Hubert H. Humphrey Building
Room 337A
200
Independence Avenue, S.W.
Washington, D.C.
PARTICIPANTS:
Committee Members:
Lisa I. Iezzoni, M.D., M.S., Chair
Hortensia Amaro, Ph.D.
Richard K. Harding, M.D.
Vincent Mor, Ph.D.
M. Elizabeth Ward
George H. Van Amburg
Staff:
Carolyn Rimes
Olivia Carter-Pokras, Ph.D.
Patricia Golden
Ronald W. Manderscheid, Ph.D.
Page
Lisa I. Iezzoni, M.D., Chair
Rachel Block, HCFA 3
Stanley Nachimson, HCFA 12
Lisa Herz, Medstat 42
James Hadley, HCFA 65
William Clark, HCFA 85
Shruiti Rajan, Urban Institute 98
Committee Members
Review and Comment on OMB Directive on Race and Ethnicity 199
Hortensia Amaro, Ph.D.
Olivia Carter-Pokras, Ph.D
Improving Data on Health Status of Racial/Ethnic Groups 214
Edward Sondik, Ph.D., NCHS
Jennifer Madans
DR. IEZZONI: What I want to do is any final, like, edits on it and then if we can vote.
MS. ARAKI: Where do you want to put that one? Here.
DR. IEZZONI: Maybe up there. And Lynnette, would you mind being the recorder?
MS. GREENBERG: I didn't mean to -- and maybe you didn't take it that way -- but to suggest that you couldn't look at surveys being done, you know, even outside of the apartment --
DR. IEZZONI: I had actually talked to Carolyn about that.
MS. GREENBERG: Okay.
DR. IEZZONI: Does anybody have any suggestions on this version of the charge work plan? Okay. Lynnette, George has a suggestion.
MR. VAN AMBURG: On No. 5, I don't think the subcommittee is really going to recommend any sampling methodology procedures. I would take that and just eliminate that part of the sentence and say the subcommittee may make recommendations for collecting surveys for improving the quality and quantity of data, just cut out that middle there because I don't think we are going to provide that type of technical recommendation.
DR. IEZZONI: Okay. That was actually a recommendation of one of the commenters, that I add that in.
MS. GREENBERG: Well, it had been in an earlier version and perhaps it is a question of linguistics. I don't think, obviously, the subcommittee is going to develop a sampling plan. We certainly are attempting to get a few more people on the committee, who really have survey --
DR. IEZZONI: How about sampling approaches rather than methods because that is vaguer and it is more accurate as to what we are --
MR. VAN AMBURG: I would be happier with that.
PARTICIPANT: Could I make a comment at this point since you are on Item 5? I was looking at the document in terms of the ability to capture -- we are really not looking at sampling per se, but how we can get detailed data for populations living in concentrated areas.
So, it is not in this document but I can see -- I think what the subcommittee would be looking at is encouragement of the current methodology as part of that sampling. I think it is a broader issue than that.
DR. IEZZONI: I guess it wasn't part of discussion either, I guess.
Marjorie, do you have a suggestion on how to deal with this?
MS. GREENBERG: I just thinking that at the end of that sentence we could say the public health initiatives for these populations at the national and community level or something -- at the national and sub-national level. I don't know. Something like that.
DR. IEZZONI: Would that work?
PARTICIPANT: I think to get it in there so -- I don't know how much you can do on this, but it is such a key to thinking about getting this data in the future, that I think you don't want to exclude it and you want to begin to pull it into your discussion.
DR. IEZZONI: Lynnette, I think Marjorie's words were good.
MS. ARAKI: So, it reads, "The subcommittee will make recommendations to federal agencies conducting surveys, e.g., the National Center for Health Statistics, on sampling approaches, data gathering procedures, additional or supplemental studies and other means for improving the quality and quantity of data to inform public health initiatives for these populations at the national and sub-national levels."
DR. IEZZONI: And community levels. At the national and community --
[Multiple discussions.]
Any other suggestions for wording changes?
DR. MOR: I have a question.
DR. IEZZONI: Yes.
DR. MOR: My question is that there is no time frame.
DR. IEZZONI: There never is.
DR. MOR: As long as there never is, that is fine.
MS. GREENBERG: No, no, that is not true. We used to have that work plan.
[Multiple discussions.]
DR. IEZZONI: That is why there are no dates on this and there will be no dates on this. This will be modified.
MS. GREENBERG: If you do have like --
DR. IEZZONI: Medicaid managed care, about a year.
MS. GREENBERG: Generally, we do have milestones.
DR. IEZZONI: But we also -- typically these charges aren't dated, are they? The work plans typically aren't dated either, are they?
MS. GREENBERG: Yes.
[Multiple discussions.]
Because otherwise there is really a version confusion.
DR. IEZZONI: Okay.
DR. MOR: My point really is that there is actually -- there are about three of these or four of these things that are current right now.
DR. IEZZONI: And that is what is good because it means we can tick them off once we finish them. So, we can say we have done --
DR. MOR: We could also finish them.
DR. IEZZONI: Not necessarily, Vince. You haven't been on this committee long enough.
[Laughter.]
PARTICIPANT: It takes awhile to finish --
DR. IEZZONI: Some of these things, you know, the demonstration that started in 1982 and it is still going on.
MS. RIMES: But I wouldn't use that as an example.
DR. IEZZONI: But anyway, don't feel uncomfortable with this. Those of us who are older than you and wiser, having served on the committee are very comfortable with this level of vagueness.
DR. MOR: I can quote you.
DR. IEZZONI: No, hopefully, the tape isn't on. Yes, it is. The last five minutes.
Okay. Other suggestions?
DR. HARDING: Could we have a footnote on No. 2, U.S. territories and islands, defining what that is or listing them?
DR. IEZZONI: That is kind of what I meant by review federal health programs. We need to define what that is absolutely, but maybe not in the charge unless there is a one sentence or -- is there a way to define that to make it clearer?
DR. AMARO: You might want to make sure we are using the right terminology there, you know, so our charge has the right --
DR. IEZZONI: Right.
DR. AMARO: Is it U.S. territories and islands or is it just U.S. territories?
MS. RIMES: You have the wording.
DR. IEZZONI: Yes. I used wording that Lynnette kind of -- I listened to Lynnette. On No. 2, are we using the words correctly? Does "territories and islands" cover everything that we are talking about?
MS. ARAKI: Well, yes. It would cover the Virgin Islands -- but we don't necessarily refer to them as U.S. islands.
[Multiple discussions.]
DR. IEZZONI: So, Lynnette, could you put the appropriate language in here for what we are talking about, including the Puerto Ricos and the Virgin Islands. So, the whole range -- you know, basically the land masses that we feel responsible for and provide federal money for health programs, but are not the 50 states or the District of Columbia.
MS. ARAKI: The District of Columbia is excluded from this?
DR. IEZZONI: Yes, it is, but we would like clarification -- if staff would look into that for us, we would like clarification as to what D.C. really is.
DR. CARTER-POKRAS: Well, D.C. is included in many of the sampling frames for the surveys.
[Multiple discussions.]
DR. IEZZONI: It was just Richard's question as to what is really D.C. It is not an island, although it feels that way sometimes, or a territory or an off-site state. It is more a --
PARTICIPANT: It is a philosophical question.
DR. IEZZONI: Any other word changes? Going once -- shall we enact some business here? Do I have a motion?
DR. MOR: Move to pass it.
PARTICIPANT: Second.
DR. IEZZONI: Everybody in favor of the motion to pass the work plan.
PARTICIPANT: As amended.
DR. IEZZONI: As amended, by Lynnette. Yes?
[There was a chorus of "ayes."]
Okay. Good. All right.
Lynnette, you will amend the work plan according to the editing changes that we just made and you can tell Don Detmer that we have a work plan and a charge from our subcommittee.
MS. ARAKI: So then you report this out in June and the full committee adopts it.
MS. GREENBERG: September.
DR. IEZZONI: September. It is July 21st.
MS. ARAKI: It will be dated today.
DR. IEZZONI: Yes, date it today.
DR. CARTER-POKRAS: I have a clarification question. I am just wondering, the subcommittee is interested in hearing about all major health concerns, not just chronic illness. Right?
DR. IEZZONI: Right.
DR. CARTER-POKRAS: Because the second sentence says concerns about chronic illness -- under the bullets, it says identifying major health concerns and the second sentence says concerns about chronic illness.
DR. IEZZONI: It says public health concerns and then concerns about chronic illness. It is actually language that I worked on with Hortensia to try to be quite broad and inclusive. That is what our goal was.
MS. GREENBERG: So, the public health issues
were --
DR. IEZZONI: Yes, and it says below -- you have got bullets, identify major health concerns affecting these areas, such as public health issues, chronic illnesses, mental health concerns and "such as" always allows for other things.
We actually are going to be talking just the bunch of us for the next about hour and 40 minutes or so. We will take a break before Dr. Sondik comes. We have a lot to do in this time actually.
Let me just sketch out what my thoughts are. About three o'clock -- maybe we should break right before Dr. Sondik, after we have talked about the OMB Directive 15 revisions. So, let me start there first because it is clear.
The OMB Directive 15 revisions, you all got the Federal Register reporting of that last Wednesday or something. It actually came out very recently. So, poor Hortensia and Olivia haven't had a chance to connect with each other to even begin to think about it. So, what we are going to do at 3 o'clock or so, maybe a little bit earlier, so we can have a break before Dr. Sondik comes, is that Hortensia and Olivia would just like to hear from us about any thoughts that we might have based on our reading, if we have had a chance to read --
[Multiple discussions.]
It is a very long document. It is very informative, though. I mean, it is obvious that a lot of work was done to lead up to this document. So, we will talk about that at 3 o'clock.
Then Olivia and Hortensia will have to get back to us with their letter. And just to anticipate, that letter is due September 8th. So, we are going to have to be dealing with it over the next month when a lot of us are going to be unavailable by e-mail.
DR. CARTER-POKRAS: Actually, Hortensia is gone the entire month of August. When is your last day in the office?
DR. AMARO: The first, I think.
DR. CARTER-POKRAS: So, we basically have to get our stuff done before the first.
DR. IEZZONI: All right. So, 3 o'clock, if people can be thinking in the back of their minds, that is what we will do. A little bit before 3:00, we will talk about the OMB Directive 15.
Before that, we need to come up with a very specific work plan for our year long effort to look at Medicaid managed care. Let me tell you why it needs to be specific.
We have the potential of having a contractor do some work for us. We might even have the potential of having a contractor, such as Mathematica or one of the other organizations, that has been very involved thus far with looking at some of the programs that are ongoing with Medicaid managed care. They could potentially be a contractor, so long as they have a work order, I guess, agreement with NCHS or --
PARTICIPANT: They are on the ASPI task force.
DR. IEZZONI: Are they on the ASPI task force? Okay.
So, if we want to ask one of these organizations to do some work for us, we need to be quite explicit about what that work is, so that Marjorie and Lynnette and Caroline can begin to work to get, I guess -- is the technical term a task order in place?
MS. GREENBERG: Yes. Developing a scope of work for the task force.
DR. IEZZONI: Developing a scope of work. Okay.
So, I think what I would like to do at the end of our little discussion is to really focus on what we might want a contractor to do and there might be two or three different things that we would like them to do and we might like them to do one of the things now and then wait for awhile and after we see what that shows and come back and come up with another contract later on. We can kind of couch it any way that we want, but we need to be clear about what we want.
The second thing is we need to decide among ourselves how much time we are willing to commit to this and what we really expect for us to be able to have accomplished by the end of the year. Okay?
So, what I would like to do, though, is really start off our discussion by being quite broad. This is kind of going back to something that I heard very clearly Don Detmer say at the June meeting, but maybe the rest of you didn't hear it because you weren't oriented the way that I am, but Don said at some point I would like to know what that camera is really for. Okay? So, that is why we handed out this camera thing.
It came around with the subcommittee charge. The two of them were together when Gracie handed them out.
So, let me just talk through the camera a little bit and then I would like to shut up and listen to the rest of you talk.
Do people have that? Everybody has got it.
Let me tell you where the camera came from. And poor George has heard this about 13 different times, so he is going to sit there and be polite and just listen again.
Back before you new members came on board, we old and cranky members were trying to figure out how we were going to deal with the fact that we had this techie invasion, this potential for the committee to basically become a very technically-driven committee. We were concerned about that because we knew that lurking around the bush was this whole emphasis on standards and ANSI --
MS. GREENBERG: This was before the legislation.
DR. IEZZONI: This was before the legislation passed but we were having -- at the full committee meetings we were having all these people come that would talk in acronyms the whole time and some of us were really thinking what are we here for. Okay?
So, Kathy Colton(?) and I, since we are both in Boston and can get together easily, volunteered together to sit down and think about the answer to this question, what are we really here for. And Kathy, who tends to think metaphorically, came up with this camera analogy and that is that what the committee is really here to do is to think about what questions we need to look at concerning the health and health care of the American people.
Once you kind of look at that broad level, then you begin to focus increasingly in on how you answer those questions. Kathy and I put data transmissions and standards at the very bottom of the camera, which is not bottom -- you can hold it this way. Okay? It is not the bottom technically. It is the lens, as Kathy would say. You know, that is the lens.
But to get to that lens, you need to begin to think, okay, here are the questions, but who are the people that we are concerned about when we are asking these questions. What are the events that we are concerned about when we say we want to identify these areas of health and health care? What is the information that we then need to begin to look at? What are the data elements that are required to produce this information? And then it gets increasingly kind of technical.
We separated the camera into the upper part, which is information policy and then the bottom part, which is standards. So, what I would really like to do today during the first part of our discussion, because I think it would really help us figure out what we want, is ask the question: What do we want to know about Medicaid managed care? Okay? Because we can only begin to think about data systems to evaluate Medicaid managed care if we first ask the question: What do we want to know about Medicaid managed care?
So, that is why we have a flip charts here is because what I would like to do is just kind of go around the room and get people's questions. What do you want to know about Medicaid managed care? And then -- because there were a lot of things this morning that came up and some of them, I think, are going to be of more interest to us and others and there might be things that didn't come up this morning that are of interest.
Yes, Marjorie.
MS. GREENBERG: I just want to -- it may or may not help, but for those who are new to the committee, also understand that part of the motivation was also -- there was this possible situation where we were going to have two committees or there were people thinking in terms of two committees, one to do all the data standards stuff and one to do this more of the survey, et cetera.
There was a real sense that that would be a mistake not to have the data standards and the data sets, et cetera, informed by this broader picture would not be a good thing. So, I think that if to some degree if there is some kind of disconnect sometimes between like the types of issues you are talking about today and some of the standards stuff that goes on in the committee, I think, it is -- in the full committee, it is important to remember that at least there was a buy-in to the linkage by having one committee.
DR. IEZZONI: Okay. But for the purpose of today's discussion, I think we are really thinking in terms of the more global perspective and we may never for this particular exercise get down to the level of standards.
MS. GREENBERG: Well, on the other hand, if what is available for Medicaid managed care ends up, you know, basically being the HIPAA transaction because that is claim or equivalent encounter data, then I think it is very important that there is a linkage there.
DR. IEZZONI: Well, but that is why we need to ask these questions.
MS. GREENBERG: Right, but not this minute, but --
DR. IEZZONI: Because these questions may not be answered by the ANSI 837 form or whatever, you know, whatever these things are. And we will get down probably to that level at some point or we could ask a contractor to get down to that, you know, to look at those kind of issues for us.
Okay. So, George, I know you were one of the people that was really interested in having Medicaid managed care be our topic for in-depth discussion this year. Can you just tell us a little bit about why you felt that and what you wanted to know about it?
MR. VAN AMBURG: Well, I was particularly interested in knowing whether the health care patterns that the managed care people were getting were better, equivalent or worse than the health care patterns in the fee-for-service, particularly access to health care. And as I talked this morning a little bit about the carve out population services, I think that is going to be a real problem is to be able to monitor whether these people in the managed care are getting the appropriate services or not.
MS. WARD: By health care patterns, do you mean behavior of the providers, the results of the care?
MR. VAN AMBURG: No. I am talking about where they are getting access to various types of services and care that they need. I am also concerned about what we didn't talk about this morning, was the provision of dependent services under the managed care.
I mean, it was mentioned a little bit this morning about family planning services, but we have got all kinds of prevention services in public health and what I have seen so far, at least, in my limited work with the Medicaid was that the whole public health arena kind of just disappeared during this managed care thing with Medicaid.
DR. IEZZONI: Okay. So, Lynnette, we should also put carve out up there because the carve outs was one of the real issues, I think, this morning and George just mentioned it.
Do we have any tape, Gracie? Can you steal some tape from somebody so we can --
[Multiple discussions.]
Okay. Hortensia, what are your questions?
DR. AMARO: Well, I thought that -- I mean, these are big questions -- so, I thought really you would be interested in knowing how it impacts access, which you mentioned, but also quality and cost and you would probably want to know something about -- quality and cost and then you would also want to know something about health status of populations and, you know, prevention, I think, is an additional -- so, I had those five points. Access, quality, cost, health status and -- I guess access to prevention -- I am not sure where to fit that in. Maybe it is sort of under access, but it is a broad -- it is not just medical access.
DR. CARTER-POKRAS: We only counted four. Access, quality, cost, health status --
DR. AMARO: Well, if you put prevention under access, then that would --
MS. ARAKI: Would it be like this, Hortensia? Provision of preventive services, does that cover your --
DR. AMARO: Access to quality --
MS. ARAKI: Quality and cost, health status, access here, provision of prevention services.
DR. CARTER-POKRAS: I heard health status or health outcomes.
DR. AMARO: I would put -- I mean, I don't know if it matters, but I -- on her list there at this point, but I do see quality and cost as separate.
DR. IEZZONI: Elizabeth, you were also somebody who was really interested in having this be the topic for this year's focus. Do you have any additional questions?
MS. WARD: I think that those cover the traditional access, quality --
[Multiple discussions.]
But I do think we also have to look at what the impact is on providers and by "providers," I mean both health plan and direct patient care services because I -- and I think, I have to think out loud or use some other, you know, sort of comment here about what it is -- that is too broad. We need to take it down, but I think we are -- by the reporting requirements, we are also continuing to squeeze providers so that providers are becoming -- there is evidence that providers are becoming less and less satisfied with their work, more and more pressured to do things that they used to be able to add on without worrying about the costs to the kind of care they provided because they feel they are getting so squeezed, they are not willing to do the kind of things -- and care environment is becoming increasingly hostile because of the unhappiness of the people who are being squeezed.
That is not true uniformly, not every provider feels that way, but there certainly is increasing -- my contact with providers continually to give the anecdotal evidence that they are being -- they feel punished -- and it is not Medicaid managed care specifically, but that is part of the population that is going under managed care and it is the managed care -- and they are uneasy about adding the Medicaid clients to that increasingly difficult environment.
DR. HARDING: Just to add to that, there is a feeling of changing of expectations that the people who are enrolled are marketed to be enrolled and they are marketed with very high expectations. And when that isn't provided, they are mad.
DR. IEZZONI: So, were their expectations on that based on the lure to get into the plan?
DR. HARDING: Yes, the marketing of the plan to, you know, come to us and we will take care of everything really and it will be wonderful and then it is not and then they are mad, but I don't know exactly how to --
DR. AMARO: Well, that is really kind of alluding to the patient/provider relationship and the impact of this whole system on that relationship.
DR. MOR: I think about this always -- I am the camera but from the micro data element up and then what information can be generated at each level of abrogation. So, I think that -- I also think what are the benefits and what are the costs in the broadest possible sense of managed care for this particular population. So, in thinking about those, some of the benefits to start with are the possibility of integration because while we all are worried about managed care, there is one we do know about fee-for- services is that you will bounce from provider to provider to provider and there was never any incentive to be -- having an integration of care.
To study integration, you must be able to construct histories. To construct histories, you must have micro level data.
DR. IEZZONI: Can we not talk about the data yet?
DR. MOR: Okay. The issue then is integration and continuity of the care process. At another level, you can also begin thinking about substitution of whether or not the way in which care is being used is different in one setting or under one structure versus another. So, does one substitute certain classes of care for other classes of care, preventive versus curative or non-curative or institutional versus home, those kinds of things you can begin to allocate because you are able to allocate across.
I also think that that potentially results in a question that I think is very, very interesting, which is what drives, I believe, a lot of behavior is the market and the market is based on what certain economic sectors operate in and the extent to which managed care results in a different aggregate allocation of funds from one sector of the health care system to another sector of the health care system as a function of different usage patterns.
You can do that under this kind of scenario by looking at the data coming up. It is this issue of transfer of funds, you know. In my world, it is the -- you know, you transfer money out of the nursing home world and put it into the home care world. It happens as a gradual process and is there a differential approach that goes on.
On the other hand, you have a whole series of costs and concerns, which actually work at the administrative level on how things work out, which is part of this, for instance, these carve out functions. Another is the changes in the delivery of public health services, but other parts of this are related to administrative operations.
How do these things function under capitated, non-capitated basis and how are administrative costs allocated? I think a lot of the real concern is that right now the providers are feeling squeezed because they are maintaining multiple systems simultaneously and so that they are increasingly being asked to be responsive to a capitated or managed care environment when it still could be the bulk of their patients are in a fee-for-service world.
So, they are actually trying to do double duty, double work. They haven't yet shifted the way in which they behave. That has a major impact on them, given the amount of overhead, administrative overhead it takes for operation. Now, how to, you know, get at that issue of the allocation of administrative overhead -- some of the big complaints about managed care is that the companies are taking all the profit and they are extracting it -- they take the efficiencies that they are getting out of the system and putting it into the shareholders, as opposed to plowing it back into the system in terms of other kinds of services.
So, those are the macro level --
DR. IEZZONI: They are very macro level, but we need to think in those terms.
Hortensia.
DR. AMARO: The other thing that occurs to me is the impact of managed care -- we talked about the impact on providers, but also the impact on the broader public health community if there are certain services that are limited or not provided and then where is the next place that people are going to have to go to. For example, the number of visits for mental health or substance abuse is cut or limited and our city public health or state public health departments going to come in and pitch in. So, the broader -- I would like us to look not just at the impact on clinical services, but also public health because somebody else will have to step up to the plate if certain services are cut back.
DR. HARDING: Who is that?
DR. AMARO: Well, I mean, if you look at what has happened with substance abuse, it is usually states through the federal block grants that end up filling in. We shouldn't have to have that. I mean, that is my opinion. Medicaid should be covering that, but for some reason for substance abuse that has not been our approach, except for a short term, you know, detox and -- so, the state has had to step in or city departments provide funding a lot of times for all those pieces that aren't provided in our medical --
DR. HARDING: What really happens is it has been transferred to corrections.
DR. AMARO: But to the extent that it hits all kinds of systems, you know, it hits the foster care system. It hits corrections.
DR. IEZZONI: Richard, did you have other questions?
DR. HARDING: The issue this morning was interesting about -- that George brought up, but, you know, I had the feeling that like the Federal Government doesn't know what questions to ask. So, the data we get is going to be suspect at best and misleading, most likely. So, we have to come up with the questions that are going to help us determine whether we have a good access, quality of care, outcomes and so forth. We don't know the questions. And they were saying kind of they don't have those right now.
DR. IEZZONI: We have three state people here and that is why -- who were -- we are happy to have them. Can you give us a sense of how some of these questions might be different from a state perspective?
MS. WARD: I don't think they are any different. There is just the size difference. And I was struck by that same thing, Richard, about if you are going to spend the amount of money and take the risk of the kind of huge data generation that they are talking about, particularly, to me, the risk, since we spend a lot of our life debating security and confidentiality, if you are going to justify the risk to that size of data, you have got to promise those people you are going to get something out of it.
And they don't even know the questions. We do it in our state. We are dumping these huge databases and we are doing it because it is -- we are doing it -- we are justifying doing it because these are people who are using public dollars.
We do have a two-tier system. Those people have
-- they can't complain about the kind of data integration we are doing because they are recipients of public dollars, but we owe them the right to have their data as secure as anyone else's, but I don't think we have justified to ourselves the cost or the purpose or the burden on the reporters for doing this. I was very interested in her comment that several of the states have now just said that they haven't gotten anything back for the millions of dollars they have spent integrating issues databases. Well, of course, because they didn't start with what was the question.
DR. MOR: Actually, her comment related to the eligibility -- integrated eligibility determination systems, that EDS and other companies have actually put together for them. It has nothing to do with the data. It has to actually do with the uniform eligibility processes and not
-- they are just looking for up front mechanisms to rationalize the eligibility process. It doesn't mean at all that they have a downstream system of taking claims or event level data.
MS. WARD: I realize that but they are still saying that the cost even at that level, they are not sure it is justifiable in terms of what it is getting them. I am concerned it is because we haven't decided what the questions are first.
DR. IEZZONI: That is why we are starting --
MS. WARD: Right. And I think that is one of the things we ought to say to many of these agencies. Until they can be more clear what the justification and purpose of this, I am not sure we should be running down --
DR. HARDING: We have got these babies rolling and they said, well, what is the termination date or when do we know that we are doing a good job or something. Let them roll and we will see. That was pretty scary to me. What is the outcome that we are really looking for? And do we have a way to --
DR. CARTER-POKRAS: I am trying to get to tag on to the comment about the need to look at the impact of providers as well as public health. One of the concerns we have at the Office of Minority Health is the impact on minority providers because we know that, say, black physicians are more likely to serve Medicaid patients and Hispanic physicians are more likely to serve uninsured.
So, right now we find out because they don't have the credentials, they are not being enveloped into this movement into managed care. They are not being brought into these HMOs and they are at the risk of seriously losing their business, losing their ability to practice medicine. Because they don't have the credentials, they are not being taken up by the managed care organizations.
But they are the ones who are providing this essential care that is not going to be covered by anybody else, like uninsured persons.
DR. HARDING: Can I come back at you, though? In my town, Columbia, South Carolina, the black doctors are taking care of the wealthy blacks. They are not taking care of the Medicaid population.
DR. CARTER-POKRAS: This is at --
DR. HARDING: No, I understand, but there is that kind of a -- that you are going to get that, a kind of flight away from it because it is so frustrating and all the other things and they are going to -- I don't know who is going to take care -- clinics and so forth, but it is interesting to see that happening in our area, that they are going towards the wealthy instead of the Medicaid in their internal medicine practices and so forth.
DR. CARTER-POKRAS: The physicians that we are hearing about are the other group, who are --
DR. HARDING: I believe you. I am just saying this is kind of an unusual situation.
MS. GREENBERG: I was just going to say from the issue of, you know, not maybe having decided on what the questions are or how they would address them, I think there is a little bit more understanding about what the questions might be, but really not much at all sense of how they would address them and whether people are there to address them because this whole issue of encounter data, obviously, HCFA has been very conflicted over it from the beginning. And I was quite involved with it when I was -- a few years ago, when I was working in the Office of Managed Care and when we came up with the McData data set and the real argument, one of the main arguments against -- and this is the burden, I think, on the issue of federal/state relations and all of those played into not requiring it, but also there was a real concern that if we require all of this encounter data, we have to tell people what we are going to be doing with it, but the only way we could even begin to justify it is to be able to, you know, say these are the types of studies we are going to do. And then we would have to do them and be able to follow back with them.
Despite, I don't know, 30 year, 20 year track record of using the Medicare claims data, you know, fee-for-service data, extensively and in many cases productively, both it seems, HCFA and their contractors, there didn't seem to be much of a sense of whether that could be, you know, applied to encounter data coming out of the managed care, whether -- you know, and the connects there. So, that really has prevented a -- you know, making a strong case for encounter data and then going down to the state level and going to the plans, et cetera.
DR. IEZZONI: Well, I think it is implied once you begin to look at the questions because if we just look at the questions that Lynnette has written up there, how many of them can be addressed by encounter data?
Let me just -- shall we anoint you an honorary member of our committee --
MS. GREENBERG: But how many of those can be addressed if you have no encounter data? I mean, this micro level. I don't know.
DR. IEZZONI: Do you have any questions about Medicaid managed care that we should put on our list?
PARTICIPANT: I will speak up if I have any.
DR. IEZZONI: Okay. All right. We have been told how this is going to play out. Okay. Great.
Anybody else have anymore questions? There is a large range of questions up there.
Okay. Let's go then to the next level of our camera. I don't mean to beat this dead horse, but I think it is actually useful because the next level is the population and we are the Population-Specific Issues Committee. And we heard from our speakers this morning about a number of populations. We heard about infants and children and their mothers.
We heard about people with disabilities. We heard about people with HIV/AIDS. We heard about the mentally retarded. We heard about a bunch of different populations. As we begin to think about what we want to focus our study on for the next year, let's just list the populations that we might be interested in asking these kinds of questions about.
George, do you want to start?
MR. VAN AMBURG: Well, clearly, infants and children is a critical one.
DR. IEZZONI: Infants and children, yes.
MR. VAN AMBURG: And children with special health care needs is another one.
DR. CARTER-POKRAS: How about adolescents and young adults because there is an age break as to when you are covered with insurance and there is a general issue without insurance.
DR. IEZZONI: From Medicaid, does anybody know how that age break splits out? Is there an age at which --
MR. VAN AMBURG: 18 or 21, depending on whether they are developmentally disabled.
MS. GREENBERG: Perhaps some of the extra services that --
DR. IEZZONI: We need to find that out.
[Multiple discussions.]
All right. Hortensia.
DR. AMARO: The other obvious ones are racial and ethnic groups.
DR. IEZZONI: And we heard about that not just from the patient level, but also maybe the provider, the doctor and nurse level as well.
DR. AMARO: I am also very interested -- maybe this fits more under events -- I am not sure, but I am very interested in considering the role of substance abuse in hospitalizations and costs to the system. I am very interested in sort of that population and how it is going to fare in this because I think it is a population that has high need for service and that this costs the system a lot and that the needs are not met well.
But maybe that can be put rather in the population center then?
DR. IEZZONI: Population, substance abuse and mental health tend to track together. They are separate, but I think that that would be a population and, obviously, I will chime in with people with disabilities.
PARTICIPANT: Migrant populations are always a problem for Medicaid.
DR. IEZZONI: Homeless and migrant populations -- are they in Medicaid managed care?
PARTICIPANT: I think the migrant health --
DR. IEZZONI: The Medicaid managed care?
PARTICIPANT: They have no care except the clinics during the months that they are working.
DR. IEZZONI: We should find out about that, the migrant --
PARTICIPANT: At least for those states that have migrant populations. I guess it is not a Medicaid issue for states that use them and don't provide care. Usually, they are not there long enough or they don't qualify in terms of the weekly salary for Medicaid.
DR. IEZZONI: Are they in Medicaid managed care programs, though, do we know?
[Multiple discussions.]
There are states that only have Medicaid managed care.
PARTICIPANT: Your state does.
DR. IEZZONI: My state. I know. I knew that as I said that. I know what Bill Weld has been up to.
MS. WARD: I would just add rural. The doctors you are describing, Olivia, are the ones we have in Washington, who are in the rural parts of the state, who can't get access to any other population to care for, other than migrant health centers, community health centers. And they are all either uninsured, underinsured or just barely Medicaid eligible.
The major health plans are not going after them. They may be credentialed, but they are just not thinking -- they have their network of providers and they claim that those networks of providers, who go out to those rural areas, that they are not picking up the provider, who are already local and the communities are very anxious because they have relationships with like local providers, who may, in fact, be more culturally appropriate for them, but the Blue Cross of Seattle is not going to add those physicians to their provider list.
They claim that they will be able to go out and provide them sort of on an itinerant basis.
DR. IEZZONI: Vince, populations?
DR. MOR: There is the long term care population, which would be in addition to people with disability, it would be old and disabled.
DR. IEZZONI: I will confess that I was somewhat confused by the discussion of the dual eligible. Is that what you are talking about?
DR. MOR: If you look at the chart that they present, is it actually the dually eligible individuals in the world, in the U.S., are sort of bimodally distributed. They are much more likely to concentrate on the 85 plus. Therefore, the impoverished institutionalized -- spend down already occurred in the community, very old folks and the younger disabled. So that they are both -- it is very much bimodal from an age distribution perspective.
So, yes, there are both. In fact, you know, probably about 70 percent of the long stay in the nursing home population is dually eligible.
DR. IEZZONI: And are they in managed care?
DR. MOR: Right now, the only thing close is this EverCare model.
DR. IEZZONI: That is in Minnesota.
DR. MOR: It is United Health Care and they have about five replicates across the country. And the ONLOC and the Minnesota example that they gave. Those are community-based managed care in that sense.
Under the dually eligible program that Massachusetts just submitted, that long term care population is absolutely 100 percent in the option. The problem of the waiver is under Medicaid all these people have a choice.
DR. IEZZONI: I am sorry. Poor Gracie. This was a disaster. I am so sorry that --
PARTICIPANT: And you are also going to include chronic?
DR. CARTER-POKRAS: Actually, I was thinking of some examples of diseases, such as diabetes and asthma, that seem to really need better management.
DR. HARDING: I was going to say the same thing. We call them the medically fragile children that are diabetic or have a bunch of different things going, epileptic, maybe -- and then have in addition to that behavioral problems or all kinds of things.
The other dual -- the dual diagnosis, you know, that I think of, of course, you know, is alcoholism and a psychiatric disorder or alcohol or drugs and a medical disorder. Those are really utilizers with diabetes and alcoholism or something along those lines.
DR. IEZZONI: And AIDS.
DR. HARDING: And AIDS.
DR. IEZZONI: I guess we should put it -- we don't have HIV/AIDS up there as one of our populations, but we should.
Okay. Hortensia --
DR. AMARO: No, I was just thinking we are coming up with all of these. Maybe later we can go back and face those --
DR. IEZZONI: Well, that is what we are going to do now. Because I think that this has been helpful. We have got a lot out there on the table, but I am thinking, okay, how are we going to deal with this huge volume of thinking because we are going to have to come up with a task order work plan.
So, let's do one more little around-the-room exercise and let's -- yes.
MS. GREENBERG: I just wondered if I could add one more. I know you don't need anymore, but since the majority of the Medicaid managed care have been addressed to mothers and children, I wonder where we left the mothers off.
DR. IEZZONI: Right. I think infants and children was meant to include mothers.
[Multiple discussions.]
DR. MOR: When they were talking about a family ID number, it is basically another ID number for mom. Once you get all the kids hooked to a mom, they can be hooked no matter what. But you have to be hooking them to a mom.
DR. AMARO: If we could do that, that would go three-quarters of the way.
DR. IEZZONI: All right. Great.
Okay. As we look at the kind of questions that we are interested in answering and the populations that we are interested in asking the questions about, what are the data systems that we need to begin to worry about? Because to answer your macro level questions, Vince, we are going to need entirely different data sources or some very different data sources than to answer a question about, for example, immunization.
DR. MOR: I bet you could answer virtually every question on the table here, except those related to certain of the health quality issues unambiguously or relatively unambiguously, not that you could answer anything unambiguously, using the existing encounter, enrollment and provider level data currently being contemplated under HIPAA.
DR. IEZZONI: Well, let's think about that for
a --
DR. CARTER-POKRAS: Not currently existing, right?
DR. MOR: Well, we just all wrote a letter to the Secretary saying do it.
DR. CARTER-POKRAS: Because we have heard from managed care plans that say that they can't -- they don't have it. They don't have individual level data.
DR. MOR: But they will.
MS. GREENBERG: The thing is, will they?
MR. VAN AMBURG: Will they because they don't have a transaction. They don't have an electronic transaction.
MS. GREENBERG: This doesn't require that the information be collected. It only requires that --
DR. MOR: Fine, fine. Were that to be in place for all transactions --
DR. IEZZONI: Okay. All right. So, if we list data --
DR. MOR: -- for events.
DR. IEZZONI: -- one is encounter data and enrollment.
DR. CARTER-POKRAS: Individual level encounter data.
DR. IEZZONI: Individual, that is by definition an encounter.
DR. CARTER-POKRAS: No, no, no. Some HMOs have -- managed care organizations have aggregate level. Okay? So, they can --
[Multiple discussions.]
DR. MOR: That is how they are asked to report it.
DR. IEZZONI: But Hortensia wants to know about health status.
DR. CARTER-POKRAS: And that you have to do the linkage.
DR. MOR: Well, I would like to know about health status as well.
DR. IEZZONI: Health status you cannot get off an encounter. You have to do the linkage from encounter data to the enrollment eligibility determination.
DR. CARTER-POKRAS: Not for health status.
DR. MOR: Well, I think that is where we should push to say that what was in the original holding place should be recommended.
PARTICIPANT: You mean back to our core data set?
DR. MOR: Back to our core data set.
MS. GREENBERG: I can't believe a new member even said that.
DR. MOR: Of course.
DR. IEZZONI: Because these are data items that we don't have. Health status is a data item that is not on the current enrollment form or the current encounter form.
MS. RIMES: But at the same time, and let me use a more bizarre example, if HCFA is going out and collecting CAPs and HEATUS and the enormous numbers of God knows what on HMO populations, including the Medicaid population, which has health status and has your ADLs, your IADLs, all of those things, so the systems are out there.
MS. ARAKI: What did you say before?
[Multiple discussions.]
DR. IEZZONI: Consumer assessment of health plan survey. We are just listing information sources right now.
MS. GREENBERG: So, consumer surveys.
DR. IEZZONI: Okay. So, those are surveys, the CAPs and the HEATUS.
DR. MOR: The biggest gaps are the nature of the information about the providers and that is the gaps that we -- this particular committee has been most upset about in this whole deliberation about the data systems and the providers is that how do you gather encounter data and how do you code encounter data for the taxi driver, for the home health aide, for those other people who are actually extending much of the care to these kinds of populations? Because we don't have any code sets to document those activities.
DR. IEZZONI: So, then, that is basically data about non-traditional providers or --
DR. MOR: And what they do.
DR. IEZZONI: And what they do.
It is not a good word, but we will call it that for now.
[Multiple discussions.]
DR. MOR: Well, taxi drivers are just sort of the most bizarre, but it is the home health aide or the personal care attendant or the OT aide or what have you. Those are people -- there are no procedures and they haven't gotten -- at present, there is no contemplation for giving them billing numbers or numbers.
Eventually, they will be rolled in.
[Multiple discussions.]
But the next thing also is the nature of the information about providers and that that is still unclear.
DR. IEZZONI: There are other questions up there, though, Hortensia's question about the impact on public health. Hortensia, where would you get that information from? What kind of information would you need for that?
DR. AMARO: I guess it is related to access in that -- I was thinking of situations where certain kinds of things wouldn't be paid for and then other parts of the system would have to pick up on that. I don't know, unless you did specific sort of case studies of cities to find out what is happening to different types of models and whether certain types of care have dropped out and had to be picked up.
MS. GREENBERG: HRSA -- again, it is not individual encounter data, but they have a data system for their community health centers, migrant health centers, and they have been very interested, obviously, in the impact of managed care on those essential providers.
DR. IEZZONI: We talked this morning about integrating data at the individual person level from a variety of different social programs, the WIC program, even the Department of Motor Vehicles. South Carolina was the exemplar --
[Multiple discussions.]
We need to include data sources from other social support services that are provided at both the federal, state and local levels by authorities under those three levels.
DR. MOR: See, Hortensia, I think about, for instance, the clinics in south Florida or something like that that I am familiar with. They have a provider number and the people they serve and the encounters they provide can be categorized in terms of paying, not paying, of what class or not what class, what kinds of services apply to a provider, to whether or not they are reimbursed or paid or not paid.
You can actually aggregate that kind of information to the level of a provider over time and look at changes in the revenue flow, cash flow or mix of services and mix of people provided.
DR. AMARO: So, for example, there was Medicaid -- under Medicaid managed care, there was some type of care that had a limitation, you know, so many visits and then city health department decides that, well, you know, this isn't enough. We need to pick up on this --
DR. MOR: As long as there is a claim coming through that provider on encounter.
DR. AMARO: -- as long as they are required to fill that out, regardless of who the payer is. Right?
DR. MOR: Right.
DR. AMARO: Because if it is the city or the state that is filling in --
DR. MOR: The complicated issue is when those events get done -- provided outside of the system.
DR. AMARO: That is right.
DR. MOR: If the public health department opens up a new clinic that is just free, like HIV testing, there is no data whatsoever on anonymous testing.
DR. AMARO: Or somebody that isn't covered -- something that isn't covered right now, like residential substance abuse treatment that is long term that is not covered, they get reimbursed through state contracts and, you know, so, something like that wouldn't come in, would fall outside the system. You know, in health care, there may be other examples.
DR. MOR: I don't know much about those on the substitute side. If those people were providers, then they would record that. But other than that, you wouldn't see them. They would be invisible unless they are providers.
DR. AMARO: Well, there are some examples in substance abuse, for example, you can bill in the postpartum period for a certain amount of enhanced substance abuse treatment for pregnant women and postpartum women, but then after a certain amount, you are not going to get paid for any of that. So, then the state or the city might fill in, you know, in certain places to give continued care.
So, you know, if somebody was providing that kind of continued care that wasn't a regular provider under Medicaid, then I guess you would be able to get it unless it fell outside of the system in the way you are saying. So, you might capture a lot of that.
DR. MOR: But that would get under, for instance, the situation of carve outs. If that woman were under managed care from some provider but that person -- that Medicaid plan had sort of said, oh, we are going to carve out substance abuse and as part of the carve out for substance abuse, when the doc calls in and says I need a substance abuse consult here, oh, fine, we will be happy to give you three. That is what they are entitled to. That is it. Bingo.
Those three, if the carve out were actually contributing data, would show up, but in the past, they might have gotten 12.
DR. IEZZONI: George.
MR. VAN AMBURG: Let me be radical here, throw out something for discussion.
Everybody has gone on the assumption that we need an encounter data system and I would like to challenge them to say why. This morning, we looked at this chart. Tennessee submitted 12 million records so far and I haven't seen one item of worthwhile information out of that yet.
The issue to me is that the health care provider has a portion of information needed to answer these questions, that whoever is doing the study would go and get when they need to get it. I mean, think of the hundreds and hundreds and millions of encounter records that they have submitted with no thought to quality, no thought to completeness, just to get the record in because it is a requirement, part of what is going on right now with the fee-for-service things, except that they get a bill attached to it.
And I am not sure state legislators and others are going to believe some of the data coming out of these encounter data systems. When they say, oh, yes, our people are getting good service. We had all these encounters. Well, how do you know? Why do you need an encounter data system?
You need the data available to go do a study if you want to.
MS. GREENBERG: You need the capacity to create the encounters. You don't necessarily need all the encounters --
MR. VAN AMBURG: Yes, but if you look at the Medicaid people and they say, oh, we need an encounter data system. We have to collect the same thing on every visit for the managed care that we do for fee-for-service. We are going to have a phantom billing system. And I am not sure that is the way to go. I am really not. And you also end up having to face that privacy issue, which is going to start linking all these encounters to things, which you wouldn't have to do if you were doing a special study.
It seems to me that the official health information on the patient is in the medical records and that has got to be the source of any analysis that you want to do, not necessarily an encounter because that is not an official record essentially.
MR. GELLMAN: Could I ask a question following up on that?
If you look down the road as information systems develop, whether you have centralized reporting of data, centralized databases, linked databases, network -- all the information is going to be somewhere on a network. So, it is going to be retrievable if you allow it to be retrievable, regardless of who you report it to and whether you have it all compiled in one place or you just -- the compilation comes at the moment of need, may not be a significant difference between one and the other.
It may be a long time before we get there, but --
MR. VAN AMBURG: But that then envisions that everything is in electronic format at the provider, when we are not there yet, a long from there yet, that the maintenance and the holding of that data is the responsibility of the provider and not a government agency and you wouldn't compile it all. You would only compile the things that you wanted to do a study -- for example, you are going to look at access to care. You might want to look at the medical care records of kids who went in the hospitals for asthma, their managed care versus my managed care. That is all you want to look at.
You don't need everything else.
MS. GREENBERG: I think what is really the perspective that is sort of missing in this discussion is -- I am thinking of Kathy Colton, but it is the plan. What does the plan need?
DR. MOR: What does the provider need?
MS. GREENBERG: What does the provider need.
DR. MOR: The provider needs that stuff.
MR. VAN AMBURG: That is right.
DR. MOR: The provider needs that stuff not about patient Q, but the provider needs that about all patients so that the provider can deal with issues of quality, their own internal quality assurance and quality control processes, their own -- they may not do it but they need to do it.
MR. VAN AMBURG: Yes, if you are looking -- I am not sure whether you are talking about an individual personal provider or plan.
DR. MOR: Well, most providers practice in groups now. I am not talking about a plan. I am talking about groups of providers. They need to have that. They need to know, particularly if someone is going to impose upon them capitation. As soon as someone imposes upon you capitation, you can't not have the information. If you don't have the information, you are lost. You will lose money by the end of the month.
MS. WARD: I think that is a key direction because if we keep looking at this as the Federal Government has to have these billions of records to answer the question about health status, I would say we didn't make the decision to go into Medicaid managed care as a Federal Government because we thought it would provide better care. It was because of financing.
Even if we had a billion and one pieces of data and proved two years from now that Medicaid managed care was, in fact, a potential way of reducing the health status, would we, in fact -- would anything happen because we had that data? Would anyone actually act differently because we had all of that?
I mean, I think if we could get the providers to want to collect it because it is meaningful because of her practice and that person is evaluating the practice, we would be a lot farther along than we would be continuing to try to demand that this be reported because it is not going to prove to be of any value because it is just a federal reporting requirement.
DR. MOR: These data are only of value to providers to the extent that there are forces in the environment that either force them to compete on the basis of quality or on the basis of efficiencies, cost efficiencies. Without that, you know, it doesn't -- just leave it all on the paper.
DR. AMARO: Don't you need this data to, you know, answer that question of whether it is cost effective to provide this service? I mean, how do we know if there is a capitated rate that, you know, patients are getting something for that? I mean, how do we track that, you know, unless we are sort of willing to just leave it wide open?
DR. IEZZONI: I am curious, George, how you feel about this notion of doing a special study to look at one of these questions versus having an ongoing kind of monitoring of these issues that would be kind of ongoing as kind of the way that nurses monitor vital signs, for example, so you can find a time when somebody is out of whack, rather than deciding that you are going to do a special study?
MR. VAN AMBURG: Well, I think there is a place for surveillance, ongoing surveillance, for things, but I think a lot of the questions we are asking are not necessarily ongoing surveillance questions.
DR. IEZZONI: Such as which would be ongoing surveillance questions from those that are out there. Which ones would you view as ongoing surveillance?
MR. VAN AMBURG: Well, let's see. Probably the provision of preventive services would be something you would want to monitor continuous on a regular basis. It is more of a process measure than an outcome measure. And I am not sure you will be able to monitor some of this stuff from encounter reports.
DR. IEZZONI: No, I didn't think that you could. Marjorie and I were having a bit of a disagreement about that.
MR. VAN AMBURG: No, I really don't think so.
DR. CARTER-POKRAS: Well, some health outcome monitoring is -- the analysis that has been done by Mary Gornick(?), et cetera, with the Medicare data, they have looked at health outcomes.
DR. IEZZONI: Mortality is really the only outcome I have taken a look at.
MR. VAN AMBURG: There is a good example in Medicare is if you are looking at cancer, you should be looking at a stage of diagnosis. Medicaid can't do that.
MS. GREENBERG: We should have put SEER down as a data source.
MR. VAN AMBURG: You have to link it to a cancer registry or some other source or back to the medical records. That is a special study.
DR. IEZZONI: I think this issue of what do states want to be knowing in a surveillance mode versus as a special study mode is an interesting question.
MR. VAN AMBURG: Well, and it depends on who in the state you are asking, whether you are asking the legislature or the health people.
DR. IEZZONI: Well, but I think that is relevant to what we are doing here because how those -- how that question is answered would have a big impact on what data systems are needed to support the monitoring versus --
MR. VAN AMBURG: But Elizabeth is right. The program is being sold as a cost savings program at the state level. Okay? And you know that you have got the legislatures and the governor's office is going to be interested in how much money they saved and how much isn't being spent. And you have another faction over there saying, well, is the population getting the services.
Those questions are not necessarily being asked right now. Those are the two questions that are going to be the key issues in the political arena and the public health people have a different set of questions.
DR. MOR: Because the state legislature -- it is not so much even saving money is it is actually putting a cap on their annual obligations, so that they are not -- it is not six months into the next year when they know what they spent last year on health care.
MS. WARD: Because they are getting killed by their supplemental --
DR. MOR: I mean, because they have got basically the same open entitlement problems that the feds have got, many of the states have got.
DR. IEZZONI: But is that an ongoing surveillance type of requirement versus --
MR. VAN AMBURG: But you will know that if you are going to have capitation --
DR. MOR: If you have capitation, you know -- the providers are at risk or the plans are at risk. You are paying this much money. That is it, but the public health people want to know what are the potential costs associated with that.
DR. AMARO: I think the public wants to know that, too.
DR. GREENBERG: The public may not care that much about Medicaid managed care. I mean, I don't know that there is a -- I am not saying they shouldn't but I think the public is, obviously, concerned about managed care.
MR. GELLMAN: The broader issue here is you have different audiences at different times for different questions and some of them may want more continuous data and others just want to know now and again and others who don't want to know the answer to some of these questions.
DR. IEZZONI: So, that is another layer of issue here that we have to deal with.
MS. GOLDEN: Some of the questions are being put forth to the public by some researchers. Even though the public may not have initially raised the issue, there is research in these fields and especially in the area of the impact on minority populations. And HCFA, I know, has entertained a number of little grants that have been looking at this, but from what I saw, there is a quantum leap between using the data appropriately and jumping to the conclusions that are being made from the data and it is being put forth to the public. So, there is an indoctrination process that is going on.
So, somewhere to get around -- turn that around might be an appropriate thing. Even though it still may be special studies, there is still that issue.
DR. IEZZONI: Okay. All right. Where do we go from here?
DR. CARTER-POKRAS: I am sorry. There is one other group and it doesn't sound like they are in the world, and that is the folks who are responsible for civil rights monitoring and enforcement because their take on this within the Office of Civil Rights is that even though we haven't come straight out and said that we need to make sure that the individual level encounter data is collected and that the linkage is being performed to make sure that everybody is getting the same quality of care, they say that when it comes down to it, the managed care organizations that we are paying their bills, that they need to be able to demonstrate in the future if there is a question, that they are complying with civil rights laws.
Okay? So, that is like a third piece as far as the need for this kind of --
DR. IEZZONI: Kind of a continuous need, rather than a special study sort of need.
[Multiple discussions.]
MR. VAN AMBURG: But you are saying that plans have to be able to demonstrate that. The individual plans have to be able to demonstrate that. It means they have to have the data available in the plans.
DR. CARTER-POKRAS: The Department doesn't specify what you have to be collecting and how you have to have it handled, that you just need to be able to demonstrate that you are complying with the civil rights laws.
DR. MOR: When I was talking about encounter data, it doesn't necessarily have to have -- be archived by the Federal Government. That is a whole other set of issues.
MR. VAN AMBURG: I understand that and I agree --
DR. MOR: I am not even sure it has to be archived by the state government.
MR. VAN AMBURG: No. I agree with that, too, but the way it is being used this morning was an encounter data system being handled by the state and the Federal Government.
DR. MOR: That was in the context of those demonstration programs. What they were going to mandate for the local plans -- so, what the dually eligible -- the reason they were getting -- they have gotten almost no data out of most of those is because those are state plans that are supposed to -- state operated plans, that are supposed to be giving money to the -- data to the feds as part of the evaluation.
I don't know in all cases -- for instance, in Massachusetts under the plan that they have just submitted as a waiver for the dually eligible, there were very detailed specs that are written up for what kind of information the plans had to maintain, but it wasn't clear that those data, the micro level data, were to be passed on to the state. At least I didn't see that part of it. But they had to be able to generate reports to assure the state.
MR. VAN AMBURG: Right. Now, in Michigan, which is a 1915(b) waiver, which wasn't on the list of required data, they are planning a massive encounter data system. Just automatically assumed that they needed that.
DR. MOR: For the state?
MR. VAN AMBURG: Yes.
DR. MOR: And the state is going to get it?
MR. VAN AMBURG: Yes.
DR. MOR: Aren't they putting this out to bid to plans?
DR. MOR: Well, the plans are going to have to submit the data. They will probably put it out to bid to someone to contract to do the processing maybe. Who knows? But it is that thinking that bothers me more than anything else.
MS. WARD: I think that is a very good way of trying to cut because I am the kind of person, who sits and loves to think about the data I can get my hands on. I am normally also drooling over that kind of stuff, but I am becoming less interested in that kind of warehousing of data because it is not particularly making -- we are not getting anywhere with all that. I think if we could take a different tact, matching up questions to benefits, to where does it have to be collected and housed and kept as a different solution to always assuming that we get from here to these massive state and federal systems with boring, billions and billions of data.
DR. IEZZONI: Let's take that exact framework, which was beautifully articulated as the jumping point for our discussion right now of what we want our contractor to do. Okay? Because it is obvious that there are a lot of questions out there, but we only have two people who are actually going to be paid from the room, who have been helping us think about it. It is obvious that there are a lot of populations we are interested in and there is some disagreement in the room about what data might be appropriate or needed.
MR. VAN AMBURG: It is how it is collected and housed more than what the data are.
DR. IEZZONI: So, as we begin to think about what we might have a contractor do, knowing a little bit from this morning about what some contractors are already doing with managed care on the demonstration projects, with multimillions of dollars. Yes, these $6 million evaluation contracts.
But I agree, the six million probably doesn't go that far once you start talking about five states. No, it doesn't.
What would people find most useful in terms of some kind of very smart, busy, well-informed people who could help us with this? What would you want them to do? What might be a way that we could us a contractor most productively in this?
MR. VAN AMBURG: Well, don't you think we ought to decide what product we want first?
DR. IEZZONI: All right. Let's start there then.
MR. VAN AMBURG: I mean, it would seem to me that we would want a product and then we would want the contractor after we get to that product.
DR. IEZZONI: Well, I will just say that I had envisioned the product being a report that would go through what are the concerns about Medicaid managed care, probably from the perspective of states to the extent that we can have that and then look at what data systems might be appropriate or necessary to be able to answer those questions and then looking at what is going on in the country around that to see whether we, in fact, have the data systems that will allow us to answer the questions.
DR. AMARO: Going back to our charge, it seemed that we said we were doing something a little bit beyond it, which is really then making recommendations about the kind of systems that would have to be put in place.
DR. IEZZONI: Right. That would be the logical kind of conclusion to that, the inadequacies, what the gaps are, you know.
DR. MOR: Well, let's think about this from the point of view of the presentations this morning. They made a number of recommendations that if we were to do something that would be particularly useful, some of it might be, you know, grapple with the issue of, you know, how are states dealing with this issue of integrating information across multiple systems. That was one suggestion.
The other was is there --
DR. IEZZONI: ASPI has --
DR. MOR: Okay. All right. But somebody else suggested that we could be helpful if we thought of -- I mean, in some sense like model contracting specs for what a state would require of their managed care plans if they were going to go ahead and proceed with this so that the plan could generate information of use to the state and of use to itself.
DR. IEZZONI: That is why you have to first define what the state wants to know, to know whether it is of use to them. You need to know what information they have and then you can begin to think, okay, what would be the specs or the ideal information system based on the gaps.
DR. MOR: Maybe another way to think about it is to say here is what the states' questions are and here is what questions of managed care companies might be, who are in this business. How would you actually present the results? I mean, how do you turn those data into information that would be usable and helpful, into some kind of reporting form?
MR. GELLMAN: I am not sure -- this is going to be argued from ignorance here since that is what I bring to the table on a lot of this stuff. But I am not sure I understand what the ultimate goal is. I mean, from a much broader perspective, what is it you are trying to accomplish? I mean, there are all these great questions and whatever, lots of data that can be manipulated all different kinds of ways to produce all different kinds of information.
Question: Ultimately, what is the goal of all of this? Are you trying to use the information for political purposes? Are you trying to use the information to create a way for the system to correct itself and make sure that it is accomplishing its goals? Are you trying to do this to provide -- what in a broad sense is the goal?
DR. IEZZONI: I think the national committee has historically been involved in highlighting areas where there is inadequate information to answer questions. There is a long history of the national committee having reports such as that, that have, I think, been useful in -- I mean, Marjorie, you can leave them here -- in, you know, highlighting this and beginning to get people to think about how to improve the data that are available.
MR. GELLMAN: The data for what --
[Multiple discussions.]
DR. AMARO: Can I give it a try? I think that, you know, if we go back to our charge and our work plan, there were a variety of populations that this committee is
-- our purpose is to look at how adequate is the information that is out there and how can the systems of data-gathering be improved to give us information on those systems -- on those populations. So, one of the key dynamics or issues that is going on that is affecting these populations is Medicaid managed care.
We said this is a big event going on. We don't know much about how the health of these populations is going to be affected and we need to think about, well, what data systems are in place that may help us start to be able to look at that. I think that is how we first got here.
It seems to me -- I was going to suggest that we heard it about three, four, five states and they are under the RWJ funding for integrated data approaches -- it seems to me that a good place to get started is to look at what have South Carolina, Maryland, Illinois, Iowa, Vermont done to -- they have been in the forefront in this and what has been the progress under this RWJ effort because it seems that they have been doing work for while and that if we looked at that, we might be able to think about some things that might be useful more broadly.
MS. WARD: I think what ASPI has done is an inventory of who is trying to do large integration. What has not been asked is for those -- are there a couple of states who are -- maybe South Carolina is the one to look at and say what is it costing them to do it and what is the value and have they, in fact, been able out of those billions of records answered the question. Has the health status of the people in Medicaid managed care gotten better or worse?
DR. IEZZONI: Yes, I had thought that we should do a couple of case studies that would do exactly what Elizabeth --
MS. WARD: And maybe Illinois is the other state, but from that inventory of states that have done sort of this whole massive -- might replicate what would be -- what we have already stated as a committee would be something like the encounter record, which this committee has already put enough work into recommending. I don't think we need to go back over that.
DR. IEZZONI: But I think that these do need to
be --
MS. WARD: -- electronically available.
DR. IEZZONI: These do need to be states that have a heavy penetration of Medicaid managed care and that have diverse populations. Like I would argue against Vermont, for example, because --
[Multiple discussions.]
Yes, South Carolina is low on Medicaid managed care. So, that would be one argument against going to South Carolina.
MS. WARD: But if they are using their data to answer the question, has the health status changed from collecting encounter data, I don't think you -- I don't know what different would it make if that encounter data were from a regular Medicaid or a managed Medicaid?
DR. GREENBERG: Well, just thinking of the whole issue of has health status changed, first of all, Medicaid data has been a problem from the get-go, forget about managed care. I know that we are making some progress. At the national level, we have never had Medicaid data. There have now been some states that we can, but do these states even have baseline? I don't know -- the ones that are going into Medicaid managed care.
The other problem is the eligibility thing, the going in and out and how you can even make a cause and effect is very difficult because of -- some of this has been addressed, I think, for pregnant moms, but we talked already today about services and -- they only go for a certain point and your eligibility is connected to all sorts of things.
So, I mean, I think this is a very complicated type of thing to be able to do in Medicaid, in particular, because of the lack of good data systems in Medicaid and also the lack of continuity.
I am not saying it is not worth -- it is the most important question, obviously, but -- or even maybe not change so much for the same -- but looking at can you look at people who are in fee-for-service and people who are in the managed care and then make some comparisons about do they have the same access to services, are they getting the preventive services, are they getting the mental health or substance abuse or whatever.
But as to the impact on the health, I mean, it really -- it requires (a) baseline data and (b) continuity and I don't know if either one exists very much.
DR. IEZZONI: Pat, you had a comment.
MS. GOLDEN: Somewhere in there is the need to know whether these systems can begin to answer such a question and to evaluate what data would be needed to answer a question and then, at least, give us some sense of the potential baseline database system by including -- which is sort of evaluating the data as opposed to the health status of the population.
MS. WARD: Why do we want the data?
MS. GOLDEN: Why do we want it and what are we trying to learn from the data and do the systems allow us to answer these questions? If not, what are the limitations --
DR. IEZZONI: Let's go back to the first -- the four bullets underneath what we thought was kind of the charge. Identify questions about health and health care services delivery. I mean, can we formulate kind of a product from this? Okay. I think -- Hortensia went back to this, too. I think that people feel comfortable with that, but let's get back to what we need to know to kind of flesh out these bullets and whether a contractor can help us.
MS. GOLDEN: -- would be to identify the major concerns. Either we need to identify them or someone has to identify them, but until that can happen on its own, then you can't evaluate whether the data answers that.
[Multiple discussions.]
DR. MOR: The first three of these are, you know
-- what questions do states have -- instead of what are the questions in general, what questions do states have, do this state have, what data elements are needed to answer those questions because that could vary from state to state? What data systems exist -- do the data exist?
MS. GOLDEN: And I think that is important. Do those questions vary from state to state based on some kind of legislation in a particular state that may not make one piece of information usable or more needed in the states and I think that is important.
DR. MOR: So, do we rely on a contractor to --
DR. IEZZONI: No, no, no, we have to hear from staff from that. Marjorie and Lynnette, have you been listening to the last three minutes of the conversation, where we seem to be heading towards a plan?
What do you need from us to be able to begin to put together a task order? We have this, which may be too general --
DR. GREENBERG: By "this," meaning what is in
your --
DR. IEZZONI: The three bullets.
I mean, if we want to find out what are the questions that states are asking about their Medicaid managed care, which is, I think, one of the those -- the first thing -- I mean, is there a survey that somebody could do of states?
DR. GREENBERG: When you start talking survey, I get kind of nervous because --
DR. MOR: Site visits. Site visits to x number of states. The number of states selected is going to be a function of how much money you have. Basically, you want to do the same thing in multiple states and really it is a question of what states they should be.
MS. GOLDEN: Could that be part of what we ask the contractor to initially do, to evaluate which states ought to be -- and come up with what you are trying to shoot -- you want, obviously, diversity. You might want difference in their -- some diversity in plans or in --
DR. IEZZONI: Well, we heard this morning that every plan is unique.
MS. GOLDEN: Of that, every state is unique. But there would be maybe enough similarities in -- they could be grouped and maybe that kind of evaluation would be the first step of the contractor, to come up with a rationale and a plan for selecting these things.
DR. MOR: What is the price per state? What do you think?
DR. IEZZONI: Well, we have to be modest. This cannot be like the HCFA Mathematica, Urban Institute contract.
[Multiple discussions.]
DR. MOR: What are we talking about?
DR. GREENBERG: I am not sure. I mean, we have had a little internal discussion or e-mail discussion and it may be -- I think it is really -- you have to decide what you want, what you feel would be most useful and kind of go from there, I guess, but it could be -- it could be that you want a study, which would probably be valuable, but of a magnitude that there are not currently resources associated with -- available to do that. I mean, we have been -- you know, we, in our current ability to put, you know, 5,000, 10,000, 15,000 -- I mean, you may be talking about another
-- a project, but the timing isn't bad because the Department is beginning to look at the use of evaluation funds for next year, for the next fiscal year.
[Multiple discussions.]
MS. RIMES: I have been recommending that we do a stage -- you do something and then decide what you want to do next.
DR. GREENBERG: Right. I don't think you are really ready to go for a --
[Multiple discussions.]
DR. MOR: If Mathematica does it, one state is going to cost $10,000 guaranteed.
DR. IEZZONI: How about the third bullet? Because that is not as -- that may not be as labor intensive, frankly, as like the first bullet because there might be some data sets already or kind of -- Vince, this is the question that you were getting at, was the information on the applications codified, you know, this morning when you were asking about --
DR. MOR: Yes, yes.
DR. IEZZONI: So, one of the contractors that HCFA already has might have a good sense of the variety of managed care efforts around the country and, you know, the extent to which they are conducting evaluations.
DR. MOR: But also you wanted to get at do the data systems that they have -- and the problem is without the questions from the state, it is tough to know how good the data systems are.
DR. IEZZONI: No, I know. I am just trying to --
because I think finding out what the questions are is going to be very expensive because it is going to take people going and interviewing and stakeholders. So, what I am trying to think about is what would be the cheapest thing that we could do first to get ourselves started, that we could leverage to begin to think about, okay, what are the states that we should go to to ask the stakeholders about what their questions are.
MS. RIMES: Can I just ask probably an insane question? Is there a possibility that even before we talk about contracting and doing something which is a pretty defined task, that there might be a need to pull some states in and pull some contractors in and have like a whole day's session, where we start asking some of these questions and start formulating? Because I am listening to you all and I am listening to what my own thoughts are at the same time.
I think we had a number of people that were recommended to us --
[Multiple discussions.]
DR. MOR: -- is apparently the guru of all gurus.
MS. RIMES: She knows exactly what is going on --
[Multiple discussions.]
But she would come in for this, is my understanding, as would Margot Rosenbaum(?). And there are a number of people out there that have already done, I think, a number of things.
[Multiple discussions.]
DR. IEZZONI: That is fine. We are going to have to find a time to do that. Does it have to be a full day? Could it be a half a day because --
[Multiple discussions.]
DR. GREENBERG: You picked the 29th and 30th of September.
MS. ARAKI: Yes, the 29th and 30th of September, we could change our agenda for that because we were originally going to do on the 29th and 30th the territories.
MS. GOLDEN: I will probably have the question about this territory issue.
DR. IEZZONI: Well, the territories is a much smaller effort, hopefully. We weren't going to have a contractor for that. So, I am wondering, would you be terribly disappointed if we substituted maybe doing this activity --
MS. GOLDEN: No, I won't be disappointed because, again, unless you have already discussed it, I really need to get some clarification of just what we were trying to get out of even that effort. I was given the names of a couple of people -- I am jumping ahead -- to contact, but I thought it was a little premature without having some --
DR. IEZZONI: Well, we need to talk.
MS. GOLDEN: So, no, I will not be terribly disappointed.
DR. IEZZONI: Okay. That is what I wanted to establish.
So, perhaps what we could do -- the way that we had it originally structured for, I guess -- what is it, the 29th and 30th or something? -- where we are going to do a full day of hearing from people and then a half a day us just sitting around and talking about it.
So, that would probably work very well with this notion of bringing in some of the HCFA contractors who know a lot about this process and then us the next day hashing through what we want. Does that sound like a plan?
MS. WARD: Yes. And think of all your bullets that give them some sort of concentration -- I think No. 3 on that page, gives --
DR. MOR: Or actually if you are going to have the other contractors in, if you can have both Mathematica and Urban in, they talked to all of these folks and they could probably actually give you a summary of what the states questions are from their interpretation, as well as probably direct quotes.
MS. WARD: And whether they think they have got the data to answer their questions.
DR. MOR: The structural questions -- the testimony around those three questions.
MS. RIMES: There is that and the one point that you made, Vince -- there is now the longer term kinds of things that need to be looked at that I know no one is thinking about.
DR. MOR: Longer in terms of how plans, how providers are managing themselves?
MS. RIMES: Or evaluating certain aspects or however it breaks out.
DR. MOR: I could tell you the market is certainly interested in how good that information on these managed care companies are.
MS. RIMES: Yes, but there are a lot of issues that you brought out and there are a lot of issues that are coming around that are more than just surveillance monitoring and no one is addressing them that I know of.
DR. IEZZONI: Why don't we do this: Lynnette, I would like to kind of get a summary of what we have kind of thought about today, just to remind us of where we are coming from. I think that it is a good idea to talk to these potential contractors or the people that we would bring in about the first three bullets here, but that we should also be thinking -- we have a half a day or a few hours the day before the full committee meeting in September, don't we?
PARTICIPANT: I don't think so. There is a breakout.
DR. IEZZONI: A breakout. That is what it was. The breakout is what I meant. Okay.
We have a breakout session. How long was that -- but we were going to be doing something different during that.
DR. GREENBERG: We haven't formalized the agenda yet, so if the subcommittee -- in fact, that was going to be one of the things I wanted to ask you all was if there was anything out of this, out of any of your work plan that you would like the full committee -- that you think a presentation at the full committee would be appropriate.
DR. IEZZONI: Okay. I think it is premature for a full committee presentation, don't you guys think so?
DR. GREENBERG: Oh, certainly from the -- you know, if there was anything you wanted --
DR. MOR: How were we planning to use the time for the breakout?
DR. IEZZONI: We were planning to look at the Bureau of the Census, the Social Security Administration and -- because they are about to conduct some big studies of disability. And we were going to look at how they were defining "disability" and how that compares with how, for example, the NHIS and NHANES defines "disability."
I would still like to do that during that breakout.
DR. MOR: The one comment I have, and I forgot to actually make it on the charge, is that HCFA, the current beneficiary survey also has a definition of "disability."
DR. GREENBERG: Oh, there are like a hundred definitions.
PARTICIPANT: Each survey has a different --
DR. MOR: But to have that on the plate as well.
[Multiple discussions.]
DR. IEZZONI: I know, we are generating a lot of hot air.
Okay. I think we should take a break. Olivia has left. We need to find her because I would like her to be in the room for the discussion of the OMB Directive 15.
So, this has been, at least, I think, interesting. I hope that you have all found it a useful way to get started to get into this issue.
MS. ARAKI: Just one point of clarification.
DR. IEZZONI: Yes.
MS. ARAKI: You want actually the contractors -- I mean, for this meeting here, you want some of these questions asked -- you want these questions posed to the --
DR. IEZZONI: No. What I would like to do --
MS. ARAKI: Do you want to talk about what questions you want to pose to them?
DR. IEZZONI: These.
DR. GREENBERG: The questions in --
MS. ARAKI: Oh, in here.
DR. IEZZONI: The three bullets.
DR. GREENBERG: Obviously, we are transcribing -- we can have the person who does the summary or minutes, we can have these -- we can probably type these up and give them to her so she can incorporate them.
DR. IEZZONI: That would be good because I think it would be good for us to remember where we have come from because questions, for example, about the impact on minority providers may or may not come up from some of these contractors, you know. And so I think that it is good for us to keep in the background the kind of lists of things that we have generated today, just to remind us of the full perspective that we wanted to address.
Okay. So, shall we take a break then until 3:00? Why don't we reconvene at 3 o'clock.
[Brief recess.]
DR. IEZZONI: All right. Let's put some dates on our calendars. How about the week of the 12th, January 13th and 14th? How does that sound?
DR. MOR: If it is Tuesday and Wednesday, I think there is a -- I have to be back by noon. I have a department chair meeting.
DR. IEZZONI: Okay. Monday and Tuesday?
DR. MOR: That is Monday afternoon and Tuesday?
DR. IEZZONI: Monday afternoon and Tuesday.
DR. GREENBERG: Monday afternoon and all day Tuesday.
DR. IEZZONI: Yes. Okay. Is that -- those are January 12th and 13th. We will take off the month of February. Now, we might want to actually go to a state, such as South Carolina or Iowa or even I was thinking Massachusetts.
DR. MOR: I like that a lot.
DR. IEZZONI: Yes, don't you like that one. We should maybe book a time in April --
DR. MOR: Actually Massachusetts would be very good -- I am serious -- for the Medicaid managed care.
DR. IEZZONI: Yes, that is what I was -- I think it actually is a good fit. I didn't want to appear too parochial in suggesting it, but --
DR. MOR: No. Actually there are all kinds of things going on in Massachusetts now where we can get a lot of -- we can get consumers to come in an testify.
DR. IEZZONI: So, why don't we book a time in April that we would do -- can we just say that we would do it in Boston?
DR. MOR: You aren't going to get any argument out of me.
DR. IEZZONI: Does that sound okay? All right.
How about -- oh, gosh, we have April 15th right in the middle of it, before or after April 15th. End of April, like Tuesday and Wednesday, the 21st and 22nd?
DR. GREENBERG: That is the third week in April.
DR. IEZZONI: Yes. Is that okay? Vince?
DR. MOR: Wednesdays, I have standing -- you know, middle of the day, but I will be there until then. I can go up in the morning and be back by noon.
DR. IEZZONI: Right. I know. That is what is so great about Boston.
DR. GREENBERG: So, Wednesdays are bad days for you?
DR. MOR: Yes.
DR. IEZZONI: Yes. I just hate to have poor Elizabeth traveling on Sunday, you know. I just think it is so unfair to her.
DR. MOR: That is fine. I will be there all day Tuesday and then can --
DR. IEZZONI: Okay. So, the 21st and 22nd, we will do Boston.
DR. GREENBERG: April. Okay. That is -- and there is nothing -- there was January and the March will be probably be a full committee meeting, but I don't think we have the dates for that. We need to get them.
DR. IEZZONI: We are probably going to want to do one other site visit. May -- see, this is when the meetings are beginning to start and it becomes awful, doesn't it, you know?
DR. CARTER-POKRAS: What about March and April?
DR. IEZZONI: March and April? Because we will have a full committee meeting in March probably. We could go back to February and think about doing a site visit in a southwestern state or a southern state.
DR. MOR: Puerto Rico.
DR. IEZZONI: No, no, they are not doing Medicaid managed care.
DR. MOR: The only place would be Arizona that is doing much now.
DR. GREENBERG: They have collected encounter data for about -- for a long time.
DR. IEZZONI: Do they have a heavy managed care penetration?
DR. MOR: Yes, it is all -- total. Mandatory.
DR. IEZZONI: I think Arizona needs to be it then.
DR. CARTER-POKRAS: Are you going to the data users conference next week?
DR. IEZZONI: No.
DR. CARTER-POKRAS: Oh, okay, because they are going to have presentations at our session, implications in managed care.
DR. IEZZONI: Oh, that one. Oh, yes, I am heading a session. That is right.
[Multiple discussions.]
DR. CARTER-POKRAS: Arizona is going to be coming and talking about their Medicaid managed care.
DR. IEZZONI: Okay. Why don't we plan on Arizona and Massachussetts being our two states then. How do people like the way that we have just chosen that? Does that sound okay?
DR. GREENBERG: Well, you may find out when you have the contractors come in that there is a better --
DR. MOR: Yes, don't have -- this is not the right thing for the contractors.
DR. IEZZONI: Let's put a place holder in February for a site visit for the committee because I think that Massachusetts, though, is a good place, Vince, don't you think?
DR. MOR: Yes, Massachusetts, I have no argument with.
DR. GREENBERG: I don't know if you want to go to Massachusetts in February, though.
DR. IEZZONI: No, but what I am saying is let's put a place holder in for a site visit in a place where, hopefully, you don't have the same weather problems that you do in Boston, but you might. It might be Iowa. Okay?
February 9th and 10th, assuming, Elizabeth, that it won't be as bad a travel for you because -- that is a Monday and Tuesday.
[Multiple discussions.]
When is President's Day?
DR. GREENBERG: Maybe -- is it the third Monday? I don't know.
DR. IEZZONI: It is not on my calendar. George, you look like you have an official calendar. When is President's Day?
MR. VAN AMBURG: Last year it was on the 17th and -- this year it was on the 17th.
DR. IEZZONI: So, it would be the 16th. Okay. So, the 9th and 10th of February. Let's put a place holder for a site visit. Come on, you guys.
[Multiple discussions.]
So, a site visit for that day. Then we have a full committee meeting in March and we have a site visit in Boston in April. Then we are not going to do anything in May. We will have a full committee meeting in June. And then shall we do -- try to -- like a wrap-up type of meeting in July, a year from now? Yes?
DR. GREENBERG: You mean, wrapping up the Medicaid managed care?
DR. IEZZONI: Well, where we are at the time.
DR. GREENBERG: -- report out to the full committee in September, you probably need a summer meeting.
DR. IEZZONI: And July is better than August. So, how about July 14th and 15th? Vince, you can't have a faculty meeting.
DR. MOR: No, no, no. I am just sort of --
DR. IEZZONI: 1998, July 14th and 15th.
PARTICIPANT: That is a Tuesday?
DR. IEZZONI: Tuesday and Wednesday and it will be a day and a half.
Okay. That was incredibly painful. It probably actually wasn't as painful for you all. You might be really bored. It was painful for us. But thank you everybody and thanks to everybody -- and we really need to start talking about OMB Directive 15.
Is everybody okay with this, please? Yes? Okay. Great.
Hortensia, why don't you lead our discussion.
DR. AMARO: Well, I was going to ask Olivia as well, because she really knows this inside and out since she was part of the writing of this, but as Lisa mentioned, we recently received this -- I guess you got it when we got it, which was around the 15th or the 16th and if people have had a chance to read it, we would like to get some input from you so that when we prepare our comments, we can take those into consideration.
I think we will definitely have to get input from you as well after we write our initial comment. The summary of the recommendations are in the last section, Section 6, which start on page 36937. Do you want us to go one at a time on those or --
DR. IEZZONI: Well, Hortensia, let me just you a more general question. On page 36876, it is the third page of this document, they ask -- they say issues for comment and then in the second column, they say that they would like -- you know, that here are some examples of questions that we would like to hear from responders about. Were you planning or thinking about organizing your comments around these kind of areas?
DR. AMARO: Well, we haven't had a chance to talk, but when I read it, I thought that this was a natural way to respond, both in terms of the general principles that they discuss and then the -- you know, the recommendations using the general principles.
DR. IEZZONI: Because some of -- what the second question is did this research use sound methodological approaches, you know, and I just don't know that -- I mean, that will take a lot of work to think through.
DR. CARTER-POKRAS: Actually, these are suggested questions. They don't have to be answered. People's gut reactions to the recommendations or that I don't like this because -- you know, I don't think it will work for my population and then give the rationale. That is still very useful information.
DR. IEZZONI: I guess what was wondering was whether Hortensia envisioned herself doing a methodological review as a scientist of how the studies were conducted.
DR. AMARO: No, not really.
DR. IEZZONI: Okay. I just wanted to --
DR. AMARO: And I think in addition to the issues they are interested in that are a particular perspective in answering the questions would be from the health data perspective because the use of these categories is going to be broader. So, I think we would want to focus our comments, you know, that way.
DR. IEZZONI: That is a good suggestion.
Well, the comment that you and I were actually beginning to talk about before the meeting started was that I am a little bit concerned about the Middle Eastern category and they say in chapter -- I actually read the detail about the Middle Eastern study that they did. I think it was in Section 4 or 5. There is actually a nice table of contents for people who haven't found it, starting on page 26878.
You know, they basically said that they hadn't done enough research to be able to make a reasonable suggestion, which is probably true. What I would propose is that they do more research on that particular category. Obviously, they won't be able to implement that for the year 2000 census, but I think that it is going to be a bigger issue than it has been heretofore because I think the population from the Middle East is expanding, especially in certain parts of the country.
George, you had mentioned that that is true in Michigan.
MR. VAN AMBURG: Right.
DR. IEZZONI: So, I think that that is one suggestion.
DR. GREENBERG: Would that be in terms of a race classification or as an ethnicity classification?
DR. AMARO: Ethnicity. They considered that they are not recommending it, but they recognized that they need more research. I think that is a reasonable suggestion is to recommend that they do more research so they can get a better sense of how it will affect --
You might want to mention some of the history, Olivia, that you shared with me about the South and Central and American Indians and the new recommendation regarding them.
DR. CARTER-POKRAS: That is another issue that really wasn't researched specifically, but they did look at responses from the 1990 census to see how many people from Central and South America, looking at birthplace, for instance, or their specific ethnic origin that they had written in, how many of them would have selected American Indian, trying to get some feel for that.
But the decision was mostly driven by the fact that the current OMB director said everybody in North America, the original peoples of North America, and so that means if you were north of the border in Canada, then you were included as American Indian, but south of the border in Mexico, you were not. So, we felt that that was sort of an arbitrary decision, especially considering NAFTA. So, that was the main reason why they went with that, not because of research that was conducted.
It looks like it is not a huge number of people that will be affected. But, of course, these are just recommendations where people may have strong reasons why they don't like that.
DR. AMARO: Do any of the committee members have any reaction to any of these that you want to let us --
DR. MOR: My biggest reaction and I have looked at some of the data and I keep trying -- I have become so habituated to the one and only slot, separate question for race, separate question for ethnicity and have a sense from my own data how that works, you know, that changing it will create complications.
DR. IEZZONI: They don't recommend that, though. They recommend two questions.
[Multiple discussions.]
DR. CARTER-POKRAS: -- and actually there is still the option of having a single question that currently is used by many of the reporting agencies. That option is still possible. For instance, the Office of Civil Rights tend to use one question rather than two separate questions.
DR. GREENBERG: [Comment off microphone.]
DR. CARTER-POKRAS: The biggest change is to allow people to mark more than once.
DR. MOR: That is what I am thinking about.
[Multiple discussions.]
I actually appreciate it and really thought that the data on the question order made a lot of sense. That was great. I liked that. But I just, you know, I am somewhat uncomfortable with check all that apply to --
DR. AMARO: What are we going to do with it was my question.
DR. MOR: Yes. Like, you know, so what? And I was like --
DR. CARTER-POKRAS: The surprise is that people have been doing this even when they have been told to mark only one. Around 1 percent have been checking more than one even with those instructions. That was a surprise to us.
DR. GREENBERG: Are you more comfortable than you would be if they had recommended a multiracial category?
DR. MOR: Much more comfortable than with a multiracial -- multiracial is just, you know --
DR. GREENBERG: Well, that is, I think -- that is how I think a lot of people felt, that this was a good compromise.
DR. IEZZONI: I actually think it is a very reasonable compromise.
DR. MOR: I do, too.
DR. IEZZONI: I am quite comfortable with it.
DR. CARTER-POKRAS: We won't know what the implications of the teams are until the changes -- we start to make the changes. There will be a work group that is going to be convening fairly soon that the Bureau of Labor Statistics is going to be chairing and they are going to look at different permutations, combinations that you can use at this check all that apply to see how we actually present that information.
MS. GOLDEN: I think Hortensia's comment probably needs to go forward -- and that is to make sure that some of this is really evaluated in a health context. Again, as I was saying, it is one thing to enumerate a population and to report on it demographically. It is another thing to use that variable as a part of an analytical construct as we do in a health audit and maybe some of the editing that is ultimately going to be worked out really needs to be looked at in terms of what we may have already seen from a health data analysis or at least it needs to be evaluated in that context.
There is a big difference in the methodology for using the information that is selected and in terms of how we edit it and code it. That is really going to be where we have a real problem, both for continuity and for comparability. If we continue, for example, putting -- using the first listed, making some assumption that people will sometimes list more than one race, have some order, to which they list it and some consistent order, if coding were to change and come up with some other methodology, you still don't have continuity with previous data analysis because it is lost in the coding process.
DR. AMARO: Has there been any discussion on how this is going to be handled in analysis for the health data systems?
MS. GOLDEN: We don't intend, I don't think, to interpret that to mean that we are going to use that as a reporting variable. I don't think we are required to use that as a tabulating variable. We are required to report methodology -- at least we are trying to assume -- and that is going to be NCHS's question, if reporting the number who give us a multiracial designation in a methodological -- in our technical appendix will satisfy what the OMB directive says as opposed to using it in a table.
DR. MOR: But otherwise in a table would it add up to more than a hundred percent?
MS. GOLDEN: These are issues that we -- well, NCHS will be raising these kinds of issues when it comments.
DR. CARTER-POKRAS: But the work group is going to be grappling with these issues. They started taking a stab at this. If you look at the chapter on multiracial, there are, I think, four different approaches that they took a look at. They said all inclusive is one of them and then they called it -- EEO edit, I think was how they called another one that they were suggesting might be useful for civil rights monitoring and enforcement, but they looked at several different approaches, four different approaches, to try to get a little bit at how you could present that information on marking more than one.
Actually, there have been other organizations that have used check all that apply data and we need to look at them also to see how they have used that information. The Federal Government is a little behind as far as grappling with this particular issue.
MS. GOLDEN: The denominator that we get from Census, of course, will be an issue and what Census does and how they put together the data and how we use it, these, again, will be the questions that we will be raising either for commentary or clarification when NCHS develops its response to OMB.
DR. AMARO: I think those are the most -- sort of the biggest changes. All the other ones have to do with sort of including a number of different alternatives for what term we use to refer to different groups and sort of adding on to the list so more people feel included and don't check "other" and there were some decisions made around -- well, we talked about the Middle Eastern, also the Cape Verdian(?), not including them at this time, although leaving it open for states to provide that detail if they have the populations.
[Multiple discussions.]
DR. MOR: We are the epicenter, as it were.
DR. IEZZONI: Okay. Hortensia and Olivia, you are going to be gone the whole month of August and the letter is due September 8th.
DR. AMARO: We will work it out. I may be available -- you know, we may be able to fax or Federal Express things to each other, but I would prefer to at least get, you know, some of the major stuff down before --
DR. CARTER-POKRAS: We need to get something out so the members have something to comment on. Sometimes that is easier to kind of formulate what you -- the comments you want to make when you have got something on paper.
DR. IEZZONI: Keep in mind, this letter has to come from the full committee, not the subcommittee. So, this expands. I would suggest actually, given that we have this timing situation, that you not go through an initial screen of what does the subcommittee think, but that you go to the full committee with the first round. Okay? And we will give our comments --
DR. GREENBERG: The only thing is, I think, the full committee would be looking to the subcommittee to have given it the most attention. We did send it to everybody.
DR. IEZZONI: Right. I mean, I am certainly around and I am happy to help, but I just -- you know, and maybe a number of us are, but I am just very concerned about the timing here.
DR. AMARO: I would think the full committee would feel more comfortable if it felt that we had looked at it closely and it sort of had the stamp of approval and --
DR. IEZZONI: Well, okay. Who is on vacation? When are people on vacation on the subcommittee?
DR. MOR: I am around.
DR. IEZZONI: You are around the whole summer.
All right. So, if we can have something the first week in August, by the first week in August, and then, Olivia, will you be kind of deputized by Hortensia to take that and revise it and then submit it to the full committee?
DR. MOR: Hortensia, can you give us an idea of like what is the bottom line of the comments going to be?
DR. AMARO: Oh, well, we need to discuss it. I feel that the recommendations are very reasonable. I think that there are areas to suggest where further research is needed and also specifically looking at the implications of how we are going to handle things, like the multiple -- you know, in health data. I think that is the most important one for us that I can think of right now, in terms of the implications, how we are going to handle it in the different health related data system in terms of analysis and reporting.
DR. GREENBERG: You might even want to say -- I mean, cc the letter -- well, you would anyway cc the letter to the data consult chairs because they are the ones who really report to, although it is perfectly appropriate to send this directly to OMB, and suggest that the committee or this -- who of this subcommittee would like to be involved as decisions are made about how -- you know, how the issue is dealt with.
DR. IEZZONI: Olivia, can I just ask, is this consistent with where David Williams was taking the minority subcommittee in his prior work on this issue? Because I know that David --
DR. CARTER-POKRAS: Right. As far as developing a response definitely. As far as the specific comments, I would have to go back and look at the earlier comments to see if the recommendations are in opposition to the recommendations.
DR. IEZZONI: Because I think that we should at least be aware of what NCVHS said previously because if we do have any significant change --
DR. CARTER-POKRAS: They definitely suggested against having stand-alone multiracial category. That I remember.
DR. GREENBERG: The main concern was about that, I think. I don't know that they were real definitive on other things, but I know there was real concern about -- we can pull that letter, but I know that was the main concern.
DR. AMARO: I really can't think of any other one that is going to have as much consequence as that one. I think that is really the biggest one.
DR. CARTER-POKRAS: As far as the Department is concerned, about half of our data systems use a single question for race and ethnicity and half of them use two separate questions. So, that will be a major change if we have to go to two separate questions.
I think most of NCHS uses two separate questions, but that is not the case for the rest of the Department.
[Multiple discussions.]
MS. GOLDEN: My question about the one question, Hortensia and Olivia, has to do with then what would be suggesting with respect to the subpopulations?
DR. CARTER-POKRAS: Well, actually, that is -- we can make a stronger recommendation because what is in the back there recommends that subpopulation data be collected and we can reinforce previous memos that the National Committee sent to the Secretary recommending that subpopulation data be collected. So, that could be part
of --
DR. MOR: In making any recommendations about collecting subpopulation data, it is really imperative that that data be collected imbedded within the appropriate context so that it does not affect the distribution of responses to those first two questions because it is very easy to affect the distribution of responses to those questions, based on the context as the data show about question order.
MS. GOLDEN: That is understood but, again -- and maybe that is additional wording that needs to be there, but the problem is on the one hand we are pushing and everything we are saying is to get more subpopulation data. If you encourage the collection of it in the single format, you have done nothing to really beef up the issue of the subpopulation data.
DR. MOR: When we were in California, we heard testimony of the enormously long list of racial and ethnic groups that the various municipalities are grappling and coping with and, you know, it is like in Miami, it is this huge, big, long laundry list and self-identification is fine, but no one is understanding what the implications of being in one class versus another class versus another class is when they just check these things. And it is not understood what the implications of that are for being able to collapse back, whether you can ultimately collapse those more detailed lists back into the reduced form.
DR. AMARO: I hear what you are saying and I agree with it. So, I think we just need to bring attention to that.
DR. IEZZONI: Okay. Dr. Sondik is coming. He is stuck in traffic. So, why don't we just all stay around in the room but -- why don't we take a little stretch here.
[Brief recess.]
DR. IEZZONI: All right. We are going to reconvene.
Dr. Sondik has come out of the terrible traffic that we understand you were stuck in
DR. SONDIK: Jennifer tells me they are digging up the city everywhere.
[Multiple discussions.]
DR. IEZZONI: Welcome. You and Hortensia have been having a kind of interesting dialogue, the two of you.
Hortensia, would you care to frame what some of the issues are that you and Dr. Sondik have been talking about?
DR. AMARO: Okay. Well, I think that the issues we have been talking about have a long history. They have been discussed by previous people before us, right, in many committees throughout the years, but most recently at the Hispanic -- what was it called -- the Hispanic Review of Healthy People 2000.
There were a number of individuals who were brought together to comment on where we were at and how we were moving along and the indicators of progress. Also, there was a meeting to look at the NHANES and other meetings that have come -- I think the Internal Data Working Group has been discussing these issues for awhile -- around -- and I guess the major issue we discussed was what steps can be taken to improve the type of data and the amount of data that we currently have on different racial and ethnic groups, especially the ones for whom our traditional sampling approaches don't work very well because they are concentrated in certain geographic areas and our normal sampling methods just aren't going to get to them. Oversampling is going to be way too expensive.
So, to try to think creatively about how do we address that -- and for many years there have been suggestions like, well, we should do specialized studies or different things have been mentioned, but nothing has ever moved, I guess, much beyond that.
Dr. Sondik has, I think, really been thinking about this with Jennifer and other staff and I think has some good ideas and suggestions about how to move it forward, which is pretty exciting. I think that that is what we were going to talk mostly about today, right? Focus on that.
There are other methodological issues, but we thought this is such an important one that really deserves major focus and that we would limit the discussion to that.
DR. SONDIK: Well, this is a challenge and it is one that I see, in effect, I would think practically every week and the challenge is for us to get an adequate, sufficient amount of information on the various subpopulations in the country that -- a sufficient amount so that we can deal adequately with the health issues of those populations.
I am not exaggerating by saying something comes up every week with respect to a particular -- one particular subpopulation or another and we face a real challenge in being able to use our resources in the most efficient manner possible to get information on these populations. The Hispanic Healthy People review was particularly memorable because it really honed in on this issue and I would say the -- it was the principal issue discussed during the review.
While there was information on some -- on the -- and I underline "the" Hispanic population in the country, it really was not broken down, except for, perhaps, one measure, as I recall, by specific -- any specific subpopulations. And the committee that has been called together, non-Department people, I think all non-federal people for that matter, were quite vocal in saying that we appreciate seeing some of this data, but some of it is ten years old, 15 years old.
It is interesting that there is data from the Hispanic HANES, but that goes back to the early eighties and we haven't seen anything since then. Several people brought up the lack of information on the Puerto Rican population, both in Puerto Rico and in this country. And when we -- it may have been me or someone else pointed out that our plans for HANES included some oversampling of Mexican Americans, this was not exactly greeted with a chorus of huzzahs that you had really found it at this point.
It was that there are lots and lots of needs. At the Data Council, the representative from the Indian Health Service, Tony DeAngelo, often makes it clear that there is not sufficient -- there is not a sufficient amount of information on the various tribes and perhaps the -- within the category of tribes, the various geographical locations, to adequately deal with their health problems.
We actually have met a couple of times now with the Indian Health Service, with the goal being to find a strategy where we can help to collect this information or at least catalyze the information and find out -- try to arrive at some ways where we can at least assist, if not go further than that, in getting the needed information on the tribes. In fact, they point to a book called The Gold Book, which actually is nearing its -- I am not sure anyone here is familiar with this, but it is nearing its 50th anniversary as being the -- Olivia, I am not overstating this, I think. You are probably familiar with this -- as being the principal document that gives information on the tribes.
There is other information but sort of the compendium of information is now that old. So, we have decided to focus efforts on taking a very hard look at the health of the American Indian, taking this document as at least the baseline, if you will.
We have some thoughts within NCHS as how we can meet some of these needs, but to tell you where I am coming from and I think where some others are coming from, I think we need some guidance to the Department on how we can meet the challenge, a departmental strategy, if you will, something where the Data Council, for example, knows when we are going to be getting particular types of information on particular populations.
These events are scheduled and planned for and budgeted for several years ahead. As it is now, most of these efforts are up to the individual agencies and in some sense I think that that is highly appropriate. We all have our budgets and we have difficult decisions to make within those budgets and I think we have to make those decisions. But I think there are some cross-cutting events, some of the surveys, for example, that deal with issues related to Healthy People, like the prevention surveys, the prevention supplements, to the Health Interview Survey, which deserve contributions from across the Department in terms of budget.
So, let me give you a couple of things we have been thinking about in terms of NHANES. Not only were we -- well, our original plans were to oversample, but having thought about these issues, we have another couple of strategies. One is we have been very strongly considering taking one stand from NHANES each year, which means one of the areas in which we do our sampling -- there are 15 of these -- taking one and in a sense floating it and using that one to address the problems of specific populations in whatever area of the country seems to be appropriate.
In particular, we are thinking about doing that for the Puerto Rican -- for the next year or two for the Puerto Rican population in the New York area.
Secondly, there is a proposal that was offered several years ago by the HANES staff, which has a tremendous amount of appeal. This is called the HANES RV, for HANES Recreational Vehicle. It fits right into HANES IV, for HANES 4 -- I always get a little chuckle whenever I say it but I think I am the only one. Isn't that true? I am the only one.
This is really quite a very nice concept because, as you know now, for -- to conduct HANES, it takes a couple of sets of four trailer trucks each that travel around the country and that enables us to get an enormous amount of information on those populations that we visit, but it really kind of logistically hamstrings us in terms of getting populations that are more diverse in other areas of the country, areas that are perhaps more difficult to visit.
So, this proposal was offered several years go when it seemed as if there was not going to be sufficient funds to do the full blown HANES and we have resurrected now and we are looking very carefully at the cost of it and seeing how we can build it into the HANES plans as an adjunct to HANES that would enable us not only to do HANES as we outlined it, but to enable us to deal with the special needs of special populations wherever they are. Whether it is taken an excursion into northern Minnesota in the winter, which we don't do with the trailers or it is visiting an Indian tribe in Arizona or wherever it might be, this would greatly increase our flexibility and enable us to get the objective measurements that HANES -- that are really the hallmark of HANES.
So, those are a couple of strategies for us. We oversample in the Health Interview Survey, but not -- we don't go after specific subpopulations on a regular basis. I don't know of any effort we have that is dealing with the Asian population, but correct me if I am wrong. Although we may get sufficient samples in some areas of the country, like in the L.A. area, to tell us something about that. But it is going to be relatively small.
So, the problem that really I guess I am putting in front of you, Hortensia is putting in front of you, is that we in general have a limited set of resources, but it is clear that we need to get information on these populations. Now, when I say "we," I mean, I think we need this information from the -- in some fashion, whether it is from the Federal Government, whether it is some manner that foundations are involved, whether it is some way that the private sector is involved and there is federal involvement in that, we need some way of doing this.
As it stands, I think there really are very significant data gaps and I think your advice to the committee and then to the Department in how we should deal with this, I think, would really be very useful. And I underscore what we have talked about a few times in the Data Council is the need for a strategy, a plan for this, I think would really be useful.
Olivia may have some comments or Pat.
DR. CARTER-POKRAS: We have a little success story. We have managed to develop an inclusion policy that we are trying to get cleared within the Department for racial and ethnic data within the data systems. So, we are hoping that once this gets approved that this will move that forward because a surprising number of our data systems don't have complete information on the racial and ethnic identification of people participating, say, in a particular program.
MS. WARD: Can you be more clear about what that will look like if you get approval, just to be a little bit more specific?
DR. CARTER-POKRAS: It will say that all the data systems within the Department have to collect racial/ethnic data unless they have a good rationale not to. An example would be, for instance, the patient registration files that the Indian Health Service has because they prefer to collect information on the people they serve, which would be Indians and members of federally-recognized tribes.
So, they have information, say, on a particular tribe, the tribe that they are a member of, whereas, they wouldn't have the full-blown categories within OMB Directive 15 or whatever the successor is to that.
MS. GOLDEN: That is an example of an exclusion.
DR. CARTER-POKRAS: That is an example of an exclusion, right.
MS. WARD: Your policy would change --
DR. CARTER-POKRAS: No, we would not require the Indian Health Service --
[Multiple discussions.]
That would be an example of a rationale or their justification for not collecting racial and ethnic data from everybody within their data system. But we would expect that the data systems right now are not will have to come up either with a very good rationale why not or they will have to start collecting that information.
MS. WARD: Do you have a sense of what the percentage is of programs who currently don't who would now have --
DR. CARTER-POKRAS: We don't have an exact number but a few years ago I had to collect some information on a like a 24 hour basis for one of the congressmen who was asking a question on that very thing. I found to my surprise it was something like a third of the data systems that I called up either did not collect it at all or did not collect it on everybody and I called about 15 or 18 data systems in that 24 hour period.
DR. IEZZONI: George or Elizabeth, coming from states where you often have articulated that you need to know more detailed information about the health of populations than might be available through something like the NHIS or NHANES, do you have any suggestions or thoughts about --
MR. VAN AMBURG: It is an intriguing option because, you know, there are some pockets of unique populations. We have a Middle Eastern population, of course, in Michigan that we don't get good data and we would love to have this selecting around to do that. That would be a really great way of doing that.
MS. MADANS: A similar approach could be used on HIS as well. I mean, as Ed said, HIS is large enough that without -- there is oversampling on HIS, but you get a fairly large proportion if you put two or three years together. But one could take some of the sample size and some proportion -- we don't know what that would be -- and say, well, this year we will do a special study. And it wouldn't be part of the national database, but it would be somewhat special.
There is also our telephone survey that we are developing and I think the committee has heard about which would give us, hopefully this is operational, many HIS's on all 50 states and as many local areas as we could afford.
I think this is kind of a generalized approach to get information on population groups, as Hortensia said, that we don't normally get, which would be states and local areas.
DR. IEZZONI: Is there a possibility of a state/federal partnership, though, on some of this? I mean, could some of the -- have any states taken -- I guess, New York, you said you were going to be doing something special on Puerto Ricans in New York City.
DR. SONDIK: That is what we are planning on doing.
DR. IEZZONI: Is New York City helping to fund that or --
PARTICIPANT: No, but it is a good idea.
DR. SONDIK: Why don't you discuss some of the arrangements and --
MS. MADANS: Most of the interactions we have had at the state and local level has been with the telephone survey and we have done some things in California. We just spoke with them last week about doing an expanded HIS in California. The state and local area integrated telephone survey, which is SLAITS, was really designed to get information on health care reform and welfare reform at the state and local level. It is a federal survey, but we would like to have as much usefulness at the state level. So, we are starting small. I mean, all of these things are starting small and it is very different than the way we have done things in the past. We are not going to 50 states.
We hope to collect data in about five or six states but we were talking about 14 states and working with them to develop a questionnaire. Most of the questionnaire we would like to have as standards, a phone survey and it is added on to the immunization surveys. We are already asking 20 minutes of questions on immunization status. We can talk more about that survey, if you would like.
We want to give states the option of adding on maybe five minutes of questions and also changing the sample design. Our funds pretty much will allow us to do a straight sample of about a thousand in every area. But we could change that sampling strategy. Especially in the area of welfare reform, we really want to oversample low income. A state may want to oversample a particular county and then there are a variety of different ways of cutting up that pie.
So, we are working with the state representatives to do that. Really very informal at this point and ad hoc because there is really not a whole lot of money. It was funded primarily by ASPI last year. But we have talked to
CSDE(?) and various groups about our plans for state and local area data collection. The flexibility in that is that you can add on to HIS. You can do something different, but you can have multiple surveys going on in the same area because we have a very large telephone frame.
So, if you wanted to, you could have four or five -- and get the money, you could have four or five surveys going on. Because of the survey integration that the Department has, a lot of the questions kind of follow through from HIS, SLAITS, HANES and the Medical Expenditure Panel. So, there are these connections among all of the surveys and we are also doing some statistical matching to SIP and CPS so that we can get all of the -- try to make -- do some matching, do some modeling.
DR. AMARO: I think that these ideas are a really good compromise between, you know, probably the ideal of what some people would like to see, but it is too expensive, which is, you know, the oversampling or yearly studies on these different subpopulations, which just wouldn't be feasible, but it would provide some data that could be gathered on some kind of interim basis on different populations that would be more regional in nature, especially for groups that are not distributed, you know, equally across the country so it would be consistent with the goals of providing more local data that health departments could use.
So, I think it is -- you know, it is a nice compromise of some of the strategies we started talking about and versus not doing anything. I think it is a reasonable approach.
MS. MADANS: I think we want to imbed all this within the national data collection so that you have the comparability with the national data as opposed to really special studies, which would go off on their own. You want to make this somehow part of the whole overall structure so that it kinds of fits in without making it all one survey.
I think in the past we have tried to do that so that everybody would be included in the national sample and, you know, you lose by doing that, I think.
DR. SONDIK: Let me mention something about the Hispanic review that I found useful. I guess you all know about Healthy People and it covers a variety of objectives from different aspects of health outcomes and services and behaviors and the measures come from a variety of different surveys or activities and not solely from NCHS by any means.
I found it actually quite useful to have this session in which we were dealing with not, for example, well, what do we get from NHANES for a particular population, but dealing more with the needs of the population on a broader basis and having people who could speak to the different aspects of that in the same room. In other words, focused more on need than on particular mechanism for collecting the information.
It seems to me that in carefully looking at what came out of that, that ought to give us some ideas about how the different agencies can work together in collecting some of this information. So, for example, before we march off and look at the Puerto Rican population in New York, we ought to be looking with SAMHSA, at what SAMHSA has on that population, what they want, you know, what they need and seeing how we can join forces in a federal-federal relationship, let alone a federal-state or federal-state-local partnership.
Sometimes I think we tend to look at these things mechanistically. What can we get from a particular mechanism about a particular population instead of thinking more broadly about the health needs of the population across the board. So, I found it actually -- you know, I have brought it up many times, but I really found it extremely useful because it highlighted for me -- to me, the importance of looking at the needs of the population across the board and at the same time taking a hard look at what it is we do have and how hold it is and how current it is and think about how we can build this in more into what it is we are doing.
There are many, though, as we are all aware, of subpopulations, if you will, small populations in this country, that have these needs and sometimes, you know, we only know about them, I think, when a particular crisis arises or when that population takes some actions for its voice to be heard.
I guess there is some merit in working that way, but I think we also do want to know what it is we have on which population at which point in time. I say that particularly with my other hat, which is the advisor to the Secretary on health statistics. I think it is important we have a sense of what it is we are going to be getting when from which group in this country.
The groups, by the way, don't have to be racial or ethnic groups, obviously. You know, there are very significant issues about what we know about SES groups, issues about relationships between SES and access, relationship between that and managed care or the changes in the health care system. I see these are all part of, you know, if you will, an agenda that we need to follow up on.
DR. MOR: I was going to sort of hold off for awhile and as soon as you said "managed care" and the context, it sort of raised a whole series of issues for me. Fundamentally, though, if one assumes that geographically concentrated subgroups of the population have different health experiences than others and that sometimes the geographic concentrated groups are relatively small would thereby be missed by normal sampling, even oversampling, like NHANES or HIS, that really does a number of things. One, it calls into question the current two-stage HIS sampling strategy, which is geographically based.
Secondly, the same thing with managed care, since many of the phenomenon of interest occur at different rates in geographically concentrated areas. The sensitivity of any effort to look at what is going on is going to be quite geographically sensitive to the choices of locations, which sort of drives you right away to some kind of population data collection and up and above to sort of, you know, very clever notions of rotating options, as well as the state thing.
But is there any notion that it is possible to add items on a census basis that might pertain to health and/or experience of disability, which at one point then could be crosswalked to more detailed information so that one could actually do a micro-macro extrapolation where you really do have a full denominator that you can then begin to do some calibration on.
MS. MADANS: Are you talking about the census?
DR. SONDIK: I wondered if this new survey that the census is proposing, the --
MS. MADANS: The American Community Survey.
DR. SONDIK: Yes, I was going to say American Family Survey, but that is something I received in the mail the other day I threw out, I think.
MS. MADANS: You could have been the million dollar winner. We need the money.
DR. SONDIK: I think that was it.
DR. MOR: What about the American Community Survey or is that not going to be detailed enough?
MS. MADANS: Well, I think it is really a question of content on any of those. I thought the American Community Survey was going to pretty much be the long form, but I can be wrong. And they are fighting terribly over what is going to be on the long form. They, in fact -- Michelle isn't here, but the last couple of weeks they have been -- there has been a lot of negotiation with Census about what the disability questions are going to be and I think they have come up with a whole new set.
DR. MOR: A whole new set that hasn't been tested?
MS. MADANS: They are going to test --
DR. IEZZONI: That is what we hear about in September at our breakout session.
MS. MADANS: There are six -- the last I saw of this -- I mean, there was this whole -- I don't know how much you have talked about this, but the Census didn't want to ask any ADL/IADLs and that caused a fury. There were little furies going on all last week.
So, I don't know. I mean, I think that it would be wonderful if they had something that was useful, but they tend not to follow the Department too closely, our department, in what they want to do.
Now, maybe with the American Community Survey, we have some input in with -- that would be close enough to, I think, a census.
DR. MOR: Right, because the point is that if there are a few items, many of those items are actually correlated with a lot of other more detailed items that are standardly obtained.
MS. MADANS: I think there were six. I mean, I could send you the six they are looking at. I think it was -- there was an ADL and IADL, getting around, shopping, going to the doctor and a work disability -- there were six. It actually looked much better than the ones that they had before.
If they get the data sharing bill passed, then we could actually use the American Community Survey as a sampling frame for the HIS, which would take us pretty far into identifying some of these population groups of interest.
DR. SONDIK: In fact, I think there is a hearing next week on data sharing.
MS. MADANS: Because right now we can't use this.
DR. SONDIK: So, that is a possibility. From your more theoretical construct, I think it really is a possibility. There is another activity that is perhaps not quite along the same line that is taking place in the Department, where there is an effort to look at these -- how the health care delivery has been transferred and to see what its implications are in significant part for our health care surveys and the other related types of activities that go on in the Department. Because the frames that we are using are really structured more on the basis of the facilities than they are on the basis of things like, for example, the financial arrangements.
So, it is conceivable that in coming up with that framework, we will be able to a priori get -- be able to identify some of these populations.
But I am -- when I mentioned before other populations on a -- subpopulations on a geographic basis or SES, I am actually getting at that type of thing. I think there are lots of different populations in this country, perhaps even more so than we have been studying in the past because of these care arrangements and since in significant part, the work we do serves health services research, I think we have to understand how to collect our information in such a way that we really can be supportive of that research.
I think that is an open question as to how best for us to do that. Maybe the strategy you are getting at, linking this to something that -- a measure that is in the census and then linking that to the other data that we have or the American Community Survey, even better, I think, would enable us to do that.
But that is -- well, I guess, it is not that far down the road. The census is, but the American Community Survey is --
MS. MADANS: That is after the census.
DR. SONDIK: Oh, it is after. I thought they were going to begin it before.
MS. MADANS: I thought it was 2005.
DR. SONDIK: I thought it was much earlier than that.
DR. MOR: Because there are also some technical issues about the sampling frame and the sampling theory. If you really do believe that things are as geographically heterogeneous and/or nested, then it seems imperative to collect and be able to assemble information that can be aggregated up or somehow or another those individual representations of the people who respond to surveys can be hooked and nested automatically to their local area, to their market, to some other characteristics.
DR. SONDIK: I think one of the areas in long term trends, in terms of long term trends in which we are going is that I expect that we will have a great deal of information from these various health care systems, administrative information. I think some of the things that we are now doing through national surveys, we will be able to do through a build up of this from the local -- if you will, the local areas. But we really need to understand the characteristics of that information, the quality of it, appropriate weights and all of that type of thing.
I see that it is down the road, but I think it is a -- I am quite confident that we will see that within the decade for some measures of information.
DR. MOR: Just one last question. I don't remember what the current status of the ability to generate information about the -- and to have available for examination and analysis, information about the environment in which an individual respondent is. Is that going to be available as part of the files, do you know?
Linked to, like, you know, the link of somebody's part of this --
MS. MADANS: Contextual data.
DR. MOR: Contextual data. That is the right word. I am sorry about that.
MS. MADANS: Well, we have a plan. I mean, I know you know that our problem with contextual data is the confidentiality problem and that putting on very limited amount of contextual data allows you to identify the area, which given the amount of information we have on our sample people, for a reasonable subset of the sample, you could probably identify individuals and that is against our confidentiality law.
So, what we are trying to do is make the data available to researchers without actually releasing the data. Smoke and mirrors. There are a couple of ways of doing that. The long term goal is to develop an automated system whereby we would provide not the real data but a test data set, which looks like the real data for you to develop your analysis on.
Then we would have the dedicated computer, where you could submit your programs to run on that computer and there would be an automated testing of the program and the output to make sure that you weren't getting confidential data. And it would all be done automated, so it would be fast. You could run things in batch. Nobody would ever see it. And we are developing or trying to develop that system. So, you would have a certain data set to work on on site, but then the actual complete data file with all the added on data would be in a secure place and you couldn't get listings, things like that.
One step back from that is to actually have somebody look at it. You would still do the same thing, but rather than having it being automated, you would look at it. We are actually doing that. We have a pilot project now at the NSFG with trained people in Atlanta. They are not -- I mean, not Atlanta, in RTP. They are actually running the program and then sending the results back.
We are trying to set up data centers. We would start with one at NCHS so people could come in and work at. This is similar to the census data centers, but we could have them in different places.
If we had the same data sharing bill, that would allow us to share data with the other statistical agencies, would allow us to have sworn agents. We do not have sworn agents now, like the Census Bureau has sworn agents. So, there is no way for us to have a contractual relationship with a researcher. Census has sworn agents. If we had sworn agents, we could make more data available.
We will make limited amounts of contextual data available, but very soon we run out of what we can make available. So, there is this kind of range of options that we are developing to make the maximum amount of data available to the researcher without actually putting out a public use data tape that has the individual -- all the individual data and 3,000 census track level variables for that person.
It is not the best. It is not what people want. We are also trying to set up some fellowship programs so that we can encourage folks to come for a year. We are starting something we hope this year with the American Statistical Association to have an ASA fellow. They would be not NCHS's employees but would have the right kind of contractual relationship and would have access to confidential data.
DR. SONDIK: We do have some state files that we have put out on CD ROM that are HIS data and they have been slightly corrupted by noise, so that you get the right marginals, but the data is protected to -- corrupted to protect the innocent in this case.
So, we are working on that type of thing.
MS. MADANS: The 1995 HIS is a much nicer -- it is a state friendly design. It is in every state and there are at least two SPSUs in every state. So, we will be able to put more information out at the state level. The state files that Ed is talking about are on the old design. So, they are even more limited. They are top coded.
It is one step. If we can deal with the dual frame HIS and the SLAITS, we can -- and actually draw samples within the state and within counties, as long as the sample is large enough and the county you are taking it from is large enough, we can put out county level files.
The problem now is that either the counties are too small or the samples are too small and you can't make estimates anyway. So, then, you have to actually provide the contextual -- the linkage information on the file and that just links you to everything.
The answer to all our problems is triple the size of the HIS and triple the size of HANES.
DR. AMARO: I guess that I wanted to sort of go back to a comment you made when you started around guidance to the Department on how we can meet the challenge of starting to integrate these strategies into a broader effort. As you go forward with this, I guess my question is are there ways of working with the committee that would be helpful to you in moving forward. Are there specific things that you think we should be looking at to support this effort in a broader level?
DR. SONDIK: Well, what I was hoping was that the committee would consider the extent of these problems, the extent of the gaps in some of these areas and consider whether it wants to make a recommendation to the Department on ways in which these gaps can be addressed. I guess I always go back to a framework in which I try to think of some -- what I think the critical questions are for the particular population and ask whether we have sufficient information to answer those questions.
So, in a sense I am -- I guess I am positing before you as to whether you wish -- you would want to take this particular topic on as something in which you would want to give the Department some guidance, how to deal with these issues related to specific subpopulations.
DR. IEZZONI: We are hoping to have some new members, who will hopefully help us think about that as well as some of the members of this subcommittee. We are specifically hoping that the Secretary might appoint somebody, for example, who is an expert in survey design, you know. And, so, I think that -- let us keep that as something that we will be hoping to help you with and make a specific plan for how that might work once we get our new members appointed, which will, hopefully, be soon?
DR. GREENBERG: Hopefully by the September meeting.
DR. IEZZONI: Hopefully by the September meeting.
DR. AMARO: I should say that in our subcommittee charge and work plan that is definitely Item No. 5.
DR. IEZZONI: That is not in order of importance.
DR. AMARO: No, it is not, no. But it is within our broad charge and work plan, working on this, as well as other issues related to methods that we have talked about before.
DR. SONDIK: I would imagine this and related topics are going to be coming up quite a bit with the Data Council and what we at NCHS would be happy to do is to keep you apprised of whatever developments we have in this area. Certainly, as our efforts with the Indian Health Service move along, we will keep you up on that.
DR. IEZZONI: I think that would be great, Dr. Sondik. We are very fortunate to have Hortensia on the subcommittee, who brought this issue to our attention, but some of the rest of us do not track these issues and, so, are unaware of them and, therefore, unable to help you. It sounds like Vince might be a ringer, too, from the nature of his questions in terms of being a user of the data sets.
DR. MOR: Yes, I use these data sets a lot.
DR. IEZZONI: So, I think that that would be helpful if we could --
DR. SONDIK: Okay.
DR. AMARO: Also, that if you are holding any meetings -- I don't know if there is going to be a follow-up to the NHANES to discuss these issues more, but if there are, that we can certainly have some kind of, you know, some kind of crossover between the two groups.
DR. SONDIK: Okay.
DR. AMARO: I just wanted to say that we are just getting started with our work plan, so I think people are feeling pretty overwhelmed with the work we decided this morning we are going to take on. So, you know, you are getting some of that, but that it is definitely within our work plan. It is a priority. We are really hoping that when the new members come with specific expertise in that, that there will be, you know, two or three of us who could work really closely with you. This is definitely an area of interest to us and it is within our work plan.
MS. MADANS: I think the other is publication dissemination. There is a lot of data that we, obviously, can't analyze everything and then people don't know we have and so in areas that you can -- say if you publish this or made this known, that this might be useful to the user public --
DR. IEZZONI: One of the things that I am hearing is that people are able to take, I guess, the Medical Expenditure Panel data right off the AHCPR Web site. Is that something that people have heard of? I had a colleague at the School of Public Health who was telling me about that. Because he was going to be using it to create an analytic file.
MS. MADANS: I know they put out a CD ROM, but it is awfully hard to do that.
DR. SONDIK: We are experimenting with a number of different ways to actually get data off of the Web. There are two principal areas. One is in Canada that is being used and we are -- which I understand is really quite good, but I haven't seen that one. There is another one that the Census has developed and we have an experimental setup in which we have HIS data that you can address with this system. The system is called Ferret(?), highly appropriate, and it is quite easy to use and allows you to get fairly detailed data, not so detailed that you violate confidentiality, but it really is a major goal of mine to make the data as close to the user as absolutely possible.
But what is interesting is that when you start to get close, then you get into the points that Jennifer was raising before and you start doing things that are really a lot more -- seem to be a lot more complicated, like corrupting it by noise or -- which is actually a very -- I think is really quite a good strategy, or get into this business of actually allowing people to use what amounts to a toy data set so that they can get a feeling for what is in that information and then submit their analyses and we will run them and review them.
Everything is straightforward except the review.
DR. IEZZONI: Right. Exactly. And it needs to be done like this.
Okay. Any further comments?
[There was no response.]
Should we adjourn for the day?
[Whereupon, at 4:37 p.m., the meeting was concluded.]