, so that even when you get ready to take it to the next level, --
DR. MAYS: So can I ask you all to take a crack then at the paragraphs that would go for six, based on the discussion? Do we want to talk a little bit about what you see in terms of cultural appropriateness?
DR. HUERTIN-ROBERTS: I think appropriateness concepts, methods, the interviewers need to be trained. I think that needs to be part of interviewer training.
DR. MAYS: Translation issues; you're talking about the language?
DR. HUERTIN-ROBERTS: Yes.
MR. LOCALIO: Just a bit of clarification on this. One of the things that I gathered this morning was that the term contractor came up a lot. Contractor I assume is like Westat and a bunch of places like that, who I think are doing a lot of work --
DR. MAYS: Westat does a lot, I know. I don't know who else.
MR. LOCALIO: And they are hiring the interviewers. There is a difference between training Westat, which may be operating out of -- where is Westat, ten miles away?
DR. MAYS: No, Westat has several different offices also. They have got a California one now, because they have been doing CHIS and all this other stuff, so they are in Fresno.
MR. LOCALIO: It is a different idea from training Westat in D.C. and training local organizations or Westat local offices.
I just wanted to throw that out, because we have to identify exactly who is going to be doing the interviewing, for example, in all of these.
MR. HITCHCOCK: They have subcontractors. I know in Hispanic HANES Westat did that in and around New York City. They handled small --
MR. LOCALIO: There are subsets. In the work that I have done, there is Mathematica in New Jersey, but they had a subcontractor in New York City to do the Puerto Rican part of it.
MR. HITCHCOCK: Yes, exactly.
MR. LOCALIO: And to what extent -- how do we make recommendations about subcontractors, or what is our role?
DR. MAYS: Again, remember, we are sending this at the big level of the Secretary, so I think that what we say is what is required, and then what you hope.
See, that is the problem. What you hope is that whoever is giving them the budget to do it, like if it was coming from NIH or if it is coming from NCHS, that they then insure that what was also suggested is followed.
MR. LOCALIO: So it might be that we would recommend that there be standards for contractors and subcontractors who are actually doing the survey?
DR. MAYS: I don't think we would get that far down.
DR. HUERTIN-ROBERTS: What we would say is that the interviewers need to be trained in cultural appropriateness.
DR. BREEN: The other thing you could say is that you should try to bring the interviewers from the community --
DR. MAYS: That is an alternative as well.
DR. BREEN: -- inasmuch as that is possible. That has actually -- go ahead.
DR. BURWELL: To follow up on what you're saying, the tenets of community-based participatory research would have interviewers of the community, and the HHS or whomever would provide the technical assistance and training for them to conduct the survey.
DR. MAYS: Let me just say one thing. I hate to throw cold water on this, but having sat on the IRB, as much as everybody likes to say community-based participatory research, there are studies in which you do not actually allow a person that is of the community, going into their own community, to actually be the interviewers.
So we have got to be very careful what we are saying, because somebody will take it to heart, and the IRBs will slam you. So it is more that the notion of community-based participatory research really is probably persons who are bicultural, bilingual. I don't want us to get into a particular research paradigm, because there are problems depending upon what the study is.
DR. BURWELL: I guess because the research coordinating council and the IWG community-based participatory research are really trying to get the Department to go in that direction.
DR. HUERTIN-ROBERTS: I'm a strong supporter of CBPR, but I don't think we can mandate that that is the only model that people use. I think it is certainly one model.
DR. BURWELL: It is a MOM.
DR. HUERTIN-ROBERTS: It is a MOM, and it is a good one, but I think that there are other solutions to the problem. But I think that is a very good solution to the problem.
DR. MAYS: I think some of these, like for instance in NHANES, where you are asking people very confidential stuff, the IRB would never let you have somebody who even lives probably too close by. It is an interesting concept. It is part of why some of these studies cost so much, because you can't do this.
So as long as we put what it is we are striving to do, I'm fine with it. But I want to stay away from the method to do it or the person to do it. I don't know if we at the level of the Secretary will tell them about the contractors. But if it is the issues, then they have to follow them.
DR. GREENBERG: I agree, but also, following up on what Russell was saying, I think it is important to articulate the importance of trying to come up with some standardized methods and approaches. That does not mean that you don't have to adapt them so that they are culturally appropriate and certainly language appropriate and all of that, and involve the community, et cetera. But I do think that -- obviously, I'm a big believer in standards, but at the end of the day, if you do things too differently, you really can't compare the data, and it has much less weight.
There is a body of research, and World Health Organization also has been doing a lot of this kind of thing, these international surveys or whatever is comparable, but I think there is research that has been done and can be built on. But it may be that there is a need to bring this together in a more systematic way than has been done, so that everybody isn't re-inventing the wheel, which is different than just one standard, but are actually brought together, published, disseminated, et cetera. I think advocating for that is a role of the committee.
DR. MAYS: Can we then -- Dale, can you go up to the beginning of this set of recommendations?
MR. HITCHCOCK: This was the outline. I can switch back.
DR. MAYS: No, there is one about something, maybe I'm not remembering it.
DR. BREEN: This is the outline.
DR. MAYS: Yes, okay. Let's go to this, when we were talking about flexible approaches. Maybe then what we need to do is to wrap in the beginning this notion of --
DR. BREEN: Best practices?
DR. MAYS: Yes, some standardized -- standards are important. Usually what happens is that there is not money to do some of the actual testing to figure this out. I think that may be the issue. Usually they do a little bit of testing, and then everybody has to go. But it may be necessary to consider the funding of some of the methods.
DR. GREENBERG: That is one thing that is always underfunded. But if what you want is to be able to make comparisons and have relatively comparable data, you are going to have to --
DR. MAYS: Here is what we can do, is talk about funding methodologies that would pay attention to some of the special issues that racial and ethnic minorities are likely to encounter, such as collecting data in populations without telephones, what is the implications of that for the Department's surveys.
DR. GREENBERG: Actually, quite a lot of research has been done on that, as you heard earlier.
DR. MAYS: They may have done it, but -- HHS does it so that they can determine what to do about a survey, but it is not published, it is not out.
DR. GREENBERG: I think people from NCHS have published a fair amount about --
DR. MAYS: Do you think so?
MR. HITCHCOCK: ASA and places like that.
DR. BREEN: I think the federal government is America's biggest wallflower, it really is. We don't publish anything or publicize anything we do. All this stuff has been done. If you go to the methods papers that they publish, 12 pages in from the front of the web page, you will find all this stuff. But you guys spend lots of time on the website, and then you've got to look at these ugly-looking reports and read them, and they are pretty technical. But the work has been done, but it is not really out there in a way people can use it.
DR. GREENBERG: I'm just saying, of all the methodological issues, the one related to telephone coverage has probably been covered better than some of these other ones related to cultural appropriateness and translation and all of that. But that doesn't mean that it is readily accessible for people to use.
I was very interested to hear Marcie's mentioning about some of their findings, that if they asked people about continuous telephone coverage versus intermittent, they find out that that helps with the adjustment, because that is a lot more efficient than having to always use the household survey. Often you don't have a household survey.
DR. MAYS: But it would be wonderful if in some way -- this is the body of knowledge that is available. It would even help states in their planning to realize that someone has already done this, and things don't have to cost what they think, because they now can make adjustments based on this report from SLAITS or whatever.
DR. GREENBERG: Well, that is the sophisticated states. Some of other other people doing these surveys don't even realize they have to make adjustments.
DR. MAYS: We won't go there.
DR. GREENBERG: In a standardized kind of way, so that people can do the right thing or the most effective thing without having to figure it out themselves always. I think that supports this whole idea.
DR. MAYS: Let's ask how to do this then. Russell, I want to ask you, this will be your piece. You can talk a little bit about the issue of standardization, but also with some flexible approaches that takes into account some of the specific issues of racial and ethnic minorities.
MR. LOCALIO: That is actually the letter, right?
DR. MAYS: Yes, this is going to go back into the letter.
MR. HITCHCOCK: Is it procedurally appropriate for us to contact by e-mail people to come and talk to us?
DR. MAYS: Oh yes. It is an informal discussion, so it's fine.
DR. GREENBERG: Yes, one member can always communicate whatever.
DR. MAYS: Particularly in the service of getting some work done.
DR. GREENBERG: Oh, absolutely.
DR. MAYS: I think we want to bring up the notion of best practices, and that we recognize that the Department has done quite a bit of work, but the ability to be able to harness those findings -- how do we get them to do it? What do we say we want people to do? Is this where we come and have a consensus meeting?
DR. BREEN: We want people to employ best practices. That is the goal, right?
DR. MAYS: Yes, but we also want to get people to --
DR. BREEN: Then we have to know what they are, in order to do that. What CDC and NCI have done with what they call evidence based procedures, which is kind of the same thing, only that would be for public health service delivery and that kind of thing is, they have started to put together these community guidelines, which are actually published. That is the second level, away from the 12th page on the website, so that they get disseminated. They are also glitzier looking. They get disseminated by mail, they are on the website, and they are a bit more obvious.
Then NCR recently came up with something called Planet, which is a really glitzy cancer control website, which basically provides a lot of things, including evidence based stuff, whatever you want to do. If you want to do an intervention or something, you put in your intervention what you want to do, and then you say, I want the best practices, and up they pop.
DR. BURWELL: Where is this website?
DR. BREEN: It is cancercontrol.gov, and then Planet is on there.
DR. GREENBERG: Planning?
DR. BREEN: P-l-a-n-e-t. I forget exactly what it stands for.
DR. GREENBERG: Planet?
DR. BREEN: Yes.
DR. GREENBERG: Like Planet Earth?
DR. BREEN: Planet Earth, the Daily Planet, yes. But anyway, the best practices need to come to light in some way like that if they are already existing, and if they are not already existing, then there probably needs to be conferences to have them. But it sounds like they already exist, is that right, in many, many cases?
DR. GREENBERG: Some do. There is a lot of work that has been done. Whether it has been succinctly pulled together, I don't know, but if you're not starting from scratch --
DR. MAYS: Here is where it would be useful. I'll give you an example. NCHS may do things one way, and then somebody else will do them another way. So part of this issue, and that has been the struggle, this is why I was hoping either Jackie or Jennifer were on, -- see, what is different is that at NIH, you have developed some of these by independent contracts, or independent grants, right? There are R0-1s that you have drawn this from?
DR. BREEN: Some of them, yes, but people have come together and evaluated best practices using meta analyses. Evidence based medicine has replaced expert opinion. What people have done is taken studies and evaluated them, compared them, meta analyzed them, whatever, in order to come to a point where they have a sense that there is evidence adequate to say that this will work in X, whatever community.
Part of the task is to see what has succeeded in a population for it to work. If you want to try that in another population where it hasn't been shown to work, that would be a great study, and very fundable.
DR. HUERTIN-ROBERTS: I think the Planet and the community guide are both great tools for dissemination, but it took a lot of work to get there, to decide what even needs to be on there. I think that is where we are. There is information here and there, but there is no consensus on it at all. There has not been any effort to compile it and integrate it and evaluate it in any way.
DR. MAYS: That is why I'm asking how you got to determinations that this was best practice, or I'm sorry, evidence based procedures, so that that is what we can talk about suggesting.
DR. BREEN: Cull through the literature.
DR. MAYS: Who?
DR. BREEN: Well, we had a lot of contractors. They weren't Westat, but other contractors who helped us go through the literature. One person was assigned to get this done, and he did it. So he just took the lead, and I would say John Turner spent the last year and a half of his life, maybe 50 percent of his time, and CDC has been active doing it, too.
DR. GREENBERG: Did you do any consensus conferences?
DR. BREEN: Not for that.
DR. HUERTIN-ROBERTS: For the community guide, that was discussed at different meetings. The CDC thing on cancer control. So the community guide was definitely a part of broader meetings, and that was addressed and put out for review and put out for discussion of these meetings and so on.
DR. MAYS: Can I suggest then that maybe what we need to do in this instance is to consider whether to suggest a consensus meeting? I think what you want to do is to bring -- like for example, there are several academic survey sites. There is Westat there is NCHS, and there is NIH. You would want to bring them together.
For example, I could call Tom Smith at NORC, and he has little papers that he had to do. He has information. Then you go and you talk to Jackie, and you find Jackie Lucas has one on something for NHIS. I don't think you can go to the literature.
DR. BREEN: There is a -- I forget what it is called, but it is the survey methodology organizations, and they are the ones that are already doing this, and would know what the best practices are, because that is their content, that is their meat and potatoes.
(Simultaneous discussion.)
DR. MAYS: It is called health survey research. They have a meeting every other year. Marcie Cynamon is actually the person who has gotten two of the publications out from that group. It is a very small meeting.
MR. HITCHCOCK: It might not show up if you were to do a MED or something like that.
MR. LOCALIO: It is hard to get that stuff.
DR. MAYS: Yes.
DR. BREEN: I think it would be good to assign the task of pulling this information and determining best practices to somebody, rather than just saying a conference should be held. The conference will be held, people will go home, and we will be left in the same position we are with these hidden websites and whatnot.
DR. GREENBERG: What might happen is, some agency or whatever would take this on, bring it together, and then the next step might be a conference meeting.
DR. HUERTIN-ROBERTS: A conference and maybe a network, but I can't see this being settled in one conference.
DR. MAYS: What is the recommendation then, so that I understand? What is the recommendation here?
DR. BREEN: Develop best practice guidelines.
DR. HUERTIN-ROBERTS: Some person or body be charged with developing, or compiling.
DR. GREENBERG: This is the obviously ideal appropriate thing for NCHS to do, not drawing on their own work, but everybody else's work.
DR. MAYS: Only if -- that is the problem with NCHS. They have a way that they do things, and they really are loath when you try and do it differently, I'll be very honest about that.
DR. GREENBERG: The thing is, NCHS should be able to do this as the federal statistical health agency, but not to just say this is the way we do it. A lot of this best practices that we are talking about for these purposes are for types of surveys that NCHS doesn't even do.
DR. BREEN: How about if NCHS and Census -- I think there should be a couple of different groups that are heavily involved in this, and really know what they are doing, because then you will get debate, which is what we want.
DR. GREENBERG: If you are making this recommendation to the Department, it goes to the Data Council. They decide how to do it.
DR. MAYS: That is steeped in NCHS.
DR. BURWELL: No, the Data Council is in every component of the Department. NCHS is just one.
DR. MAYS: Okay.
DR. BREEN: The Data Council is co-chaired by ASPE and AHRQ.
DR. MAYS: What would you do if you had this recommendation? Aren't you the co-chair?
DR. BURWELL: Well, Jim Scanlon and Data Council would have to decide that is what we need to do, and then we would work on it.
DR. MAYS: No, no, I mean if you were given this by the Secretary to look at this issue, what would you do? You wouldn't do the work, would you?
DR. BURWELL: Dr. Stinson might give us some funds, but we --
DR. GREENBERG: There is some money for it, yes.
DR. MAYS: What happens if we made suggestion to that, like say GAO and Census and NCHS? What would happen then? Is that too detailed for us to give to the Secretary? The Secretary only is dealing with health, and the Census which is in Commerce, has a whole lot of expertise that you would want to bring in. So that is why I'm a little concerned.
DR. BURWELL: This is a consortium of federal statistical agencies. That is probably where you might suggest an organization like COFAS, because it represents beyond the Department, every statistical agency.
DR. GREENBERG: But you do want to focus on health.
DR. MAYS: Yes, but the expertise is also outside of health.
MR. LOCALIO: One of the things from this morning that was interesting was when Marcie mentioned how much they are dependent on Census in order to post stratify their data to get the estimates they want.
DR. BREEN: Every survey in the country is dependent on the Census.
MR. LOCALIO: So even though Census ia a different Department, everybody is dependent on Census, so they are in it whether they like it or not.
The other point I wanted to make is, I am amazed from Februry until last week how much progress was made on the various federal websites with the privacy rules, in terms of consolidating information. We went from absolutely nothing, almost nothing, to the ability of somebody to get on the web and go to NIH, to OCR and other sites, all from a couple of websites, and get information, not just in English, but in Chinese, Tagalog, Spanish, Italian, about a dozen languages. That is an awful lot of work in about four months.
So somebody got the message -- is it us? -- somebody got the message to get serious and get this stuff out there, because people are really hurting. Well, it is the same concept here. The type of information is different, the issues are different, but it can be consolidated if somebody takes some leadership in doing this.
DR. BREEN: It needs to have priority from the top and then it gets funded, and it gets done immediately.
DR. MAYS: Here is the recommendation, so that we understand it. That is, we request leadership from the Department on evidence based procedures for the conduct for the collection of population-based data in racial and ethnic groups.
I'm losing my ability to just spin this, but it is something like that, that many arms of the Department combined with other groups such as Census -- what did you say the group was?
DR. BURWELL: COFAPS. Census is a part of it.
MR. HITCHCOCK: That is where Kathy Wong came from, right?
DR. BREEN: Aren't they more of an advocacy group, though?
DR. MAYS: We'll use them. We'll use them for now.
DR. BURWELL: I go to those meetings, and just about every federal statistical agency is represented in that consortium.
DR. MAYS: So we will say something like, in the conduct of its work, many departments in the federal government have conducted methodological assessments to determine best practices for the conduct of its work. However, it is difficult to --
DR. BREEN: Work needs to be consolidated and easy to find place, or --
DR. MAYS: However, the public could benefit from evaluation, consolidation and dissemination of this work. So then do we say something like, groups like COFAPS might serve? Or do we just leave it at that?
DR. BREEN: We can leave it at that.
DR. MAYS: Okay. I didn't write down everything I said.
MR. HITCHCOCK: It is on the screen up there.
MR. LOCALIO: The consumer of this information is a lot of people. It is federal, state, local agencies. It is the population groups, researchers.
DR. GREENBERG: Foundations.
MR. LOCALIO: Foundations. There are a lot of consumers for this information.
DR. MAYS: Should we then add a sentence that, users that would benefit from this include? Do we need to say something like that? Users that have indicated that they can benefit from this, because in our hearings we have had that. Users that indicate they could benefit from this include, and then we'll put some of the list of what you just said. Communities.
MR. LOCALIO: State and local governments, foundations, agencies, federal agencies. Is that broad enough?
DR. MAYS: Yes, that's good, because that is compelling. That is good.
DR. BREEN: Say it again for Dale, so he can write it up.
DR. MAYS: Users that indicate that they could benefit from this include community groups, -- he has got local and state government, foundations.
MR. LOCALIO: Federal agencies, foundations, as well as researchers.
DR. GREENBERG: Bobby made a good point. It is not only evidence based procedures and best practices for collection of the data, but also --
(Simultaneous discussion.)
DR. MAYS: What is the inter-agency collaboration?
DR. BREEN: Dale hasn't finished up there.
DR. MAYS: Where do we find these things?
(Simultaneous discussion.)
DR. MAYS: Survey design and reporting. Take me back to inter-agency collaboration. What we want here is to make a recommendation that the surveys, which we will say what they are at some point, that we realize that there is cost. What else? What do we say to bring about why there should be an inter-agency collaboration?
MR. LOCALIO: Expertise.
DR. HUERTIN-ROBERTS: Pool resources, coordination of efforts.
DR. BURWELL: It is responsive to the President's initiatives.
MR. LOCALIO: Unnecessary duplication.
DR. MAYS: You all know these buzz words. Let's get to the hard part here. Some of this we can do fairly easily, but we have to get to the hard part. That is, what are we actually going to recommend in terms of surveys?
DR. BREEN: Well, Marcie suggested we use special surveys, because it covers everything we want; special studies.
DR. MAYS: She said special populations.
DR. BREEN: Studies of special populations? Is that the term she used?
DR. GREENBERG: But these recommendations are going to focus on populations as defined by race and ethnicity, right?
DR. MAYS: Yes.
DR. GREENBERG: That doesn't mean that you're not looking at income groups or age groups or whatever.
DR. MAYS: But it is to deal with the disparities that exist.
DR. GREENBERG: Yes, I think that is the focus here, whereas obviously like what Marcie was saying she is responsible for, that is broader.
DR. BREEN: I would like to propose, and this is for two reasons, one because I think it is a very reasonable way to do it, and also, the literature backs me up on this, and also because it will increase our count to a reasonable number. That is this guideline for the national health care disparities report. It says, the national health care disparities report should present analyses of racial and ethnic disparities in health care in ways that take into account the effects of socioeconomic status.
If we keep in mind that we want to continually do both of those, I think we are going to be better off in the long run in a million ways. So I would incorporate that recommendation.
DR. GREENBERG: There are actually some guidelines from CDC not even do identify race and ethnicity without --
DR. BREEN: Socioeconomic status? Okay. So there are a number of guidelines then that we can probably draw on for that.
MR. LOCALIO: Translating that, I think what they are saying is, we want to make sure that if there are disparities, they are based not on income.
DR. BREEN: No, they want to make sure that if there are disparities, we can be clear as to whether they are in a regression model, we can parse out SES versus race. Also, you can often find within a racial ethnic group income or education or a socioeconomic status gradient. So I think you want to be able to any of those things with your data.
MR. LOCALIO: I want to back up to something I said in the beginning to Vickie. I have some real concern about the definition of the racial and/or ethnic subpopulations. I just have some real problems on who is going to identify them and how, what are they, what are the ones we want to focus on.
This is the next step, talking about, once we have identified them, we can talk about disparities. But we must understand that -- this came up maybe this morning -- there are big differences among the Mexican-Americans, the Cuban-Americans and the Puerto Ricans, even though they may all be called Hispanics.
DR. BREEN: In 1991, Audrey Burwell and others had a conference, an agenda setting workshop with recommendations, but if you look at these, you have the OMB categories.
DR. MAYS: Before we do that, let's talk about what has changed and what the issue is.
DR. BREEN: Because it does take into consideration what Russell said.
DR. MAYS: As of 1997, we now collect data so that people can identify themselves with more than one race.
DR. BREEN: Right, and we have three percent of Americans or less doing that. But we also want to take into consideration the '97.
DR. HUERTIN-ROBERTS: I see that as a separate issue, I really do.
MR. LOCALIO: It may be happening in only three percent, but I was able to look at the American Indian and Alaska population, and I think I found that 800,000 Hispanic, American Indian, Alaskan, out of a total of 4.5 million. That is a big chunk. That is not three percent. That is a big number.
DR. BREEN: No, that group has the largest --
MR. LOCALIO: When you get into some of these groups which I think everybody is interested in, we face these problems.
DR. MADANS: Vickie, I just wanted to let you know I joined. This is Jennifer.
DR. MAYS: Hi, Jennifer.
DR. MADANS: Sorry to interrupt.
DR. MAYS: No, that's okay, thank you.
DR. BREEN: The reason I brought this up though is that it has got a lot of subgroups under the Hispanic, black, American Indian and Alaska Native, white, not of Hispanic origin. Then we have to divide up Asian and Pacific Islander because they are grouped here and this is pre-2000.
Under Hispanic, it has got Cuban-American, Mexican-American, Puerto Rican. Under American Indian and Alaska Native it has only got the Pacific Northwest tribes; we might want to discuss that. Under Asian Pacific Islander, it has got the same groups that were shown to us in Los Angeles at the Asian Pacific Islander hearing that we had. Cambodian, Chinese, Filipino, Guatemanian, Hawaiian, Mong, Indian, Korean, Laotian Pacific Islander, Pakistani, Samoan, Thai, Tongan, Vietnamese and inter-racial or ethnic.
If you allow people to respond to those, then the analysts can group the mixed race as they wish, cull them out. But I don't think we can -- well, I don't know, I don't think we can -- the government is working on standards to try to figure that out because it creates huge problems with the denominator. But I think it is not premature to say what we should do analytically, but data collection wise, I think it is fairly clear.
DR. MAYS: Tell me whether or not -- we have to decide this, whether or not you think that all those groups, that we should be able to have population-based data for all of those groups.
DR. BREEN: I think in the long run, yes.
DR. HUERTIN-ROBERTS: Let me step back though and ask the question, it is our charge to decide which groups we should recommend data be collected for?
DR. MAYS: I think that in the hearings, what happens is that many groups have come to us and put it on our radar, but they have almost no data. So the question is, as we make recommendations, I don't know whether it is left to the Department if we give them just these broad names of doing racial and ethnic minorities, that they do the right thing, or that we take what we have heard in the hearing and say that these issues are particularly critical for American Indians and Alaska Natives, for a lot of those API groups.
There are segments of those groups in which they have no population data. We have no data about their health status. We only have data about those individuals who have come into treatment for either particular conditions like cancer because of the SEER registry, or because of the Indian Health Service.
DR. HUERTIN-ROBERTS: Is it possible for us, rather than recommending a specific plan of who should be targeted, let's say, to just include the number of groups that we have heard from as having problems in collecting data? So there is an example of this breadth of need, but we are not in the position of saying, we need to talk to this group, this group and this group, because we are not supposed to be developing the plan, the long range plan. We are saying there should be a long range plan.
DR. BREEN: Maybe what we could do would be to put something like -- I think Paul Ong showed a slide, or the guy who spoke before Paul, the very last person to speak --
DR. MAYS: Noah Malone.
DR. BREEN: Noah Malone, showed a slide which was pretty similar to this part of this slide, where he said there are 12 major API groups in the country.
So if we were just to say that, that the expert researcher on this population group indicated that these are the main groups nationally, and they should in the next ten years, if that is how we define long run -- we should have reasonable population-based data on these core -- at least some localities in the United States, where they are concentrated, something along those lines, it seems to me.
I think we have to say something pretty specific, or else a lot of these groups are just going to continue to fall through the cracks. They are lobbying to get better data, they have indicated they want better data, and I think we are trying to set a stage where they can continue to do that even more effectively than before.
DR. MAYS: I want to hear from Marjorie, in terms of what flies in terms of these letters.
DR. GREENBERG: Well, I think you could generally say subpopulations or whatever. It is too general, probably. So I basically agree with what Nancy said, particularly if you have heard from these groups, you have heard from researchers that the groups that at minimum they would like information on are these. I think you should say that.
Probably to prioritize within those is not appropriate, because you don't have a basis for doing it. Again, that would depend partly on interest, resources, communities who would come forward and so forth.
MR. LOCALIO: You can always prioritize population-based. You always have that option.
DR. GREENBERG: But within these smaller groups.
DR. BREEN: Can I make one more point? I think the other advantage of collecting information on socioeconomic status -- because I think if you have both population size of these small populations and the sense of health status and socioeconomic indicators, you may have very small populations that are extremely poorly off, and you might want to skip over some of the larger groups in order to focus on them, just because their need is so great, and you want to be able to provide some baselines for them to move forward with eliminating health disparities.
MR. LOCALIO: It is a product of the severity of the need times the number of people at risk. So if you have a large group and they have severe problems, that would be first. But a large group with not severe problems might not be first.
I just want to say, I like your idea. It might be nice to say you want to name a dozen or so and say, we have heard from these dozen, but there may be more.
DR. BREEN: Yes, so it is not these only.
MR. LOCALIO: So it is not -- the question is not limited to these dozen. These are just an example of the type of populations we have heard from.
MR. HITCHCOCK: The problem with severity though is, you really don't know.
(Simultaneous discussion.)
MR. LOCALIO: You cannot determine the order until you actually define the population and done a survey and asked the question.
W do have a chicken and egg problem here, but we first have to identify the people you start with. You have to have a starting point.
DR. GREENBERG: There too, you maybe start with the less expensive or resource intensive surveys, and you take these, like community HANES or whatever, which are quite costly, to target it maybe more on where you have identified in your preliminary data that there is a real need, or there is a real problem.
DR. MAYS: Is Susan still on? Susan Queen isn't on. Susan Queen sent us a segment for --
DR. BREEN: Special population surveys?
DR. MAYS: Yes. Maybe what we can do is talk with -- I think what Susan has written is actually very good, because it gives you the Departmental language, combined with the Departmental survey.
Jennifer, did you get a copy of it? I asked Susan to send a copy by e-mail.
DR. MADANS: Yes, she did. I did get it.
DR. MAYS: Great. So I think that what we want to do is to play around with this section and include in it, Nancy, your other suggestion about listing specifically some populations. It is probably the lead-in to her section.
What we need to be careful about is to remember to say that we have heard from some of these groups, to talk about these groups and to say that from some of them, we have heard their need. I hadn't heard from everybody on that, which is why we were continuing the other half of the hearing.
But I think you're right, there were 13 or more groups that they put up at the hearing. As a matter of fact, I have the original. It is in some colors, it is only in English. There are no surveys. So I'll pull that material and add that as a little bit of background to it.
Then later, when you have the section talking about language and everything, it is perfect.
Now, for the American Indians and Alaskan Natives, the question is, do you also want to try and generate all of the groups that go there? That is like, what is it, 50?
DR. GREENBERG: 500.
(Simultaneous discussion.)
DR. MAYS: So I don't think we want to do that, but I think we want to mention that there are that many tribes, and the extent to which there are some that we have not actually gotten data from, as well as make sure that we take into account the notion of urban Indians.
DR. COOPER: Is there any way that those are collected?
DR. PAISANO: I wanted to point out, Nancy was talking about the racial groups. We are talking about federal surveys, so we have to keep in mind the directive, and the directives basically say you have to have data -- is it the five major races, plus multi-race?
So I think when all the surveys are designed, that is what is going to apply for getting survey instruments cleared, as well as allowing to collect the subpopulation information under these major races.
DR. BURWELL: OMB's revised standard is the minimum. You can collect as much detail as you want.
DR. PAISANO: That's right, but I'm saying that it needs to be pointed out, although it is the minimum, it allows all this other collection.
DR. BURWELL: Yes, it does.
DR. MAYS: Are you saying that we encourage them while the minimum is what is required, that to solve some of these problems, that going beyond the minimum would be useful?
DR. PAISANO: Yes.
DR. MAYS: Because what is going to happen in privacy, it is just going to collapse them all down.
DR. PAISANO: Right. I'm just saying that in the guidelines it says the minimum, and because it says that, I don't think there has been an effort by anybody to go beyond the minimum.
DR. BREEN: We want to encourage people to go beyond the minimum. I think that is the whole point of this letter and the report.
DR. MAYS: But just so that you know, at the federal data sets, when they go beyond the minimum, by the time they give it to you for public use, it is going to be back at the minimum. Even then, we can't get some of the data because the numbers are so small. So in some of the population-based data sets, they won't give you very much, or else what they have done is scramble them to the point that you worry because the numbers are so small in terms of American Indians and Alaskan Natives or something like that.
DR. GREENBERG: Isn't that why you are wanting to recommend these target surveys?
DR. MAYS: Yes, exactly. I don't know if it helps us to remind them to increase the data collection in those big population-based studies. Unless you go to the data center, you will not really get to use it.
DR. BREEN: In fact, I think that is the argument that should drive the letter. The OMB minimum ends up being the maximum in most federal surveys, because of the sample size and frame. In order to go beyond this minimum to get more detail and more complexity, which is reflected in the actual population, we need to do these special targeted surveys.
DR. GREENBERG: That is one method.
DR. BREEN: We need to get information on other groups, and special targeted surveys is one way to do it.
DR. GREENBERG: One approach.
DR. BREEN: What are the other ways? We should give them a choice if we're going to say it is one way.
DR. BURWELL: You're focusing the letter on special surveys, but there a lot of different statistical ways you can do it, but they have to be looked at in terms of the population, in terms of the cost, in terms of other variables.
DR. GREENBERG: Over sampling.
DR. BURWELL: That is one. Nested studies, nested sampling designs. There are a whole number of things. One grant that we had used an ecological type of approach to tag, so to speak, a certain racial and ethnic group like you do deer. You tag them again when you come across them.
MR. LOCALIO: Tying this in to what we talked about a few minutes ago, there are statistical methods that people have used and are developing to be able to augment surveys, combine surveys, so that you can get the requisite sample size from the population of interest.
You can use BRFSS data, which may be particularly rich because it is state level and it is done pretty frequently, and use that to supplement some of these other surveys, where you don't really have the amount of detail that you need.
But I guess tying this back to what I said before, we have to be able to say that given the limitations on the budget, this type of information on how to do this needs to be reviewed, compiled, consolidated, disseminated, along with the possibility of having supplemental surveys to fill in the gaps.
DR. BREEN: So are those tracks then to these targeted special surveys and in investigation of other statistical techniques?
DR. MAYS: That is exactly what I want to do at this point, is raise some issues about the statistical techniques. The reason I want to do that is because -- I'm glad Jennifer is on, because she can help me, and Dale in terms of the sampling frame.
We usually store it with the Census as the basis of the sampling frame. We move from there and what happens is that there is a limited number of individuals who are being used for several different surveys. So what happens now is, if we keep -- and remember, in terms of the racial and ethnic minorities, they are weighted to represent a lot. So you might have one African-American male, and that person is weighted up to represent 15 African-Americans males.
So what happens when we start doing these statistical methods is that you really are relying enormously on a person to represent a population, and what people tell you -- and I've done enough of these now in terms of both telephone and face to face, particularly in terms of racial and ethnic minority males, who you get does not represent necessarily, particularly in terms of health care and services, who is in the study.
An example of that is, you have to look at the high number of men who are in ward care, jail or all the institutional things we want to talk about, who come into the community. They are usually in none of the surveys. So when you talk about social service or health care needs, you underestimate.
So I get a little nervous if we do too much linkage, because we actually have very few people in terms of racial and ethnic minority groups sometimes that are being represented that we are making great pronouncements about. It is particularly true for males.
So I'm not saying don't do it, but I want to back off of it a little bit, because I think this is what is creating the service need problems for us, is that we are not getting as much accuracy about the needs as if we then went and did another sample, where you have a possibility of that next male being different than the male that is in the other, and then when the two studies are published and nobody can figure out why, then it at least leads you off to do yet another study.
MR. LOCALIO: But I think what you could do is, suppose the two surveys were published and they had a sufficient number of common elements, you could take the data from both of them and combine them. You would then have a combined analysis that is more powerful than each of the ones separately.
MR. HITCHCOCK: Statistical matching?
MR. LOCALIO: No, I'm not talking about record linkage, I'm just talking --
DR. BREEN: He is assuming the two samples are independent.
MR. LOCALIO: Yes, they would be independent samples, but they may have slightly different answers to the same question. Then you can say, very likely the true response is somewhere in the middle.
DR. MADANS: Vickie?
DR. MAYS: Yes?
DR. MADANS: You're concerned because of record linkage, that there aren't -- that we are missing people in the sample?
DR. MAYS: No, no.
DR. MADANS: I think there are two issues here.
DR. MAYS: What I am concerned about is that if we don't call for additional surveys. We only use what we have, and what we try and do is more linkage among that. But we are actually just continuing to use the same people from the same sampling frame.
DR. MADANS: There is a difference between saying a frame is not encompassing an entire population of interest. Most of our population-based surveys are based on a civilian, non-institutional population. If you want to pick up the other parts of the population, you really have to do a different survey on a different population and combine them.
But in reality, what happens is, we over sample -- we generally tend to over sample minorities in the big national surveys. We actually relatively get more of them than we would of the other population, so their weights are actually smaller.
DR. MAYS: But they are drawn from the same urban area.
DR. MADANS: So I guess I'm not as concerned with that problem. I think the basic issue which several people have mentioned is, it is not enough to say you should collect the data by these categories; you have to say you have a sufficient sample to analyze in those categories. Then once you say that, then there are ways of getting the samples, all of which have different costs associated with them and quality associated with them.
The easiest thing to do is increase the entire sample. That is easy and expensive, but it is straightforward. Once you start doing -- every one of these other kinds of approaches has statistical downsides as well. If you do targeted studies, and I'm not saying we shouldn't, I think we should, you no longer have representative data for the whole country. Some people think that is very important, because there is big regional variations.
So you have to look at the pros and cons of the different approaches.
The next step I think on this is, and it goes beyond just the racial and ethnic data but other things as well, we tend to do these surveys very independently. If you really want to be able to have maximum flexibility, you need more of an integrated approach to the large data collections, but then allowing them to pull in other data or get other data so that you have a more integrated set of data collection, rather than a single set, a lot of individual data collections, none of which really meet your needs.
DR. BREEN: Jennifer?
DR. MADANS: Yes?
DR. BREEN: This is Nancy Breen from NCI. One thing we were talking about earlier with -- who was the one that was sitting right here?
DR. MAYS: Vickie Burt.
DR. BREEN: Vickie Burt from the planning division for NHANES was that it turned out that your center has thought about a whole lot of the kinds of things we are trying to think about here. She said that they had discussed the possibility of going -- in order to get better data on small populations in the country, they had discussed the possibility of going to areas where there were big concentrations of those populations, not worrying about the national representation.
She didn't say this, but the way I was thinking of it is, if you use your national denominator as your population base, then this group is going to end up so small. But say Indians in the Southwest are quite an important part of the population, or in the Plains states, are quite an important part of the population, or Asians in Los Angeles, Latinos in Los Angeles.
There are a lot of examples like that all across the country. Apparently there has been some consideration of that. I don't see any problem with that, especially when you have a phenomenon, something like -- and I'm not exactly right on this, but something like two-thirds of Asians live in the three Western coastal states.
DR. MADANS: We did actually do it. Hispanic HANES, those were not national samples. The Mexican sample was only in the Southeast. The Cuban sample was only in Florida, the Puerto Rican sample was only in New York. So it has been done. I'm not saying it shouldn't be done, but you then can't talk about Mexican-Americans. How focused do you want to be? It is no longer a national sample. You can't compare -- what is your comparison group?
MR. HITCHCOCK: The coverage was very high, Jennifer. Before you leave the Hispanic HANES, it got up to around 90 percent of those three groups, I think, by just going to those areas where there were --
DR. MADANS: That would be true today. If we look at just Puerto Ricans, if we did it the same way we would be under 70 percent, I think.
DR. HUERTIN-ROBERTS: But we are not suggesting substituting these studies for the national surveys, so we will still have the national data.
DR. MAYS: Jennifer, say for example if we did a targeted survey, and you do have some data in the national survey, why couldn't we do some kind of standardization to see how the targeted survey relates to the national survey, and then derive information that way?
DR. MADANS: If you are willing to make the assumption that you are targeting a survey to a geographic area, there is no interaction with geography and there have to be other of those kinds of people living elsewhere, then you can try to make some modeling. Generally, if you do the targeted survey, there is not enough people in the national survey to do that kind of modeling; you don't have enough of them.
So don't get me wrong, I'm not saying we shouldn't do targeted surveys. But if you just do a lot of little targeted surveys, you don't have that ability to get a complete picture. If you are doing lots of them, then you probably go back and say, maybe we should do a bigger national survey with over sampling, which is what we do now, and you can do differential over sampling rates, which is exactly what we do now, depending on where people live.
It is easiest to do this if people are geographically clustered. If you know where they are, then screening becomes easier. It is much harder when they are spread a little bit. You look at the census map and you try to figure out where people are so you can design your sample.
I think that it is useful to do targeted studies, but you have to be careful which groups you use, when is it better to do national, and how much more is it going to cost you, and what kind of comparisons can you make.
DR. MAYS: Can you answer those questions? Do you have something that answers those questions?
DR. MADANS: We have general -- we have done general position papers about -- for example, taking the HIS. If for example we were to expand the HIS, let's say double the sample, we know how much of the major groups we would have. Then we would say, all right, if we really want to get one of API Asian groups that we know is more focused in one area, we could do just an over sample in that area, and then we would get -- as part of the major data collection, then we would get more information as part of the main framework.
But my own personal opinion at this point is that sample sizes on our surveys right now are too small to really get a baseline of the heterogeneity in the population. They need to be a little bit bigger.
We have to add areas to get some of the larger subgroups. So if you could get the base a little bit bigger so you have that to work from, then I think it is easier to design second level, third level add-on surveys, targeted surveys, that can respond to some of the specific needs of these smaller populations.
I think the main decision you want to make is, do you want these targeted surveys to be within the framework of the larger survey, or are you just going to fund a lot of targeted surveys that are very focused on the particular needs of that population, which is another way of doing it. You have more of an independent approach to that data collection. It is a data policy decision and how do you think you really want to approach it, and how are you going to get the major bang for your buck.
DR. BREEN: Jennifer, we have been thinking in terms of a long range plan and shorter term immediate planning. Do you have answers to those questions? I think our most immediate desire that we want gratified is better information on small populations for which we don't have adequate data to analyze at this point.
I think in the long run, I presume this group would agree that we would like better national surveillance that would allow us to analyze the smaller populations in the context of the national population and the state population, too, if that were possible.
So can you comment on how you might go about that?
DR. MADANS: You mean a short term solution?
DR. BREEN: Short term and long term. You were saying you could do this first or you could do that first. Yes, we want choices, but you have probably thought about this a lot.
DR. MADANS: We probably have a preference on how we would like to do it. Right now I can't tell you what the company line is, because we start down these paths, and all of them cost money.
I think the first thing you have to realize is, none of this is possible within the current budget of these surveys, unless you reduce the ability to make a national estimate irrespective of race and ethnicity considerably.
DR. BREEN: I don't think we want to give that up.
DR. MADANS: Okay, but if you say that you are not going to do that, then all of these cost money. I think if we understand that, then I think there are some options that we could include, start doing immediately.
Someone mentioned BRFSS. We have spent some money and some time thinking about doing dual frame with the HIS, which is a similar kind of thing, by increasing the sample size using a phone frame. Now, a lot of these groups don't have such great telephone coverage, there are problems there, but at least they have a household survey to back it up.
So I think there are some things that you could do that would at least give you some data, not for very small groups, but for some of the growing groups, where they are not very small and very focused.
DR. BREEN: When you say very small, --
DR. MADANS: Let me asked one targeted thing before I move on. One of them is looking at things over time. When you look at a particular geographic area, you have a population group that is very highly concentrated at one point in time, and you keep doing the data collection in that particular area. Then in five to ten years, people start moving, because they don't always stay in place, and you are still doing your data collection of Asians in Washington State. That population changes, and when you see changes over time, you don't know what it is due to.
This has happened to us before on vital statistics with a number of states collecting data. You never know whether you are seeing migration effects or people effects. I think you have to keep those things in mind in the long term, how you are going to interpret the data if you don't have that broader picture.
MR. HITCHCOCK: That is the national viewpoint. I think the local viewpoint is, still that information is very valuable to them. We may be wondering what is going on back here in D.C., but I think they have an understanding of what they are getting five years out compared to what they got recently.
DR. MADANS: The one thing they can't do is whether things are improving or declining. All they can say is that there is a composition change, which is not -- for service provision on their part, that is fine.
MR. HITCHCOCK: Yes.
DR. MADANS: But if you are looking at overall monitoring of health data or whether things get better or worse, you can't make those kinds of judgments.
DR. HUERTIN-ROBERTS: When you say you can't get small populations right away, are you talking about, you can't get a specific American Indian tribe, or you can't get Chinese or you can't get any Asian estimates?
DR. MADANS: I don't think it is ever possible to get specific tribes using a national survey. You have to go with the targeted. The larger Asian groups we could probably get.
DR. HUERTIN-ROBERTS: And we could get an estimate for American Indians other than other?
DR. MADANS: HIS puts out estimates of American Indians. You really need to combine three years of data. We do reports on that. There are lots of surveys that once they collect it -- if you look at CPS, the Current Population Survey, they collect according to the five categories, but they basically publish white, black and other. HIS actually publishes the five basic -- I think it is the four Asian subgroups and the Hispanic subgroups. We caution that some of those estimates are not very stable, and you really need to combine multiple years of data. But if you have to combine three years of data to triple the sample size, you can do it in one year. If you have to combine three years of data that is only 50 percent increase of sample size, then you would also get the data in one year.
Now, the more you want to cut it -- if you want to look at a very small age group with one sex in a certain region, or in an SES group, obviously you start having sample size problems. But if you want some basic information, you can get it from surveys of that magnitude if you increase them some, and then you can do the focused targeted groups surveys for groups where it would be very difficult to do it on a national level.
Just to clarify, the issue on the data center for HIS has been more around the Hawaiian groups. It is not a sample size issue for American Indians so much. It is just that when they broke the category of American Indian and Native Hawaiians, when in the past we had them combined, and so we have problems with consistency, because the Native Hawaiians is a group we can't put out. So you have a consistency problem, and you have to go to the data center to fix the consistency.
MR. HITCHCOCK: So Jennifer, you are basically telling us this is a very complicated, thorny area. You are not saying we shouldn't go ahead and make recommendations.
DR. MADANS: No, I think you should make the recommendations, and lay out that there are various options. I think it is fair to put in that these require not only methodologic work, but require additional funding. I don't know if I would suggest -- and again, my personal opinion would be that it may be counterproductive to focus so much on the targeted studies as opposed to trying to make that an expanded part of the basic national survey. I think you would get more if it was an integrated approach than if you just suggest independent targeted surveys in different geographic areas for specific subgroups.
MR. HITCHCOCK: You may be right. I think OMB might say you were right, too.
DR. BREEN: And OMB has an air of consistency that I think would fly better with the Administration, too.
DR. MAYS: Jennifer, I think that what might be useful to us as we pull these together is, we have been talking a little bit more about targeted surveys a little bit more independently, partially because we now have you here to also hear the other side, is for us to be able to benefit from -- again, if you all have thought this out, like it was very helpful for us to hear from Vickie and Marcie, et cetera, about the fact that they are ready to go if they actually have funding. So we can clearly express to the Secretary that there is the foundation upon which this could be done with efficiency, and not that it is actually starting from scratch.
So I think the more that we understand that, and if you have some specific recommendations for being able to build upon -- do you all call them flagships? Like NHIS and NHANES? You call them something. The major surveys that you have, if for example we could gain a tremendous amount of information by increasing it by some specific number, it might be useful for us to be able to say that.
So to the extent that you can share that with us, it might become something that the committee can consider recommending.
DR. MADANS: I think we have written something like that. I'll find it and send it to you. When we have written things like that, we also include such national data as geography. It helps a little as well, because if you have these larger samples, you have a higher chance of hitting these population groups than if you don't. So we try to balance the sub-national versus the race and ethnicity.
What we like to do is design the HIS sample -- and this is something that is up for grabs at this point -- to have geographic spread, so that if someone came to us and said we want to look at Asians in those three states, or a particular tribe in an area or an immigrant group that is only living in one area, we have frame there, and we can just add samples to it, rather than starting all over again. So it is all part of the integrated approach. Let me send you what I have on that.
DR. MAYS: I think that would be great. I agree with Dale; I think the Department -- if we make it clear that a majority of this work is there, but that adding this on a periodic basis could really bring --
DR. BREEN: Enhance.
DR. MAYS: Yes, could really enhance the Department's efforts. I think this would fly a lot better. So anything that you can share would be great. Hi, Susan.
DR. QUEEN: Hi. I just wanted to add that back in February at the hearings that AHRQ did, AHRQ proposed the idea of two separate and parallel tracks of data collection, to get the large coordinated collection that gets the large population, and then in parallel a separate track because we do occasional, separate studies.
DR. MAYS: I was just asking whether we were going to have Reynard. Can you call his office and see if Reynard can call in at all today to share this with us? There was a possibility he was going to try. Oh, Audrey already has, she used the cell phone.
DR. GREENBERG: Do we need to take a break?
DR. MAYS: Yes, I was just going to wrap up.
DR. GREENBERG: Also, the cafeteria closes at three, so if anyone wants to get tea or coffee.
DR. MAYS: Okay. We are going to take about a five-minute break here, so people can run and get things and come back, then we will be back.
(Brief recess.)
DR. BREEN: Susan, are you on the phone?
DR. QUEEN: Yes.
DR. BREEN: Susan had given us a proposal for special population surveys. I think this morning, we started talking about doing independent surveys, and that would be our recommendation. After we talked to people around the table and people at NCHS and AHRQ and whatnot, it seemed that it would probably be more efficient, more feasible and all-around better to do them in the context of the existing surveys. Is that right?
DR. MAYS: I think we want to do both, but I think we know that within the existing surveys, we would probably get more mileage.
DR. QUEEN: My impression is that the existing surveys would never be able to completely get some of the smaller groups. You would still have a need along the way with even the screening issues, the cost issues. You sort of need both.
DR. BREEN: Susan, what we heard late this morning was that NHANES in particular and SLAITS is another example, have done surveys in targeted areas where they have been requested to do them, like New York City or like on the Navaho Indian Reservation, that sort of thing. I think that is very consistent with what you are proposing, isn't it?
DR. QUEEN: Yes.
MR. HITCHCOCK: No, it is not, because the HANES didn't really do those surveys. They didn't really do much of anything with Navaho HANES, other than maybe provide equipment. In New York basically they are providing everybody in technology and guidance expertise, but it is New York's survey.
I think what we are talking about here is doing -- a New York approach would be to do a regular HANES in New York, but then maybe stay around for what they call stands, a couple more sampling locations within New York, and then combine the data collected from those special focus stands with add-on stands, with that collected while they were in New York for the regular HANES, and out of that get a sample that would cover New York better than the regular survey would.
DR. QUEEN: Did Kathryn Porter mention the concept of the smaller mobile --
MR. HITCHCOCK: Yes.
DR. LUCAS: That is what they would use. But it is part of HANES. The way they are doing it now is not the way we intended. The community HANES's are still using HANES protocols, so that you can compare what you get in that area with the national HANES. So they are not independent surveys in that sense.
MR. HITCHCOCK: They haven't been done, either.
DR. LUCAS: One idea for New York and also D.C. when they were here is that when the basic HANES is in D.C. we then send out these little media trailers throughout D.C. and increase the sample, so that you get a HANES stand in D.C. that is representative of D.C. So that is one way of doing it.
Another way of doing it is to send out the little RV part of HANES to someplace that HANES isn't even in, and do a HANES light. You don't do all of HANES, you do part of HANES.
MR. HITCHCOCK: That is not what I was talking about. The example I used would be just staying in D.C. longer, basically, and accumulating more sample size.
DR. LUCAS: We never were going to do that. Not recently, anyway.
MR. HITCHCOCK: You could.
DR. LUCAS: You could, but they are part of the basic framework of the larger survey, even if they are not part of the sample of the larger survey. They are add on. They are not totally independent surveys that are going on, and they can't relate back.
DR. QUEEN: Are we necessarily saying that these special studies could not be part of the main project in a periodic way?
DR. LUCAS: No, but I think you design them differently. If you talk about it completely parallel or separate, two tracks, the people reading this may not realize that you may want them related in a way so that you have comprehensive data collection. If you want to do your report, you couldn't look at the regular HIS and an HIS-like interview survey that you did in an Indian reservation.
DR. BREEN: And you can compare them.
DR. LUCAS: Right.
DR. BREEN: Exactly.
DR. HUERTIN-ROBERTS: I was just going to point out that we have been talking about this dual track that Reynard had suggested of doing the national studies and then doing separate smaller targeted studies.
I think that trying to focus on one or the other exclusively is not going to get us where we need to go. I think we need both, because both approaches have strengths, but they also have limitations. So I think unless we try both, we are not going to get very far.
DR. BREEN: I think it is fine to have the parallel approach, or I would agree with the parallel approach, while keeping in mind what Russ was emphasizing before, that we want to have these things collected in a way where we can at the end ot the day have something we can compare with the national surveys that we have got. We know what the sampling frame is. We don't have these questions like we had for years with the BERFIS.
DR. MAYS: What were the questions with the BERFIS?
DR. BREEN: Well, the sampling frames were different in all the state states. The questions could be different in the different states. There were so many differences that nobody knew for a long time how to bring them together into a single something that looked like a national estimate.
MR. HITCHCOCK: The response rates were wildly different.
DR. BREEN: Yes.
DR. MAYS: What I want to do is visit the overarching recommendations that we initially have. I think that they do fit into what we are talking about.
At this point I'm not asking about any wordsmithing as much as, I just want to make sure if there are any one of these that you want to get rid of. But otherwise, what we will do is work on integrating them into the things that we have said, because they do pretty much fit.
So for those of you who have the report in front of you, if you go to -- it starts where it says overarching recommendations, page seven.
HHS should implement a multi-approach strategy based on documented needs and the results from methodological research for developing data on racial and ethnic minority populations and subpopulations, including collection of vital statistics, over sampling in national surveys where feasible, conducting followup and dual frame type sampling approaches for special surveys, sponsor targeted studies for developing data on populations where over sampling is not feasible or cost effective. Appropriate resources need to be made available to accomplish these goals.
So the gist of that we are going to work with. We have made some changes, but I just want to make sure that the gist of it as we work on this letter is okay.
The next one. The Department should assume a leadership role with other Departments -- and this is where Commerce, Labor and Education were the ones that after discussion we suggested, and the Bureau of Census and private and academic research centers, to undertake a program of methodological research into issues associated with the measurements of race and ethnicity in survey census and research. We have broadened what they should do the methodological work on, so we'll make sure that we capture that.
Next, which Nancy may have to help me with, agencies need to develop and publish crosswalks to bridge racial and ethnic data collected and published under the new OMB standards with these collected under the old standards. Obviously, the crosswalk is to yield data that is in accordance with OMB tabulation guidelines.
Now, I don't know if we want to address that in this letter. I would suggest that we drop this one right now as overarching, because it really takes us in a different direction.
Next is, whenever possible, agencies are encouraged to combine several years of data to achieve sufficient analytic sample size, something something, standard set of questions over several years so the data can be combined to an adequate sample size. We have something something; there is some editing here that I wasn't sure where it went. But this is the notion that we have been talking about, that it is the encouragement -- that one of the ways to be able to get information is by combining across years.
Is that major enough that we want to put that into this letter?
DR. GREENBERG: No.
DR. MAYS: Thank you. The Department is urged to develop consistent strategies and mechanisms for the dissemination of data about racial and ethnic minorities that result from the federal data collection efforts, when in particular data is targeted at specific racial and ethnic groups. Is that sufficient enough that we want to put it in?
DR. HUERTIN-ROBERTS: Yes, I think so.
DR. MAYS: This is about dissemination. Basically what it is a push for is to say that there needs to be -- they want to see an ongoing mechanism available to assure that there is a dissemination of data specific about racial and ethnic minorities.
DR. GREENBERG: You want the information collected and you want it disseminated.
DR. MAYS: It may not need to be a recommendation as much as it needs to be maybe discussed.
DR. JACKSON: I see there is a part A that is a little bit before that, and this is the followup to that.
MR. HITCHCOCK: Community involvement.
DR. MAYS: We will talk about it when we talk about -- that's good, community involvement. So rather than this as a recommendation, it will come under the six. That is part of what six is really about, is insuring that the data gets out to the communities as well as in some kind of way that it is public and available. Great.
I don't think we want the next one. The Department should fund a National Academy of Sciences study to examine its level -- I didn't even finish it and everybody is hollering no -- its level and methods of funding of its data collection and Data Council efforts, and whether the structure and level of funding are resulting in a decrease of data and lack of efficiency due to outdated technology and methods.
We'll just say no to that one for right now. Just say no.
So those were the overarching recommendations, so we see which of those that we want to pull. Let me now read you the one from the NCVHS recommendation for achieving the health statistics visions for the 21st century. Which of these is the one that you wanted?
DR. BUTLER: I checked some that I thought were appropriate.
DR. MAYS: Sorry, I thought it was on the front page.
DR. BURWELL: It is on the front page, the one we read earlier.
DR. MAYS: But this is underlined.
DR. BURWELL: Yes, it is the first one checked saying, develop mechanisms to rapidly modify the samples, data collection strategies.
DR. MAYS: You're on the next page. What is on the first page? There is an underline on the first page. That is what I didn't understand, what you want on the first page.
DR. BURWELL: No, there is nothing there.
DR. MAYS: Okay. Develop mechanisms to rapidly modify the samples, data collection strategies and data collection instruments of ongoing data collection systems in response to emerging issues. Is that something that we want to -- are we agreeing with that philosophy? Personally I think that is hard.
DR. BURWELL: It goes under your flexible approaches.
DR. MAYS: Oh, okay. Jennifer, are you still there?
DR. GREENBERG: And it is a recommendation that the committee has already made.
DR. MAYS: Jennifer, are you still there?
DR. MADANS: Yes.
DR. MAYS: I want you to listen to this one. I just want to hear if you have a comment about it.
DR. MADANS: Okay.
DR. MAYS: Develop mechanisms to rapidly modify the samples, data collection strategies and data collection instruments of ongoing data collection systems in response to emerging issues. Can you do that?
MR. HITCHCOCK: SLAITS.
DR. MAYS: Oh, SLAITS is an example.
DR. MADANS: Yes, you can do it.
DR. MAYS: No, you.
DR. MADANS: I think we do need -- especially on the sample design need to develop ways to do that.
DR. MAYS: Okay. So then we are fine about language that captures that.
Prepare and provide staff of state and federal agencies with health statistics responsibilities, a manual and a toolkit of data sets for use in community level analysis.
DR. HUERTIN-ROBERTS: Wouldn't this be under the best practices?
MR. HITCHCOCK: A lot of these -- the basic list of recommendations reinforce them, basically.
DR. MAYS: Develop person-based longitudinal data sets and surveys in order to develop portraits of influences on the population's health throughout the life cycle. What is that person thing?
MR. HITCHCOCK: Household means instead of a provider based survey, rather than office visits for a provider, you look at number of office visits for a person.
DR. MAYS: I think this takes us a little afield from where we are going for that one.
MR. LOCALIO: It is terribly expensive. Can somebody help me on this?
DR. BREEN: It is terribly expensive.
MR. LOCALIO: This is so much money, nobody --
DR. MAYS: Wait, wait, wait. There is one thing that we didn't talk about. That is, NIH is embarking on the national longitudinal child survey, national childhood survey.
DR. BREEN: And it is terrifically expensive.
MR. HITCHCOCK: And there is no funding to speak of.
DR. BREEN: It has taken them what, five years now to get off the ground?
DR. MAYS: Yes, but it is going forth, I am told. It is going forth. So we may want to make a recommendation -- remember, this came up on the telephone -- that says that we want to insure that that survey takes into account the recommendations we have made. That might be one of our final things to say, so we'll just say that.
MR. LOCALIO: Marjorie, what are the jurisdictional issues here on this? We are making a recommendation on a -- what Institute is doing this?
DR. GREENBERG: Child Health and Human Development.
DR. BREEN: Different Institutes within NIH.
MR. LOCALIO: So is that within our --
DR. BREEN: Yes, because it is being done under contract, I'm sure. It is not a grant, is it?
DR. MAYS: Aren't you giving out little grants to do some things, but the survey itself is under contract?
DR. BREEN: Right. You would love a recommendation like this, wouldn't you?
DR. MAYS: So we will remember to put that one in about -- but longitudinal surveys are very expensive.
DR. BREEN: But since it is happening anyway.
DR. MAYS: This is one for Russell. Adopt or if necessary develop standards for data elements commonly used in all methods of data collection for electronically transmitting data, for presenting and disseminating data -- I'm totally lost on this.
DR. GREENBERG: This is related to the standards in the best practices.
MR. LOCALIO: This is too specific.
DR. BREEN: I don't know that we have to tell people how to present and disseminate their data.
MR. LOCALIO: We are talking about similar questions. What we have been talking about has a larger --
DR. MAYS: Let's skip to recommendations on the research agenda and see if any of these -- we can always refer them to our report. We can always refer to our report. Develop and fund a research agenda for the study of methods and technologies to enhance the privacy, confidentiality and security of health statistics.
DR. HUERTIN-ROBERTS: That is part of best practices.
DR. MAYS: So this is best practices. We will refer them to the 21st Century for a variety of things we think are important in best practices. What about developing a research agenda to explore new data collection strategies? That is best practices, right?
DR. GREENBERG: No, that is methods.
DR. BURWELL: Yes, that is new methodological work.
DR. GREENBERG: But in addition to funding the surveys, you have got to fund some of the research for it.
DR. MAYS: Oh, this is the methodological stuff that we were talking about, so this goes there. Develop and fund research agenda to explore new data collection strategies that can rapidly and flexibly -- yes, this goes with this -- provide data on emerging issues on a population's health, assess the validity and reliability of items used in key ongoing data collection systems, and estimate any loss in accuracy from early publication.
DR. GREENBERG: No, this is more specific to that issue.
DR. MAYS: I think the first part we have the essence of. Our committee suggested this to the rest of the world.
DR. GREENBERG: This is part of the 21st Century report.
DR. MAYS: Yes, this is 21st Century, where the budget is different. 21st Century, the vision and the budget are totally different directions.
Provide control and access to potentially identify individual level data on a population's health and influences on a population's health. I think this goes to the community stuff.
MR. HITCHCOCK: The data centers.
DR. MAYS: Yes, the data centers.
DR. BREEN: Did you mean to skip over, develop methods to validly and reliably estimate important indicators?
DR. MAYS: I'm sorry, I didn't realize that did. Develop methods to validly and reliably estimate important indicators of the health and of the influences on the health of the state and local populations. So what about that one?
DR. BREEN: Don't we want to collect data to do that, rather than to develop methods to do that? I'm sure they are married, but --
DR. MAYS: I think we might want to say something about making sure that the data that is collected has some usability at different levels.
MR. HITCHCOCK: I think they are talking there about things like quality of life, something that we don't have agreed-upon measures.
DR. MAYS: Okay. I think we should leave that one alone. So now we know what we will take from the 21st Century.
MR. LOCALIO: The proposal then is to take some of these recommendations and re-issue them?
DR. MAYS: No, the proposal is making sure that in those sections where we have talked about these issues, that we can also refer to them, and if the language can be taken from here. But we also will refer people to this.
MR. LOCALIO: Okay, thank you.
DR. MAYS: Let's do Susan's. I think Susan has done a great job. I think to some extent we may just fold this in.
DR. HUERTIN-ROBERTS: Do we need to take the word surveys out?
DR. BREEN: All of this does refer to surveys. We might want to add something about vital statistics or something else, administrative data.
DR. MAYS: We might want to make this just a special on surveys itself. This is just federal, because what she has done is identify HIS, HANES and MEPS.
Susan, are you there?
DR. QUEEN: Yes.
DR. MAYS: I just wanted to make sure.
DR. BREEN: But she says, despite them, there is a substantial lack of data on most smaller population groups, and that they don't suffice.
DR. MAYS: I like this. As a result, there may be limitation on the various federal data systems providing data on the health disparities in racial and ethnic groups.
Let me do it the opposite way, which I think is a little faster. Is there anything in here that you have a problem with? Otherwise, what we are going to do is use it as the basis upon which to write this section. So let's do it the opposite way. Is there anything that you think needs to be pulled out? What we will try and do is to work with the difference between the broader population-based data and then to make sure that we also address population-based surveys, which is what I think she has done.
What is in here are the names. Have we settled upon how we want to describe this? We have gone from special population, we have had defined population, we have a community HANES, and we have targeted population. What should we use consistently throughout here?
MR. LOCALIO: I like special populations.
DR. HUERTIN-ROBERTS: My only concern with special populations -- my first reaction was to go with special populations. My only concern was, special populations in the context of government really does mean certain ethnic populations, and doesn't necessarily -- wouldn't necessarily subsume things like geographically focused stuff. It would make sense to me, but it calls to mind specifically an ethnic or racial focus, I think.
DR. BURWELL: And when you look in Office of Special Populations in NIH, they don't always focus on racial and ethnic groups.
DR. HUERTIN-ROBERTS: I know.
DR. BREEN: I think it is completely open-ended. I also have this historical remnant in my mind. Before we had health disparities, we had special populations in the federal government. But it also is very broad, and it doesn't limit us.
DR. HUERTIN-ROBERTS: I guess my concern was, and Dale, we talked about that on the last phone call, when we were talking about what we should call them. We didn't want to eliminate the idea of geographically limited or defined studies, community-based or regional or whatever.
DR. COOPER: (Comments off mike.)
DR. MAYS: Let me just ask, and I am probably being a little too political here, but in this day and age of some of the issues we are dealing with such as affirmative action and stuff like that, is this going to create a problem? This is the federal government, so does the federal government still do this?
DR. BREEN: The name?
DR. MAYS: Is it going to make somebody vote against it or not give it money?
DR. BURWELL: I just think so.
DR. MAYS: I just don't know.
DR. BREEN: There is another term in here, special studies. That doesn't designate population, it just says special studies.
DR. MAYS: How about special targeted studies?
DR. QUEEN: Sounds good.
DR. MAYS: We did it, we did it.
DR. BREEN: I like this a lot. I have one suggestion. That is, on the third paragraph down, it says the use of defined targeted projects in parallel with large national studies would provide two systems and tracks of data collection that would together provide a more complex picture of the overall population. I would add, complete, because I think that is the issue. It is incomplete without these population groups.
DR. QUEEN: I think I actually meant that there was to be a complete picture of the overall complex population or something like that, at this point.
DR. HUERTIN-ROBERTS: Should we actually say two systems, or should we say dual track or something like that? Two systems sounds like xerox.
DR. MAYS: And they will come looking for scale of economy. So we'll call it dual track?
DR. HUERTIN-ROBERTS: Dual track, yes.
DR. MADANS: I would still strongly recommend that you say something about that those two tracks have to be related to each other some way. Otherwise, you can have two very short systems that don't mesh enough for you to say anything comprehensive about the whole nation and related subgroups.
I tried to do that. I guess I didn't elaborate enough, where I was mentioning that they would be collected in a similar way to allow comparisons. But we could certainly elaborate.
MR. HITCHCOCK: I think the word parallel is a little troublesome. I know that is the word that Reynard used, but it is almost --
DR. MADANS: I wasn't there when he spoke, but I would be surprised if he really meant them to be parallel. I've got a lot of background. If you need anything, anyone, just give me a call and I'll try to send you that stuff. Bye-bye.
DR. MAYS: Thank you, Jennifer. So we'll stop using the word parallel, because that may freak people out a little bit. Don't you want to have a turn about something you are supposed to be doing in terms of integrated something data something or other? Isn't there some committee about surveys?
DR. GREENBERG: There was an integrated survey work group of the Data Council, which evolved into that other data strategy group.
DR. HUERTIN-ROBERTS: How about integrated dual track system?
DR. GREENBERG: I think you want to use the word integrated.
DR. MAYS: Yes, somewhere, because it even fits back into what you all have -- yes, we'll take out track and call it dual frame.
MR. HITCHCOCK: There is only one example. It covers most of them, I guess. It covers telephone, in-person, it covers regular HANES, then with the little trailers going out.
DR. BREEN: When I hear frame in this context, I think of sample frame. Is that what is meant? I'm not sure we want to use it if that is what is meant.
MR. HITCHCOCK: It could be.
DR. MAYS: It could be.
MR. HITCHCOCK: If you've got a sample of households and a sample of phone.
DR. BREEN: It wouldn't necessarily be.
MR. HITCHCOCK: It wouldn't have to necessarily be, right.
DR. MAYS: Yes, I see, in the wrong hands it is taken as a statistical thing, and that is not -- we wouldn't want them to act on it like that. Okay, we'll go back to dual track.
DR. GREENBERG: But you are talking about adjusting or whatever the telephone data with data you have from households.
DR. BREEN: How about just calling it an integrated data collection system? To say the use of defined targeted projects collected in a similar way to provide comparable data with the large national studies would provide an integrated data collection system.
DR. MAYS: We'll get them from you. Susan, I like that you go to what the HHS strategic goals and objectives are. That is really good. She has HRSA in here.
DR. QUEEN: Yes, I know.
DR. MAYS: No, that's fine, the examples are good. That is exactly what I wanted, is to make sure we tied it back into the Institutes and agencies, so that they can see that we are not coming at them with something that totally is different, but instead that it will help to accomplish that. So if this fits with others, you should also let us know.
DR. BREEN: Excuse me, though. I'm happy to have HRSA here, but doesn't the Department have a stated goal of assuring quality health care to low income, uninsured, blah, blah, blah?
MR. HITCHCOCK: Sure.
DR. MAYS: Oh, you mean it is the overall goal of the Department?
DR. BREEN: Yes.
DR. MAYS: Oh, okay. Let's find where it is.
DR. BREEN: In the next paragraph it specifically refers to HRSA needing the data to identify and address unmet service needs, so maybe that sentence could -- the thing about special targeted studies would be useful to HRSA, could be put there.
DR. QUEEN: Everything that appears after the actual HHS objectives is the HRSA strategic plan. Eliminating barriers to health care, that is part of HRSA's stated mission.
DR. MAYS: So Nancy, are you saying that before special targeted studies, blah, blah, blah, HRSA, there should be something about the Department?
DR. BREEN: No. What I am thinking, and Susan, correct me if I'm wrong here, but Dale seems to agree, we have in the Healthy People 2000 goals and also in the strategic plan, the Department has a stated goal to provide quality health care, eliminate health disparities and to have everybody insured, and everybody have an equal source of care, that sort of thing.
DR. QUEEN: What you might want to get rid of then would be the last paragraph, which is very much process specific.
DR. BREEN: I was suggesting putting the, special targeted studies would be useful to HRSA, which is that first sentence in the second-to-last paragraph before the bullets, to move it down to the last -- I guess that whole paragraph is about HRSA, so maybe just leave it. Maybe what you said, Vickie, would be better, about the overall Departmental goal, and then say HRSA as an example.
MR. HITCHCOCK: Where is this material going that we are agonizing over and wordsmithing here? Where is this particular wording going? Is it going in the letter?
DR. BREEN: Yes.
MR. HITCHCOCK: This is our three-page letter, and we're two pages in.
DR. MAYS: I think what we need to do is, here is what I'm going to suggest. Let's try with this letter, and if the letter is going to be too long at this point, then to go and decide what to cut out of it. I have a feeling that there are things in here that might be useful in the letter.
MR. HITCHCOCK: Oh, yes.
DR. MAYS: Then we'll cut it. The letter might start out five or six pages, and then we'll cut it back.
DR. BREEN: Also, Susan has already talked about it, and the lead-in is fine, so ignore what I was saying.
DR. MAYS: Okay. Our letter is probably six pages, maybe seven, but I think that when it all gets put together we will edit it down so that it is three pages. It is almost like when we are doing questionnaires. You've got to have a compelling reason to make it any longer. You've got to come up with a big argument as to why this paragraph must make it over three pages.
So if you've got something that we start cutting, come to the table with a lot of rationale as to why it has to stay, because otherwise we are going to strive for three pages.
MR. HITCHCOCK: A good example is like, if you can't have your five questions, okay, I'll ask one question with five parts.
DR. HUERTIN-ROBERTS: Our section six is for the letter?
DR. MAYS: Yes.
DR. HUERTIN-ROBERTS: So it needs to be married with --
DR. MAYS: It needs to carry with it authority, but not necessarily details, like why would you want to do this. It is compelling for you to do it, but read the report to learn more.
Dale, can you go to the outline? I just want to see if there is anything else in there that we should talk about real quick.
DR. GREENBERG: The outline of the report? Oh, the letter outline.
DR. BREEN: Yes, the letter outline.
DR. MAYS: Can I just ask a question especially of Nancy? We wanted to comment -- this came up a little bit on the telephone call -- about the American Community Survey. Is this the place to do this?
DR. BREEN: In the letter?
DR. MAYS: Yes.
DR. BREEN: It seemed to me that the American Community Survey is an option for getting data on small populations, because it is going to replace -- it is intended to replace the long form for the Census. So it is gathering data on -- it literally has a sample size of 15 million, or something like that.
So it is likely to pick up lots and lots of the population groups we want with all the features we want. So for that reason, I think it is useful. The downside is, it doesn't collect any health data, but it could.
DR. MAYS: That is what somebody said.
MR. HITCHCOCK: (Comments off mike.)
DR. BREEN: Then yes, definitely. We actually had a meeting, and Marcie Cynamon had said earlier that NCI was meeting with them about SLAITS. NCI also met with ACS about that same survey, because they are looking for a venue for it. They found out that the American Community Survey were very open to it. The problem was that they wanted to do a followup.
This committee should probably know that right now, followup surveys are problematic because the Census can't do it, because they have been banned to do followup surveys, because the Census can't release private data to private firms.
DR. MAYS: Is this HIPAA?
DR. BREEN: No.
DR. MAYS: It is just the IRB?
DR. BREEN: Confidential data to private firms. So private firms have gone to the Administration and said, if we can't do this followup surveys, we don't want Census to do them, either. So right now, no followup surveys are being done. But anyway, that is the current situation.
DR. MAYS: If I can't have it, you can't, either. That is silly.
DR. BREEN: I don't know if we want followup surveys. We might down the line, but initially we don't need that.
DR. MAYS: Do we need to say anything about this in this letter? It sounds like that we shouldn't.
DR. BREEN: Susan did use examples, so maybe we should consider --
DR. MAYS: No, there is another place where we also talked about -- we give examples.
DR. BREEN: I would put it in.
DR. MAYS: If it fits where we can do an example, we will put it in. Otherwise we are not doing anything else about it.
DR. BURWELL: The only thing I was going to say is, we gave Departmental examples. I'm not sure --
DR. BREEN: I think we would say, or the Census American Community Survey might be enhanced to include health questions, or something like that. Make it clear that it is outside our current realm, it doesn't have what we need, but the frame is great.
DR. MAYS: All right, so that is all we need to say about that.
DR. BREEN: And the background info is great.
Agenda Item: Discussion on Remaining Agenda Items
DR. MAYS: Can we talk about the process for getting this letter done? Some people have assignments in terms of pieces they are going to write. Here is what I would suggest. Let's get this first draft out with the assignments that we have.
(Discussion off the record regarding future meeting arrangements.)
DR. MAYS: Those of you who don't have an assignment, I don't want you to feel left out. We will get you one the next go-round. I realize that we also have a health plan letter, which I am going to talk to the people from AHRQ, and I will also talk to Bob. I think Bob Hungate should also be involved in this, so I will ask him to participate.
Can we take here on the agenda the last things? The timetable. Now, the one thing you notice we haven't talked very much about now is the report, because we have gone from one letter to two letters. Personally, I want to make sure we get these two letters out. If we have to push the report to November -- Marjorie, do you think that is okay, if we send the Secretary a letter and then we say come November something is coming?
DR. GREENBERG: Yes. The only thing I would say is, if you want to make sure it is November, you might want to bring --
DR. MAYS: A draft in September.
DR. GREENBERG: -- a draft to the full committee.
DR. MAYS: Right, because now I remember, November is very busy, isn't it?
DR. GREENBERG: Yes. We've got the Pacific Islander meeting.
DR. MAYS: The meeting agenda is very full in November.
DR. GREENBERG: It seems to me September was very busy.
DR. MAYS: No, because what we decided was that in September we wouldn't have outside people coming, and that we were going to have more outside people for November. Then September we have so many things we were doing that we were going to try to keep September --
DR. GREENBERG: That's true. We have the ICD, and --
DR. JACKSON: And Lumpkin is available -- is that September? We were trying to structure the agenda so that the discussion would actually occur in the first day, and the breakout and all that in the second day.
DR. MAYS: Here is what I am going to suggest, because some of the report may be on my back, whether I can sit down and try and -- because I think we do have a set of comments, whether I can sit down and try and reorganize it. If I can do that with these two letters going, then we might be able to try for September. I just have to get through APA, which is the beginning of August, and then I can clear my calendar and concentrate. By the second call in August, we will know.
DR. JACKSON: Do you know how much time you will need on the full agenda in September?
DR. MAYS: Yes, exactly, that is what I was saying. By that conference call, we will know, as to whether we can do it or not, yes.
We need to talk about the American Indian, Native Hawaiian, Alaska Native report, as to whether we can do it, and if so, the timetable we can do it. So let me put it on the table and tell you what it is about.
The person that we would have do this report -- no, let me back up, because some of you don't even know what the report is about, so let me start from the beginning.
This report would be based on what we have heard in the hearings specific to two issues, the collection of data on race and ethnicity in these populations, and also the issue of disparities, specific to whether or not we have the data to be able to determine disparities, and if not, what some of the issues are in being able to get that data.
Now, the reason these are grouped together, the rationale is that these are indigenous populations. So that raises a whole other set of different issues than if we are talking about say Latinos or African-Americans.
DR. GREENBERG: What about Asians?
DR. MAYS: That is why I stopped. I think that it is just these populations we want to deal with, because we haven't finished our hearings for the Asians.
DR. GREENBERG: It is the Native American, --
DR. MAYS: The Native Hawaiian, --
DR. GREENBERG: -- the Native Hawaiian and the Alaska Natives.
DR. MAYS: Alaska Natives.
DR. GREENBERG: So it is not going to be Asians.
DR. MAYS: I was just going to say, it is Native Hawaiians that are in here when you are talking about indigenous. We haven't finished the hearings anyway for the other groups, but the rationale here is that this is the indigenous group. So there are certain issues that get very complicated in terms of the indigenous groups.
DR. BREEN: So the indigenous groups are American Indians, Alaska Natives, Native Hawaiians and other Pacific Islanders?
DR. MAYS: No, the Pacific Islanders are not. They are not indigenous.
DR. BREEN: To the Pacific Islands they are. We just happen to own them.
DR. HUERTIN-ROBERTS: Some of the populations in the Pacific Islands are indigenous.
DR. MAYS: Like which ones would we be talking about?
DR. GREENBERG: The territories.
DR. HUERTIN-ROBERTS: Guam, Micronesia, Marshall Islands.
DR. MAYS: Marshall Islands.
DR. HUERTIN-ROBERTS: American Samoa.
MR. HITCHCOCK: Commonwealth of Northern Marianas.
MR. LOCALIO: I think we are talking mostly about Guam. The largest island is Guam.
DR. BREEN: They are indigenous to Guam. It is just that we have half the island with the base on it.
DR. MAYS: The Marshall islands, we will consider those. See, we are getting beyond my expertise, so I am willing to show my ignorance. That is the overarching issue, and asking whether or not --
(Simultaneous discussion.)
DR. MAYS: That is why the Native Hawaiians are here. It is kind of the umbrella of why it is American Indians, Native Hawaiians and Alaska Natives, as opposed to American Indians and Alaskan Natives only, because the perspective is to put the indigenous populations in, and to talk about some of the things that impact indigenous populations, in terms of their relationships are different, what it means about the data collection, who is responsible.
Apparently this is a different approach to dealing with the collection of data to some extent in the indigenous populations, because of their relationship to the federal government.
MR. HITCHCOCK: We haven't heard anything from the islands.
DR. GREENBERG: No, other than what we heard in the report from a couple of years ago, which you could -- since you are trying to bring forward past recommendations, you could do that. I don't think that we should -- I would not recommend expending the resources at this point to start trying to --
DR. BREEN: Update that.
DR. GREENBERG: Yes.
DR. MAYS: Yes, we have some of that in the hearing, and then it would be a continuation of that. So maybe you would say that that report would go for --
MR. HITCHCOCK: The infrastructure was what they need.
DR. HUERTIN-ROBERTS: But infrastructure is important in this group, too.
DR. MAYS: I know you have to go. I want to really thank you for your contributions. I want to thank you for the issues that you raised.
MR. LOCALIO: No problem.
DR. MAYS: I really appreciate your being here.
MR. LOCALIO: Thank you, I'm glad I could come. To whom should we e-mail our assignments?
DR. MAYS: Let's e-mail them to Dale.
MR. LOCALIO: Fine.
DR. MAYS: Yes, e-mail them to Dale, and we'll try and pull them together. Thank you very much.
MR. LOCALIO: I'll talk to you on the fifth.
DR. MAYS: Of August.
MR. LOCALIO: Yes. Thank you.
MR. HITCHCOCK: The sixth.
DR. MAYS: Yes, the sixth. So that is an issue. I think we have to decide on that, because that is part of the perspective that apparently is a significant perspective, as it is being raised relative to some of these issues. Maybe we don't have to have -- we are going to have the hearing in November, and I think that hearing from those individuals in November could help us to set the date that we want to have this report finished. Maybe that is the way to think about it.
DR. BREEN: I was just going to say that I think that another thing that the indigenous populations have in common, and again, it is a little like the Asians, in the sense that you can't generalize about every single tribe or every single island in this group. But there is a tendency toward lack of resources both among individuals and communities in these groups.
I think that if we think back to the earlier part of our discussion when we were talking about why we would collect data, and we said one reason would be that the population was fairly large, and another would be that the population was in need, these populations I think will be characterized by big-time need.
DR. PAISANO: On the indigenous issue, I think a lot of it is politics, because American Indians and Alaskan Natives tribal governments have treaties, and a policy for precedent on government to government relationships. This does not apply to Native Hawaiians at this time.
DR. GREENBERG: It does or it doesn't?
DR. PAISANO: It does not, or to any of the Pacific Islanders.
DR. BREEN: But they are territories.
MR. HITCHCOCK: Yes, there are three territories and there are three freely associated states. The freely associated states are much like tribes. Palau for instance is a freely associated state and is not a territory of the U.S. Its governor is the equivalent of Bush, for instance.
DR. BREEN: And by need, I wasn't meaning so much health need, because I think we don't have the data to determine that for every case, but I was thinking more socioeconomic need, in terms of educational attainment and income.
DR. PAISANO: Yes, I'm just saying the politics of the issue.
DR. MAYS: Let me just then ask, because that is part of this proposal for this, is to deal with these indigenous issues, because of their relationship to the collection of data in particular. Am I hearing that it doesn't seem to fit? Or am I hearing that it is just different?
DR. BREEN: I would look at them as two pretty separate groups. The American Indians are different from Native Hawaiians and other Pacific Islanders. But I would put them in the same report, in two sections. That is the way I would do it.
DR. PAISANO: Yes, and I don't think that would be a problem, because if you look at a lot of the federal legislation, it usually only goes to American Indians and Alaskan Natives based on the sovereignty issue, and in some cases includes Native Hawaiians, depending on the legislation. It doesn't always include the other Pacific Islander groups.
DR. MAYS: Maybe the other Pacific Islander groups, maybe we don't want to put them into this report in some way. Again, the issues I think of territories -- also, it is the third part, which are then the territories. We have some territories stuff, and we may be better with taking the territory stuff and putting it back with the other territory stuff, as opposed to trying to put it in this report. Then I think we are getting three concepts, and I think that is beginning to be --
DR. GREENBERG: Cumbersome.
DR. MAYS: Yes. There are some places where the Native Hawaiians have some things that are like American Indians and Alaskan Natives, and then there are ways in which they don't. But the concept of being indigenous is there for all of them. Then we have special issues.
DR. HUERTIN-ROBERTS: When we are talking about Native Hawaiians, we are talking about indigenous populations within the United States. So that is one reason to include them. That is one way to bound this report.
DR. MAYS: That may be a cutoff.
DR. BREEN: Maybe that is the way we want to do it. The hearing that we are planning though for San Francisco before the APHA is to focus on other Pacific Islanders. We did hear from the Hawaiians, and we decided that they were important enough to hold another hearing. So do we want to leave them out of our report?
DR. MAYS: Leave them out of which report?
DR. BREEN: I guess we are calling it the indigenous people report, or the second report.
DR. MAYS: Let's not get in trouble now. We're not calling it the indigenous people report.
DR. BREEN: Everywhere else in the world, that is what people who came first, first nations, are called.
DR. MAYS: First nations is better. Indigenous people they don't like. First nations, they like.
MR. HITCHCOCK: Are you saying, what do we do about the other groups that we have heard from and that we are going to hear from, besides Native Hawaiians?
DR. BREEN: Yes.
DR. GREENBERG: The last hearing you have in L.A.
DR. MAYS: Is on Asians.
DR. GREENBERG: That is not going into the report that Olivia is working on.
DR. BREEN: No.
DR. GREENBERG: No, okay. So where is that going to be?
DR. MAYS: I'm sorry. We haven't decided where that information would be. We are using information we have gotten so far, but we haven't decided.
DR. BREEN: You know where logically I think they go best? I don't know if it makes sense in terms of the mandate of this group, but I think Asians and Latinos actually are most similar in terms of -- because they are the immigrants. They are the ones with large proportions of immigrants in this country, with the language barriers and cultural barriers, with access to care barriers based on those things.
DR. MAYS: We talked about doing the language stuff, and we heard about language relative to surveys. But a lot of the language stuff we haven't really done that. We quickly wrote the SCHIP thing about the language.
DR. GREENBERG: So some of what you heard at the Asian hearing will probably go into this report, this first report?
DR. MAYS: I think it is really driving this notion about these targeted surveys, because there is the group that really has almost next to nothing.
DR. GREENBERG: We heard that African-Americans and Hispanics in some of the big national surveys are a substantial proportion of those surveys, way beyond the population, and enough to do an analysis on. So I would think some of this on the Asian population does drive this first report.
But the groups that you did not adequately hear from --
DR. MAYS: Are more the OPI, the other Pacific Islanders. For example, the Palaui are going to be coming to APHA, and there are some other little meetings, and we will hear from them. The Samoans felt we didn't hear sufficiently from them.
DR. PAISANO: Maybe we should include them in this --
DR. MAYS: No, I'm just saying, there were some that we asked before and they couldn't make it.
DR. GREENBERG: And they don't necessarily have to live in the territories. They could live in Hawaii, they could live in California, right?
DR. MAYS: Right, as long as those territories feel that that person can represent them. Believe me, we have gotten into this before about these issues about the people who are on the mainland and the people who are in Hawaii and who speaks for them. We have got a little trouble with the Samoans; they didn't want someone else speaking for them.
DR. GREENBERG: Actually, the Samoans were not large, but a population in the continental United States
DR. PAISANO: There are a significant group of people in California.
DR. MAYS: Yes, they are a significant group in California.
DR. GREENBERG: They might have cultural things and everything that are similar, but they have got different issues than the Samoans living in American Samoa.
DR. BREEN: One of the points that was made was that one of the big issues of living in American Samoa or Guam or the Marshall Islands or a lot of the fairly isolated Pacific Islands is, they don't have a good health care infrastructure, and often what brings them to the United States is that they are sick or a close family member is sick.
DR. HUERTIN-ROBERTS: Maybe we should have a first nations report including the Pacific Islanders. All we need to do is, in the report discuss and address the fact that some of the groups we are talking about, the American Indians and Alaskan Natives and to a certain extent Native Hawaiians, do have this special relationship with the U.S., and that needs to be addressed and researched, and there will be other issues with the other populations.
DR. MAYS: Let me address that then with Paul. But the contractor that was a possibility is Paul Ong. If anything, you could probably speak to Paul's work. I think this is a contractor that really hits the mark very well. I think he is on what, his second or third report for you?
DR. BREEN: Yes, he is doing a second report for NCI. We funded CHIS in order to get data on American Indians and Alaskan Natives, Asians and Pacific Islanders. We commissioned a report from him initially to find out -- we asked him to look at the different census regions to see whether the demographic and socioeconomic information about Asians was similar in all of those regions to what it is in California, because we wanted to know whether we could generalize from the CHIS data about Asians to the rest of the country.
We had a similar one done by Matt Smith on American Indians. Then the second one that he is doing now is, he is looking at the census PUMS file on Asians, he is comparing that to the CHIS random digit dial sample on Asians for California, and he is also comparing it to the list sample. What we are particularly concerned about is the list sample and whether there might be bias.
But Paul has been working -- I met him when he was in grad school a long time ago, and he was working on Asians then, at the University of California at Berkeley.
MR. HITCHCOCK: He does good work.
DR. BREEN: He does excellent work. He just published an edited book -- I was reading a review of it the other day -- on dissemination, affirmative action.
DR. MAYS: This is their civil rights project, I think.
DR. BREEN: Oh, okay.
DR. GREENBERG: Is he with the Asian American Health Forum?
DR. BREEN: No, he runs his own Asian study center. He is a labor economist by training, or an urban economist, maybe both, and he specializes in Asians.
DR. MAYS: He just came off of the census --
DR. BREEN: REAC, it is called, Race and Ethnic -- help us out.
DR. PAISANO: Advisory Committee.
DR. BREEN: Advisory Committee.
DR. MAYS: And he was specifically on the Asian one. He also has close contact with Matt Smith, who is on the American Indian/Alaska Native one.
DR. BREEN: Matt said he is president of it now, or chair or whatever it is called.
DR. MAYS: Recently I saw that he is chair.
DR. BREEN: Now, Matt might be willing to help out with this thing. I know at one point you wanted him to do it, which I think would make sense, because American Indians are going to be the largest population of these populations we are looking at.
DR. MAYS: Yes, he was going to work with Matt, but I think the contract would probably be to him, and then him to deal with Matt, because I think he would be responsible, and bring that on.
I think that if we are going to also include these other populations, one, I would need to go back to him and talk with Paul. Two, because I am learning my lesson, I want to make sure we are very clear about what we want relative to this. So I think it would be better, even though he -- I think if he knows that he is going to do it, he can begin the work this summer. But I think I want to make sure we have clarity, because he is actually talking about analyses he can do from data, et cetera, to illustrate these issues.
But what I want to do is to make sure that we go back and forth with iterations on what this would look like. The only wild card, and I don't think it should be that wild, is we might hear something we hadn't thought about in our hearing, in which case we can't then ask him to go yet down another path.
If we are going to call it a first nations report, and we are going to include these other Pacific Islanders, while he can do basic analyses with what he has, if another emerging issue comes up, it just won't be something he does, but we just add it into the report ourselves. That would be what I would suggest.
DR. BREEN: You can add on to the contract, too. Modify the contract.
DR. MAYS: But part of what I think you are trying to make sure happens this time, Marjorie, I have written stuff which we have sent, and he has written stuff back which you sent. For example, the notion of the indigenous came up. He has been on Census, so he actually knows this very well, and he wants to do some analysis that uses actually the census data to illustrate some of those points.
So I think we should go back and forth maybe a time or two more. But what is also very helpful is that -- because Nancy has written a scope of work with him, I think we have some sense of what he really can produce. I think he has done well in terms of what he really produces. I think he is on the mark.
DR. GREENBERG: It sounds feasible. My concern is that we bring to closure the contractors we currently have. I think we have a plan for that.
DR. MAYS: Yes.
DR. GREENBERG: If Olivia is supposed to include some of the Asian stuff in her report, --
DR. MAYS: It is in it.
DR. GREENBERG: It is already there?
DR. MAYS: Yes, she has done it.
DR. GREENBERG: Okay, fine. Then if we want this additional contract, that we be very clear in our scope of work what we want him to do and what the time frames are.
DR. MAYS: See, I think I would make the time frame a little longer. We might strive to get it by November. I think I would stretch the time frame to February.
DR. GREENBERG: Yes, because actually, the committee is meeting in early November, and APHA is after that.
DR. MAYS: Yes, that's right. It is. We meet November 5-6, and APHA is the 13th and 14th.
DR. GREENBERG: Yes, so it will have to go to February.
DR. MAYS: Yes.
DR. GREENBERG: That's all right.
DR. MAYS: Can we talk about cost with you? I'll get back to him and talk about how much we have.
DR. GREENBERG: I think we need to know what the scope is.
DR. MAYS: That's what I say. I'm going to go back to him with the additional part of this OPI part, and then ask him about cost.
DR. BREEN: You had a copy of the report that he did for us, and I was wondering, are you thinking of doing something similar? This was pretty basic, and he can do more complicated things, but he just looked at the regional distribution of the demographics. He looked at age and gender, and then he looked at the top three ethnic groups in the different regions to see what they were.
DR. MAYS: I think some of it was that, but it was a little more -- for example, is that something that is available to us?
DR. BREEN: Hey, we work for the federal government, Vickie. It is all in the public domain.
DR. MAYS: It is? No, it is in a contract. I don't think so.
DR. GREENBERG: No, it was done under contract, then it is available.
DR. BREEN: Sure.
DR. MAYS: It is?
DR. GREENBERG: If it's under contract, yes.
DR. MAYS: Oh, I thought it was just the opposite.
DR. GREENBERG: No, unless it has some privacy issues or something, but this wouldn't.
DR. BREEN: You already have this, so you should be building on this.
DR. MAYS: I thought that is your thing.
DR. BREEN: No, not at all. It is one Department.
DR. MAYS: Then let me take that and work with Paul. Thank you, I didn't know we could do that.
DR. BREEN: We've got one just like it, except done by Matt and it is about American Indians. You should have that one, too.
DR. MAYS: Can you send me that one?
DR. BREEN: Yes.
DR. MAYS: This is so nice.
DR. BREEN: Does the full committee want this? Should I send it to Gracie and she can distribute them?
DR. MAYS: You can do that? Of course. If you send me the other one, then yes.
DR. BREEN: If we can build on that, we might as well, because we may want some additional stuff.
DR. MAYS: Then this probably makes it easier.
DR. BREEN: Or that may be sufficient.
DR. GREENBERG: I wondered if Nancy, since she didn't have an assignment on the letter, she could work on the scope of work, the one he is going to do.
DR. MAYS: She has already said she would, yes.
DR. BREEN: Okay, we need to figure out what we want him to do.
DR. MAYS: I think what we should do is get your other report, and then I think maybe what we ought to do is have that be -- I'll have to talk to Paul, because I told him I would let him know after this, but the August 6 call, if everybody has it at that point, then maybe we can have a discussion with the entire group on the sixth.
DR. GREENBERG: You mean on the scope of work?
DR. MAYS: Yes, the checking in on the scope of work. That will have given me a chance to read both reports for the staff to talk a little bit about, is there other issues that aren't covered, what those are, and then we can deal with scope of work.
DR. BREEN: What I would suggest, having read the report and knowing what this group wants to do a little bit more, all these reports do is to say, this is the demographics and the socioeconomic situation for these groups in California, and this is how they compared in the other census regions.
I would say that it would be useful to try to get a sense of what some of the underlying problems that we can measure with the census are for these groups that we know are related to health disparities, that might need to be eliminated to get rid of health disparities.
DR. MAYS: From the Census?
DR. BREEN: Sure, income and education, the socioeconomic status measures. Then if a population is particularly old or particularly young. All these things have an effect on health, if the gender is biased in a particular way.
For example, what we are going to find among American Indians, and I wouldn't be surprised if we found it among some of the Island populations too, is that the reservations are going to contain older and younger people much more than work age people. The work age people are leaving. Why? Because there aren't jobs on the reservations.
So these kinds of things aren't emphasized in these two reports, because we are literally trying to figure out, can we generalize from this California data to the rest of the country. I think we want to get in a little bit more to causation and association and things that we think might be underpinning health disparities in the followup report.
DR. MAYS: That is really his specialty, is to talk about contextual issues.
MR. HITCHCOCK: We want to capture some of the flavor of the hearings, too.
DR. MAYS: You have something too to contribute?
DR. COOPER: We'll be coming out before the end of the fiscal year with our race and ethnicity report.
DR. GREENBERG: The what?
DR. COOPER: The race and ethnicity report, talking about drug abuse in different racial and ethnic groups. So that is going to print in August.
DR. MAYS: So you can probably share -- can you share it with us before the August 6 call?
DR. COOPER: I'll try.
DR. BREEN: We have one coming out on cancer, but it is in clearance. Can anybody help with that?
DR. MAYS: What does that mean?
MR. HITCHCOCK: You mean Departmental clearance?
DR. BREEN: Departmental clearance. Is yours going through Departmental clearance?
DR. COOPER: It's already gone.
DR. MAYS: Can the committee say it needs to use it? Can they hurry up?
DR. BREEN: Maybe.
DR. COOPER: I don't think that will get it very far.
DR. MAYS: Okay. Let's do the last couple of things here. Debbie, can you tell us, are there any other things that are going to the executive subcommittee for them to look at, in terms of reports and stuff like that for the meetings, so we will have an idea of about when those things are going?
DR. JACKSON: Standards probably has something.
DR. MAYS: About when are they going?
DR. JACKSON: I'm looking at the two items, Standards and probably NHII. Time wise, we are trying to look at two weeks before the meeting is having everything ready to look, and then a week before that, if we were going to try to implement getting material out to the executive subcommittee.
DR. MAYS: So it is about three weeks before?
DR. JACKSON: The first or second, yes.
DR. MAYS: September when?
DR. JACKSON: First or second, in preparation for that first week for the executive subcommittee to take a look at that.
DR. MAYS: We were thinking somewhere around those. We have a call on the 20-something.
DR. GREENBERG: What is the executive subcommittee to look at?
(Simultaneous discussion.)
DR. MAYS: Gracie, if we push it up to the 19th, we might then see if we can't schedule a call for the very first of September or the very end of August. So that gives us a time for another iteration of whatever we are working on.
DR. JACKSON: We have the Labor Day weekend right before that.
DR. MAYS: So that gives us an idea. This corresponding liaison assignment, this is what we were talking about in the executive subcommittee, in terms of trying to make sure -- we have people that I think would do well to be assigned to the other groups, so that as the information -- it just means that they are on e-mails, as I understand it. They are not assigned to come to meetings or anything like this. But if there is an issue, they know to bring it to us. Then if we have an issue, then they can pipe up on the e-mail, or they can also be on the conference calls, I think.
So I just want to make sure that I understand the ones that probably we would do, which is just to the other subcommittees, right?
DR. BREEN: Yes.
DR. MAYS: So we will make those assignments in there. We will talk about those assignments with the lead staff. Audrey and Dale will talk about who are not committed. It is not like we have ten people.
Anything else?
DR. BURWELL: Are we going to do anything with the states?
DR. MAYS: Not until we can finish this. We've got such a big -- our fine china is broken pretty soon, so I don't want anything else on the plate.
Are there any other things that people want to put on the table? Anything else administratively? Do you all have any questions for us, so that we can take it into our planning process?
DR. JACKSON: We're going to put our piece on the agenda for September for the full meeting.
DR. MAYS: We'll appreciate that.
DR. JACKSON: If that will work?
DR. MAYS: Yes. The only other population issue is, are you putting us on the November meeting to take care of that panel that Jim talked about? I sent Jim a note, but I haven't heard back from him.
DR. GREENBERG: Which panel was this?
DR. MAYS: This is the panel on population health, in which we would add Dan, Barbara and this outside person.
DR. GREENBERG: Right.
DR. MAYS: So you need to tell us about that.
DR. GREENBERG: Wasn't this to look at how the conceptual model or whatever in the 21st Century report was relevant to the subcommittees?
DR. MAYS: Yes.
DR. GREENBERG: Each of the subcommittees' mission?
DR. MAYS: Yes. But I remember now, Barbara said that she thinks she had to give a talk in November, so she is not sure. I will write her back and ask her, but we need to get some sense --
DR. GREENBERG: September probably isn't that good for it.
DR. MAYS: No, no, we weren't going to go it in September. It would be either November or February.
DR. GREENBERG: Dan Friedman is leaving the department of health at the end of September.
DR. MAYS: Going where?
DR. GREENBERG: Looks like he is going to be independent. He is not leaving Massachusetts, he is just going to be retiring from the state, the department of health.
DR. MAYS: So we just need to get some sense of that, and we also need to deal with -- and you suggested at the executive subcommittee level, about the liaison issues to the Board of Scientific Counselors. That meeting is coming up in October.
DR. GREENBERG: Right, I guess it is.
DR. MAYS: So if we are going to do something, if somebody is going to do it, they need to get it calendared. So if that prompts you to have the subcommittee conference call, or by e-mail or something, that would be helpful if you would do that.
DR. BREEN: What is the Board of Scientific Counselors going to do with this committee's work?
DR. GREENBERG: We have a commitment of some type of liaison relationship. We have even talked about a member on both, but that isn't happening currently. We talked about the potential as an occasional joint meeting, but that isn't planned currently, either. So at a minimum, we would want to have a liaison from the National Committee, and this subcommittee seems to be the appropriate one, who would attend the meetings of the Board of Scientific Counselors to represent the National Committee.
MR. HITCHCOCK: Are those meetings always held in public?
DR. GREENBERG: Yes, it's FACA.
DR. MAYS: So no more administrative stuff. Does any of the members, anything you want to add, anything you want to take off? Thank you very much.
(Whereupon, the meeting was adjourned at 4:25 p.m.)