[This Transcript is Unedited]
Hubert H. Humphrey Building
Room 705A
200 Independence Avenue, SW
Washington, DC 20201
DR. IEZZONI: Thank you everybody for coming. This is the Subcommittee on Population for the National Committee on Vital and Health Statistics. We'll talk in a few minutes about why we are here today, and a little tiny bit of tomorrow.
But let's first go around the room and introduce ourselves and give a one sentence summary of who we are.
My name is Lisa Iezzoni. I'm chairing this subcommittee. I'm from Beth Israel Deaconess Medical Center in Boston.
MS. RIMES: I'm Carolyn Rimes. I'm staff to the subcommittee.
DR. QUEEN: I'm Susan Queen from HRSA, and I'm staff to the subcommittee.
MS. WARD: I'm Elizabeth Ward from the Foundation for Health Care Quality in Seattle, Washington.
DR. NEWACHECK: I'm Paul Newacheck from the University of California at San Francisco, and a member of the committee.
MR. HANDLER: I'm Aaron Handler. I'm with the Statistics Office, Indian Health Service, Public Health Service. I'm a staff to the committee.
DR. HENDERSHOT: Gerry Hendershot, National Center for Health Statistics, and staff to the committee.
DR. LOLLAR: I'm Don Lollar with the Office on Disability and Health at the Centers for Disease Control.
DR. PLACEK: Paul Placek, National Center for Health Statistics.
MS. COLTIN: Kathy Coltin from Harvard Pilgrim Health Care and a member of the subcommittee.
[Additional introductions were made around the room.]
DR. IEZZONI: Great. Before we get started, I want to thank our staff members who really have pulled together I think a good program for us for today, and a tiny bit of tomorrow. I'll explain what I mean by that in a second.
Unfortunately, one of our speakers, who is a member of the committee, Andy Kramer, won't be able to be with us because of continuing problems with orthopedics, but hopefully he'll be returning with us again soon.
We also have a number of other committee members who are missing -- Barbara Starfield, and Dan Friedman. And tomorrow we will be losing Elizabeth Ward and Paul Newacheck at ten. So what we have decided is that tomorrow we will start the meeting at eight o'clock, and we will adjourn tomorrow at ten, because basically the only people will be here will be Paul, Kathy, and I. So that's just a tiny bit of housekeeping.
The subcommittee about nine months ago decided that we wanted to look at filling in that unresolved place in the core data elements, which was functional status. And we spent a little bit of time thinking about how we would pursue investigating that. And finally, at our February breakout meeting of the full national committee, we decided to very much focus our efforts in this area.
And what we are basically going to be doing is focusing on the feasibility, the appropriateness, the value, the potential problems with having ICIDH codes serve that role as filling in information about functional status on administrative transactions. And I need to underline that, that what we are talking about here are administrative transactions within the health care setting.
So this is very separate from what the National Center for Health Statistics usually does, which is do surveys, and ask people detailed questions about their functioning. This would be on administrative transactions such as encounter forms submitted by physicians.
So today we are going to be hearing from a number of speakers who each will have a slightly different perspective and level of expertise in the whole area of functional status measurement. And we hope that they will speak to us about their area of expertise, but we also hope that they will give us some insight from their point of view about how ICIDH might work in the capacity that I have just described.
And let me, before we get started with the first speaker, ask if any of the committee members want to expand on what I just said, or kind of give additional charge to the speakers? No, okay. So we're set.
So I understand that Michele Adler, who is on the first panel, will arrive when she arrives. So Don, you are it; Don Lollar, from the CDC. But you've thought about this a lot, so we're looking forward to hearing from you.
DR. LOLLAR: I've thought about it a lot, but I haven't thought a whole lot about it from the -- well, I have thought a lot about it, but not from so much the proxy standpoint that you have as the panel. Because my first notion is that if you assume proxy --
DR. IEZZONI: Don, do not look at the specific verbiage in this as indicating what you need to do.
DR. LOLLAR: I'm going to assume this is a free association about the issue of proxy.
DR. IEZZONI: You're a psychologist. You can free associate as much as you want. Don't assume that any particular linguistic decisions were made.
DR. LOLLAR: It did communicate an area of concern. So to that extent, then I'd like to address that.
DR. IEZZONI: I think Andy Kramer was going to speak a little bit about proxies.
DR. LOLLAR: To the extent that a physician is a proxy response when they put in an ICD code or a CPT code, then I think if you are going to look at another health care professional putting something else into a patient encounter form, then we're talking about proxy. If you are talking about somebody else beside that person, beside the health care professional, and you are calling that proxy, then I think we have missed the boat already.
DR. IEZZONI: Why don't you just tell us what you wanted.
DR. LOLLAR: Okay, assuming that, the first thing I think is if we are talking about trying to put ICIDH to codes onto some kind of form, as I mentioned I think previously, I think where you do have substantial problems with at least reliability, if not validity of data is if you start looking at participation or environment codes.
As much as I would like for those to be a part of administrative records, and in the future that may well be able to be done with our continued improvement in computers and electronic data collection. I don't have any question at all that several years from now, maybe by 2010, this committee will be looking at other sections, both for perhaps participation and environment as codes that ought reasonably to be a part.
DR. IEZZONI: Can I just interrupt, Don, and make sure that all the committee is on the same page about what Don is talking about?
DR. LOLLAR: Do you want me to say a little bit more?
DR. IEZZONI: Yes, I think that would be helpful, because some of the committee knows a little bit more about ICIDH than other parts of the committee. Unfortunately, we kind of had to rearrange the schedule, and we're not being introduced to ICIDH until after you speak.
DR. LOLLAR: We know about the WHO history. You know that ICD is the primary classification, and then what we are talking about is the complement from what I would call a companion document to ICD, which is why it seems to fit appropriately. Now what it does is talk about the influence or impact or associations of disease injury at several levels. The generic terms are the body system level, the personal level, and the interaction with the society level.
The body function level for ICIDH is called impairments. The personal activity level is activity limitations. And the interaction with the environment, that is the kind of outcome measure is called participation. This revision has included a fourth component which is the environment.
Now the point about this that separates it from any of the other conceptual framework, and there are several, is that it also includes a classification and coding system so that each of those four elements ought to be able to be more clearly delineated, and in fact put into an alpha-numeric code.
So when I say P, meaning participation, and E, meaning environment, and AL I usually say for activity limitations. So what I am talking about is that participation and environment, as important as that outcome is, and as important as the environmental factors, whether they be barriers or facilitators are, because they are farther away in much of the thinking that goes on in health care settings -- is that appropriately gentle? -- it is not as much on the screen in health care settings.
And because there is much more personal element in the understanding of how much someone is participating in the society in that there is a much more perceptual notion of that. And whether or not an environmental factor is facilitating or a barrier, and how much that may be a facilitator or a barrier.
So given those issues, my notion is when you are talking about functional status, and the closest element in the ICIDH to framework that approximates functional status as I see it related to health care settings, is the activity limitations; the person level activities.
Now some of my colleagues would say that we're already giving away the store when we don't fuss for or fight about participation, because in some sense you could call participation, for example, working, traveling, boating, all sorts of those societal settings or societal roles as being also activities, and they are on level. But at this point I think the activity limitation dimension is the one that allows us to most comfortably stay attached the health care setting, health care environments, various ones, across the spectrum of care that include administrative data. So that's the first part.
Now my assumption always has been, as I have said before, that activity limitations -- if we were going to try to operationalize this, we've got several things. One is you really need some data to see to start with, how much it's going to be used. How cost effective it is going to be to even mess with it. And so it is relevant, I think, to know at least without getting into surveillance survey tools, it is informative or instructive to know what those tools suggest about activity limitations.
And our office did such an analysis of the NCHS national health interview survey disability supplement, trying to take the phase 1 that had to do with disability, and looking at the questions that would most closely approximate each of what we would say are the eight chapters of activities limitations in the ICIDH-2.
For those with visual difficulties, you may have some problems, but I brought 20 copies of that for you to look at. Now the reason I bring this up is merely to say to you to the extent that any reporting by someone in a survey of an activity limitation, never mind diagnosis, but an activity limitation, if you look at the third row under total all ages, you notice that this table has total at the top, and then age 5-17, 18-64, and 65 plus.
There were 45 items, I think 42 items that were used to delineate the 8 chapters. You notice they are: seeing and/or hearing; learning; communication; movement, which are generally the simple movements that we would call reaching, grasping, holding; moving around, which is generally the mobility; personal care; routine activities; and behavior.
You notice that in the third row we see 19.35 percent of the population reporting one or more -- that's unduplicated 19.35 percent. And you notice that for example in 5-17 year old, the learning is 9.38 percent, and that goes down when you get to be 18-64. We evidently must get smarter. But then we get dumber as we get older, because then the learning goes back up after 65, the learning problems, back to 9 percent.
The point is that when we are looking at activity limitations, if we assume that when someone comes in for a clinical encounter, that not everybody is going to have an activity limitation, unless you use that term very broadly. Then we are not talking about requiring that element be included on every patient encounter form. That is, that it isn't necessarily relevant in every patient encounter form. If we were to assume that, I think frankly it would substantially undermine the data that is real.
Because I did, by the way, spend some energy thinking about proxy, that's where my notes are all.
DR. IEZZONI: That's okay. Tell us what you thought.
DR. LOLLAR: We do know that observable characteristics are the key to the reliability of an proxy respondent stuff. The farther away you get from observable characteristics, the more difficulty you have with reliability. On the other hand, you can't assume that different observational ratings, one of them must be invalid. It is possible to have two valid proxy responses, and yet they be unreliable, one with the other, because they are done in different settings, by different people, with the person using a different device.
So there are problems with proxy stuff. But we know that because of our concern primarily in our surveillance work. I guess the thing that I'm intrigued by, an article that has not been published -- in fact, it just came across our desk for review -- looking at functional limitations and the inclusion of functional limitation in hospital notes in elderly patients. It's from Yale, and some of you may have seen it or been a part of it.
But what you found was that physicians, when they are asked about functional problems, will often say there aren't any, or they won't mention them in the notes. Most of the research has suggested that if there is no mention, then that means there is no problem. And in fact, what this study clearly showed was you really can't assume that at all. That in fact because the physician doesn't put anything about functional status in, that therefore there are no problems functionally.
The point about all that is if we are going to do something like this, it has to be addressed globally. What I mean by that is it has to start at the top. It's got to be dealt with in training programs for any health care professional, because I think if it's seen as anything -- as I have said, the data field doesn't necessarily have to be used with every patient, but it needs to be available for every patient.
I think if we want to pilot it in certain ways so that we see where it seems to be most useful, then I think that is really crucial. And as I said to you folks last time, we are poised to try and be helpful in that regard from the standpoint of not only energy and commitment, because we think it's going to make a large difference in terms of what we know, and what we can understand about health services, health service utilization, and health costs for people with disabilities.
So we are not only interested and committed to it, we are even willing to put some money toward it, although we realize that our pot of money is substantially less than others' pots of money that would dwarf our efforts, and make this easily doable.
I do know that what we heard last time from the folks who were representing the financing side is that when you have enough trouble getting physicians or health care folks to put diagnoses on the forms, what in the world do you think you are doing trying to add another element that may or may not be helpful to many folks? And all I can say to you about that is that I think we are moving more and more in a direction of wanting better data. And I think having more people that are affected, that is, that are experiencing activity limitations.
And I think not to do this is to miss a real opportunity for adding to the science. And it probably will help monetarily, but that's not my area of expertise, so I won't talk about that.
Well, I will stop there. The other data that we got was just on the number of different activity limitations. Of those, 19.53 percent of folks who reported, 45 percent of those reported having one activity limitation; 24 percent having two; and 30 percent having three or more. That doesn't have anything to do with you guys. It has to do with my notion of how one would do this on the patient encounter form. And that is the literature indicates that some physicians are willing to put two diagnoses down, but the average I think is 1.5 on most patient encounter forms.
So my assumption is we might could go for two activity limitations, but I wouldn't want to push it any more than that. And again, if you decide you want to pilot it, there are numerous ways to pilot it. You can pilot it by age. You can pilot it by medical specialty. You can pilot it by looking at the primary diagnosis that is given, and only limitations related to that diagnosis. There are numerous ways to try and make this a very good fit, and a reasonable extension of what already is being collected.
But the bottom line is we also know that diagnosis doesn't really predict function. And it doesn't really tell us a lot. When we try to make assumptions on the basis of diagnosis about numerous other things, it gets us in trouble.
DR. NEWACHECK: Don, can you just help us a little bit by showing how the eight activity limitations that are in the handout you gave us, correspond to the ICIDH-2 booklet that we got?
DR. LOLLAR: If you look at the second section, activity limitations. I don't have that in front of me. The A section. Do you have a page number on that?
DR. NEWACHECK: I have page 46 as activities.
DR. LOLLAR: Let me tell you what we have done. There is a difference between Beta-1 draft and Beta-2 draft. In the Beta-1 draft seeing and hearing was Chapter 1, learning was Chapter 2, or activities of learning and applying knowledge is what they are calling Chapter 1 here. WHO took seeing and hearing out of activity limitations and put it in impairment.
Then they took the purposeful sensory activities, 110 on page 46, and put them under learning. We don't believe that's a wise move. We believe that learning is certainly substantially beyond seeing and hearing. And that seeing and hearing need to be separate as activities, i.e., reading newsprint, seeing people across the street, being able to recognize people. Those are activities, as opposed to actual impairment like visual acuity, for example.
So the distinctions and the importance of that particular thing we see differently than WHO. So that's why Chapter 1 for us is seeing and hearing. Learning is 2. I assume that every one of them after that you've got communication is 2, and we've got it as 3; movement is 3, and we've got it as 4. So everything moves back a chapter, until you get to 8.
And they have Chapter 8, performing tasks and major life activities, and what I have called that is behavior, because there is the other conceptual thing, which I am not sure we need to get into here, is the distinction between the person level activity limitations, and the societal or social participation. Where is the interaction with the environment line drawn?
And we think WHO has pushed that line too close to activity limitations, and so we have tried to say this is personal behavior, and it has to do with things like from our standpoint, maintaining your own emotions, maintaining your own behavior, maintaining your own aggression. That is, being able to manage that within yourself. It has certainly social implications. So those are the two changes.
DR. NEWACHECK: So when you say "we" you are talking about CDC?
DR. LOLLAR: A group of us that have been working on this for six years, which includes the Canadians, the Australians, the Brits. That decision is going to be made at the WHO level very soon.
DR. NEWACHECK: Yes, because I was going to say this Beta-2 is a year old now.
DR. IEZZONI: It's older than that I think, isn't it?
DR. LOLLAR: No, the Beta-2 is just now a year old. It came out in April 1999. So it's a year old.
DR. NEWACHECK: So when will Beta-3 come out?
DR. LOLLAR: There won't be a Beta-3. There will be a ICIDH-2 voted on by the World Health Assembly this time next year I think it is. So there is going to be an interim meeting, we were just told, June 26-27 in Geneva, where we will get 2B, and we'll be real intrigued to see what changes have been made, particularly in A.
What I have said, and I may have said this to you last time, as I perceive it, the medically oriented folks have a great deal of interest in impairment. The advocacy community have a serious commitment to P and E. For me, the crucial link is the activity limitations. It's the thing that can help us pull things together. As you know, my notion of this, as I said to you last time, is the crucial element for activity limitation is the issue of naked person limitation, versus compensatory use limitation.
And that my suggestion is that on that data field there be three digits, one for the chapter, that is what's the area where there is the limitation? Two is what is the limitation, the extent of the limitation without any assistance? And three is with currently used compensatory mechanisms, what is the level of functioning, or what is the limitation?
I think not to do that, again is to put the environment way out. So I think the environment does relate at the activity limitation person level, but it's on the basis of how do the compensatory mechanisms assist the person with the particular person level activity that we're looking at.
DR. NEWACHECK: So you would have like 24 fields potentially then?
DR. LOLLAR: No. Having filled out a gazillion patient encounter forms, they don't have little individual things. So what I call a field is actually a place for the diagnosis. That's one field for me. The second diagnosis is another field. So I would have two fields that would include three digits each.
DR. IEZZONI: Don, are you going to be able to stick around until lunch?
DR. LOLLAR: Until lunch, yes.
DR. IEZZONI: Good, because I think Paul and Gerry are going to try to walk us through ICIDH. So Paul and Gerry, it would be good for you to think about maybe how we can refer back to the numbers that Don has produced to kind of give us a sense, as Paul was asking, of how you might be able to anticipate what the rates of identification of these types of deficits or impairments would be.
Don, let me just be clear. Your specific proposal is adding ICIDH coding fields, two of them, to physician encounter forms?
DR. LOLLAR: Yes.
DR. IEZZONI: What about nursing homes, home health care, other types of care settings where in fact the whole issue of function and impairment is more obvious?
DR. LOLLAR: Are you saying whether or not you ought to have more fields?
DR. IEZZONI: Well, how do you feel about that? No, but how do you feel about that, knowing all the other data collections, the MDS, and everything else that goes on in those other settings, and OASIS now. I don't now what it's going to be rehab -- Carolyn?
MS. RIMES: It's going to be a surprise. Right now it appears to be MDS. But once the regulations are published, things are certainly going to be enunciated.
DR. LOLLAR: I think the thing that has been frustrating to me is that what has happened is that instruments have been being developed over the past few years, OASIS, minimal data set, all of these kinds of things. The instruments have been being developed, and I'm sad that they haven't been being developed consistent with the ICIDH-2 terminology.
If folks had just said, gee, we wouldn't think about doing anything around diagnostic procedures without using ICD, and WHO did ICD. Why in the world, as we are developing the measures, don't we develop them consistent with that framework, with those concepts? If we do that, then there is a natural convergence of what it is you are measuring with the classification and coding scheme that you have got.
So to the extent that we end up having to back code, like we did the NCHS into ICIDH-2, we can still do that. You can probably do that with OASIS, particularly people like Hendershot and Placek and all these really clever people can do this back coding stuff. So you can take OASIS -- with some work, those instruments, I am almost certain, can be put into this coding scheme.
Now what you have to understand is that what I'm talking about at that HCFA-1500 patient encounter form is baby steps, because the others are substantially more. If in fact what we are talking about is the tree level, some of these other instruments allow you to look at it at the leaf level, past the branch to the leaf, and even to the stem. Is that not fair?
PTs will tell you whether or not the middle finger of the right hand, the muscle in the finger -- we can take every muscle, every muscle group. The classification scheme allows you to do that. I have stayed at the chapter level for what I perceive to be the beginning of a process. To the extent that in the nursing home settings there is a need for more defined assessment or reporting of function, then the instruments can do that, but the coding system also has that capability.
And so the point is to map those onto one another. See, the whole conceptual framework, when you talk about WHO calls activity limitations, OTs have called the ADLs for forever. We have used functional limitation. We have used all sorts of English words that talk about what people do in everyday life. I don't care what it is, that's what we are talking about. We are talking about can you move around, can you communicate, can you take care of yourself, can you learn stuff, how is your behavior?
This is not rocket science. That's what I continue to say. Don't make rocket science out of something that isn't rocket science. It may be very difficult to find out whether someone has some medical condition that has symptoms that can go to four or five etiologies, and you have to really look at the blood work and the MRIs and stuff. Activity limitations is not like that. We are talking about stuff primarily that you see.
And we have to just be clear enough, and I think we have to keep egos and money out of it as much as we can when we are doing this, which is what it usually comes down to. I developed this instrument, damn it. The functional independence measure is God's gift, and it's the only way that we can measure this. Well, having been a psychologist, that has come throughout the annual of every discipline. Somebody gets an idea that is a generic idea, puts some operational terms to it, and all of the sudden it's now owned by somebody.
And I think we have to be clear that if we are talking about this, there may be different instruments that allow one to move toward putting down the coding schemes that you want, but I think you have to be clear that the constructs that you are measuring are universal. You can do that across languages, across cultures, et cetera.
DR. IEZZONI: I just wanted to underscore for our subcommittee, this issue about the very kind of care settings where functional impairments may be most obvious, are already having imposed upon them huge data collection mandates. And we don't want to be in a position of advising another one that is inconsistent with what appears to be kind of cast in stone at this point, at least in federal regulations.
DR. LOLLAR: And I honestly don't see us, this is not a different instrument. It is a way of coding what is already being collected.
DR. IEZZONI: Okay, so it would be some other entity back coding data that were already --
DR. LOLLAR: A computer program that says we're going to take the OASIS, and we're going to say if you have this here, it's going to mean that there is a personal care. Again, we can say personal care, but what we are talking about is dressing yourself, feeding yourself, going to the bathroom by yourself. This is not rocket science.
There are different scales that measure that, and we can measure it. If you want to get into the metrics, now you get into the validity and reliability of the metrics. You all can debate that from now until Jesus comes.
MR. HANDLER: I have a comment that is sort of from a statistics and layman's point of view. My father-in-law was diagnosed about three years ago with Parkinson's disease. He has a lot of different things wrong with him about functional disability, and it keeps changing. It fluctuates.
About two months ago he had an operation because he had blockage in carotid artery. After he recovered from that, he had trouble swallowing. Liquid was going into the lungs, which could cause heart failure, because bacteria could form in the lungs, and get into the heart. So that had to be rectified.
What comes to mind is there are so many different things wrong with him, how do you give so many codes on an administrative form? There are so many different things to observe, and they keep changing. From a layman's point of view, it's hard for me to figure out what would need to be put on an administrative form to describe exactly his situation.
DR. IEZZONI: Yes, that's a classic truncation of coding slots problem that has been discussed at length for hospital admissions, where in the past we have found that truncating the number of diagnosis codes means that you only get information about the most acute and serious things that are going on with somebody, and you don't learn about the longstanding diabetes. You don't learn about the longstanding hypertension. Those kind of things simply aren't coded.
But I hope that he's doing okay.
MR. HANDLER: He is.
DR. IEZZONI: Aspirations can be very dangerous things.
MS. GREENBERG: Just in relationship to what you were talking about, about the relationship between the ICIDH and some of the measurement tools such as FIMS(?). I know there are some studies being carried out during this Beta-2 testing period. I know Canada in particular is doing one, comparing some FIMS data to the ICIDH. So hopefully, that will shed some light on it.
DR. LOLLAR: About a year and a half ago Social Security asked us to use some of the projects that we already had looking at participation and environment, to measure activity limitations, and to try to look at that issue. So we started looking at the FIMS to see whether or not it covered all eight of the activity limitations, so that we could get a broad notion of the activity limitations.
And because it really only covers about four of them, maybe five if you push it, we decided to go ahead and take the emphasis from the NHIS, which by hook or crook, did cover all eight of the chapters. So yes, the FIMS does a very good job of two or three of the activity limitation areas, but not all eight.
DR. IEZZONI: Don, I know that you have had this idea for a while about adding ICIDH to physician encounter forms. Have you presented this to physician audiences? And what kind of response have you received?
DR. LOLLAR: I won't name the AMA person that came to me when I talked about the idea and said, this is a wonderful idea. She is one of the major staff members at AMA in the geriatric medicine.
DR. IEZZONI: I know, she would say that, but she's an outlier.
DR. LOLLAR: That's right, she's an outlier. But in fact, we have talked about now getting the next step, which is the folks internal to AMA. As I may have said to you, I think some of the physicians would see it as an opportunity to deflect some of the emphasis around why it is that if you have this diagnosis, and this procedure, why it should take you this long.
So the activity limitation, if the person has that activity limitation, it really does become a buffer to explain from the medical standpoint, why one is spending more time, which is another part of the HCFA form.
DR. IEZZONI: Let me just say that from my observation and experience, physicians now often do not see their patients walking into their examining room, because they only have seven minutes or ten minutes for the patient. The patient is already in there, already undressed, already on the examining table, put there by a health aide. So the physician does not get to see whether the patient can walk 10 yards or 20 yards. They don't see the amount of time that it takes.
So in fact, assessing this objectively, and not relying on a patient's report or a family report will take more time. And so I think that we are fooling ourselves if we think that it wouldn't.
DR. LOLLAR: Well, you notice I did not say physician when I first started. I said health care professional, because my assumption is that you are talking about someone in the physician's office if you are going to do it on a HCFA-1500, if you are going to do it on the patient encounter form.
But see, what I would suggest to you is that if you are assuming already that it's not going to work, that it's going to be too expensive, then you have made your decision. But you have to assume that who is going to do it and how much is it going to cost, then certainly we need to work on that. But to assume automatically that because of the way doctors' offices function now, that we can't do anything.
DR. IEZZONI: I think we just need to identify what the barriers are, and find out whether anybody has had any experience with trying to overcome the barriers.
DR. LOLLAR: Well, that's why we need pilots. I don't have any question at all that we're going to need those.
DR. IEZZONI: Okay, great. We have a whole bunch of people who seem to have joined, so are we sure they're in the meeting. Yes, they are?
Yes, Paul?
DR. NEWACHECK: Don, I wanted to ask you about the national health interview survey data that you have. It comes up with about a 19 percent overall prevalence, which seems to have some face validity in my mind. I wondered if there have been any other data sets analyzed either in the US or elsewhere in the world with trying to measure these activity limitations, and what sort of prevalence estimates are coming out of that? And whether this is kind of in the ball park?
The other question I had was about when you get down to the behavioral component, which is a fairly large chunk, do you think that's going to be more controversial in terms of trying to get physicians or other health care personnel to code? A lot of those are fairly intrusive measures, or could be viewed as intrusive measures. I just wanted to a sense from you about that.
DR. LOLLAR: Yes, I think that's the most -- from my standpoint that and the learning are the most problematic, because on the one hand they are the ones that people report more frequently, and yet that general health care professionals are going to be least likely to really to: (1) want to deal with; and (2) not wanting to deal with it is because it is more delicate in terms of uncovering a learning problem or a behavior problem.
So I think it is going to be more delicate. Certainly, to the extent that health care professionals are medical specialties that are much more involved with learning problems, for example, Alzheimers or stroke or whatever, those are going to be relatively straightforward. Pediatricians will, if there is a learning problem, or by however that comes to them from teachers or parents or whatever, will be able to deal with it.
The behavioral stuff, I think with children it's going to be less problematic than it will be with adults. Psychiatrists will not have any difficulty with it, psychologists won't. I think if you are talking about a family practice person, to the extent that they are going to struggle with putting down a diagnosis of depression or a behavioral problem, that's another one of those things. I don't want to put depression down as a diagnosis, but I might be willing to put that they seem not to be able to be getting out of bed.
Frankly, there is some interesting potential for communicating information through activity limitations that up until this point -- mental health folks know if you put 300.4, if you put depression neurosis, you have now effectively curtailed the person's capacity to have medical insurance for the next whatever.
On the other hand, if you put behavior, which says there is an activity limitation related to behavior, at least at this point that wouldn't be undermining. So you would be communicating without having to do something that is going to jeopardize the person's standing with insurance.
DR. NEWACHECK: So in terms of other data?
DR. LOLLAR: Paul and Gerry are responsible for an activity called DISTAB(?), which is UN disability data. We have been working with the Netherlands, France, South Africa, the UK, Canada, and ourselves around what data is available. Because the UN is responsible for this kind of common emphasis, and of course the UN uses the ICIDH-2.
So they, in their new manual that has just going to be coming out soon, outlines the ICIDH-2, and with specific emphasis on the activity limitations, because globally that's again the most common sense observable, consistent, reliable questions that you can ask in a survey.
Do you all want to speak to that, as to what other countries have those?
DR. IEZZONI: Yes, let's hold off on that until we hear from them, because I want to make sure to move on to Michele.
I just have one final question for you, Don. Don't go away. You said earlier that you are working with England, with Australia, with Canada on some concerns that you have about the Beta-2 version that was passed last April. We all know that the United States has modified ICD to meet our own national preferences, and kind of our expectations about what a data classification should look like for diseases.
Might there be kind of a rump movement if WHO doesn't go with the way that you all hope they will to come up with a US version of ICIDH that would be different from the WHO version?
DR. LOLLAR: Well, I'm just a jive little CDC person, and these folks are the Hosses.
DR. IEZZONI: Already, maybe we should let Paul and Gerry respond to that question.
Paul, you were smiling. You look like the Cheshire cat.
DR. PLACEK: I don't see that happening. That was a Canada advised to putting up the boundaries between activities and participation. WHO has responded with their own version of a similar one. It's a little more refined actually, test that will take a little longer to do. So there is a dialogue. I don't think it will collapse over the boundary lines between activities and participation.
DR. LOLLAR: But what I would say is that I believe that what I've given you as the chapter headings are sufficiently consistent with what WHO has done with ICIDH-2 that if those are the only differences that we pushed for in this country, that I don't think we would have difficulty with WHO. I can't believe that taking seeing and hearing and enhancing its priority as an important sensory activity, or that trying to clarify the boundary of personal behavior is going to in some way create overwhelming obstacle with WHO ICIDH-2. Because as you notice, we are not talking about changing very much.
But I think there are some things that I think we do have to make sure. If we don't have public health utility -- I have said this to WHO for the last four years. We don't want to be John Wayne. We want to bend over backwards to be a global player and a contributor and a cooperator, but we can't do stuff that is going to confuse us if we just go straight down the line, and we get data that isn't interpretable to us, or that has several interpretations. Then it doesn't work for us. So we have to do whatever is reasonable for us to make it usable for us.
DR. IEZZONI: Thank you.
Michele Adler from the Social Security Administration is joining us. We were reminded, Michele, by Aaron coming into the room that we have heard from you in the past about how disability was going to be captured on the Census. You are now live in the field, and it seems to be generating a tiny bit of heat. But today, we want to hear about your views about ICIDH, and how that might work. But we also will hopefully be able to hear a little bit about the Census too.
MS. ADLER: To start with the decennial census, we are in the field, and the questions are working fairly well, except for Sen. Trent Lott put them on the -- he only put one of the two questions on the front page of The Washington Post. I don't know how it's going to work out.
In the National Study of Health and Activity that I'm going to talk you about today, we have the same Census questions, and they take a minute to answer all six questions. That's with an extra lead in that we don't have in the decennial. And we have that so we can compare that with other questions. Don Lollar has used the Census set of questions, and they seem to work pretty well. So we will have to see what happens on that.
This is Tom Rush from Social Security. I brought Susan Calasonus(?), she's our mental health specialist; Mel Strumenger(?), who is in charge of facilities and operations; and Michael March, who just joined us and many of us you know, on the National Study of Health and Activity. And we're here from Baltimore. We're late, but we have brought goodies. We have a couple of these slides. Then we have these very slick brochures on the National Study of Health and Activity.
I would like to thank you all for inviting me here today. It's like old times when I worked in the Humphrey Building.
I'd like to tell you about the National Study of Health and Activity, which many of you may know by our old maiden name as the Disability Evaluation Study. It's a very exciting study, and I think it may be the only major study in existence where we will have a lot of information on functioning, but from different perspectives.
We'll have information on functioning from self-reported data, and perhaps proxy data, but basically self-reports. We'll have information on functioning from performance measures. And we'll have information on functioning from a medical examination and medical records, and some observations. I'd like to say that Social Security runs the biggest disability program in the world. But the study was designed around functioning. Functioning is so important, and that's what we are looking at now.
The purpose of the study is to help us plan for the future, and to give us a much better understanding of how health and disability work together. We are looking at disability in a very broad way. So often we have talked about dis-ability, but in this study as you will see, what we are really doing is looking at the bundle of abilities and disabilities that everyone has, and seeing how that affects our day-to-day lives. That's really the purpose of this study.
The National Study of Health and Activity or we call it NSHA is really a national disability survey of working age Americans. Traditionally, working age is defined as 18-64, but we have raised to 18-69 because of the emphasis on older workers. Our goal is really to understand what's going on with working age people with disability. We were in the Office of Policy at SSA, and we have a contract to Westat to collect the data.
We have four major goals. The first is to estimate and project the size, hence the jargoning, of the potential pool of eligibles. What that means is how many people right now, and how many people in the future could get onto Social Security's disability programs just on the basis of their disability.
The second goal is to find out how some people with impairments work, while others do not. And this would lend support for our return to work initiatives. There are a number of theories and speculations about how this may or may not happen, but in the study we will ask people with disabilities, and get some good, solid answers.
The third goal is we are looking at older workers with disabilities, and seeing how the retirement decision feeds into the whole health, work, and disability.
And finally, I think something is hard to find an audience that is really interested in this. I know you are. We want to come up with a really good short set of self-reported disability measures that we can include on national surveys like the 2010 Census, so that we can monitor changes in disability from looking at other surveys. This was a lot of work to put something like this study together, and we just can't continue to do that.
But also a brief set of disability questions on all of our major surveys like the national crime survey, for example, the American housing survey, would do a lot towards understanding how policy changes affect people with disabilities. When we routinely look at how policy changes will affect older people, younger people, blacks, whites, rich, poor, but if we could come up with a really good set of self-reported disability questions, we can add that to the mix.
This is a very complicated study, and it's probably easiest to start with how the main study is going to be designed. The main study will start about a year from now with a telephone screener to 100,000 people across the country, and the telephone numbers will be picked by random digit dialing. We will have a small portion, maybe 10 percent, we haven't really decided yet, where we will have an area sample. That will be used to adjust for people without telephones.
The 100,000 people will be administered a short questionnaire on disability, 20-30 minutes and some other items. And we'll boil it down to 10,000 people. We'll have them come down a few weeks later. We will verify information on disability, and get information on other things that are important on work, work history, accommodations, education, income, sociodemographic characteristics.
Of the 10,000, 5,500 will be selected to come down for a medical examination, which will occur usually a few yards away in a mobile exam center. And they also will be administered functional measures. For the 5,500 people, we'll get medical records for the last three years.
The next step is what makes it unique to Social Security, but also makes it a really unique research project. To get on the Social Security program now, you go to the Social Security field office, fill out forms, bring in medical records. And depending on the state where you live, in your state DDS or disability determination service center, that information is put into a big folder about three or four inches thick, and whether you get on the program or not is decided on the basis of the information in this folder.
So we are doing the same thing with our respondents. The case folders are really interesting. They are not just research, but there is a lot of information on functioning from self-report, in this case when they go into the field office. And we'll have interview information, which will mirror that. We'll have functional measures. We'll have a slide on that in a moment.
But we also will get observations from doctors and interviewers. And finally, the examiners in the DDSs will make a decision on whether each particular respondent could or could not have gotten on the program on the basis of their disability.
The sample, the 5,500 is easiest if you think of we have 500 people already on the program and getting benefits. We have 5,000 people who are not on the program. And Social Security knows a lot about the 8 million people we pay benefits to, but we don't know about the 140 million people in the working age population that we don't pay benefits to, and that's why we're doing this study.
The 5,000 will consist of 3,000 who we think just of the basis of the initial screener could probably get into the program, and 1,500 people who are less severely impaired, and 500 people who either have mild or no disabilities. That will be for comparison, but if you add all those up, you get disability and non-institutionalized working age population.
We are looking at a broad range of functioning. One thing, having worked on all these disability surveys for so long, we tend to have questions on the disabled part of the spectrum. We have been looking at the disability spectrum instead of the ability spectrum. The ability spectrum is much wider. So in the NHSA we have questions on strenuous activity, physical activity. And we also have some strenuous mental activity, when you've really got your act together, and that kind of thing. And you've really got your cognitive abilities in there.
So we have a broader range of functioning, and we're looking at a broader range of ability, and disability certainly is there, but it's no longer the only game in town. And the reason that we are doing that is because in these decisions, which essentially the purpose of the decision to find out how your own particular package of abilities and disabilities fit in with your ability to work and to your day-to-day activities. That's what it is really about.
So we found out that by focusing just on the disability part, we were missing a whole lot. And from the survey we'll have self-reports, functional tests, medical exams, medical records, observations from just about everybody, and finally this DDS assessment. So we will be looking at this picture from a number of different angles.
The functional tests we have are Purdue peg board. We have a grip strength; stooping, kneeling, crouching, which is important in assessing jobs; repeated chair stance, something like this; side-by-side, semi-tandem, and tandem stance; a measured walk, I think it's 20 feet; lifting and carrying with different jugs of water that you lift and carry them as far as you can; shoulder rotation.
For people who are designated as having a psychiatric impairment, we are looking at the CIDI(?) and the SKIDS(?), so we'll have some measures of psychiatric functioning. And we are also doing an experiment on comparing the mental status exam, and the mini-mental status exam.
In terms of stages, remember, we start out with the telephone interview, our initial screener. At the next stage we have this interview with lots of components, which we have technically called the big fat interview, or big interview for short, and the functional tests. Then we'll go on to exams, records, and finally the DDSs.
This is Mel's part. It's the most popular part of the study. These are the mobile exam centers where the medical exams will take place. We have two of these built. They are completely accessible. We built them with the US Architectural and Transportation Board. We had them in over and over again to make sure it was right. The middle door is a wheelchair lift. We have a restroom that is accessible. We have a braille map.
We have gone through a lot of things to make sure that it's accessible for people with physical impairments. And we also have put in some elements to make sure it's also accessible for people with mental impairments. It looks inviting. It doesn't look scary. People don't drive away we hope. That blue line at the top is an awning, so we don't people to be upset when they see it on the outside.
On the inside the walls are white. We have huge windows. We have a lot of light coming in. The Department of Transportation has these restrictions nothing can be more than 53 feet long, 8 feet high, or 8 feet wide, or they won't let us on the highways. So we are limited by that, but inside it's very airy and spacious.
And of course all of our information is confidential, and we can't use it to make claims.
We are in the pilot right now. We have six sites in the pilot. Right now in first two sites we went through the initial telephone screener in Cobb County, Georgia, and Grandville County, North Carolina. That's where we were yesterday. Yesterday at one o'clock was the first time that they started exams. We were on Atlanta TV. It was all very exciting. We're getting a lot of good reaction.
The purpose of the pilot is to find out how we can raise the response rates and test logistics, and prepare of course for the main study, which will start about a year from now, and run from 9-12 months. We've got all kinds of study results. This is going to be a very rich data set to mine methodologically, as well as programmatically.
From the pilot though I have to say that we have a very long screener, and we have a very big interview. We are doing that on purpose so that we could look at different sets of questions and see how they work in pilot, just see if we can pare them down for the main study. But then in the main study we'll be looking at all of these different components, and coming up with a good, brief set of questions.
We have ssa.gov painted on the side of the MEC, but if you go to /policy, you can find out more about this study. Social Security has a new thing called e-news that just started about a month ago. I would recommend you all go to the SSA Web page. You can subscribe to e-news. There is a disability section and a research section. And we're in the disability on March 14, an article called "Just Truckin' Along." We have pictures of the MEC. Or you can email Tom or I and find out more about this study. In your handout there is another photograph of the MEC.
Also the brochures are what the respondents get when they come down for the exam, or when we send them letters.
Are there any questions?
MS. WARD: If you could repeat again for me how you are managing to get people who don't have phones. You went by that really quickly, and I wasn't able --
MS. ADLER: We have a random digit dialing, which comes from a universe of telephone numbers from the phone companies. That is selected only three days before we pick the sample, because of people getting new numbers and that sort of thing. But it is supplemented with TTY numbers, which apparently we didn't realize this, but the phone company has those separate. So we have put them in, so we are not excluding people with hearing impairments.
And then we are picking our sample from there. That's a little simplistic. Area probability samples are done by primary sampling units, and how we are mimicking that in the RDD sample is that counties have certain phone exchanges. And phone exchanges are not specific to certain counties. Like maybe 90 percent of a phone exchange is in a certain country, and we are adjusting for that. So it's kind of like a PSU, but we have overlap.
DR. IEZZONI: Yes, I would also be concerned about the physical examination piece. How you are getting people physically there who have disabilities. For me to arrange wheelchair transport is a really big deal, and it's expensive. And if I recall, you are not paying people?
MS. ADLER: Transportation is being provided. If you want transportation and you need transportation, Westat is contracting with cab companies, including those --
DR. IEZZONI: I know what that means.
MS. ADLER: So it's a matter of scheduling the appointment at the respondent's convenience, not the cab company's convenience. We are doing some incentive experiments in the pilot. There are two experiments. People who come into the big interview will get paid $30 in one site, and then $30 for getting the exam. That's the case now in Cobb County. And then in North Carolina they will get $50 for the big interview and $50 for the exam.
DR. IEZZONI: Michele, Don made a point earlier, and I do want to hear your opinion of ICIDH at some point, because you are somebody who has been so experienced in looking at disability, but Don made the point before you came in that he is interested in looking at activity impairments when people use compensatory devices, such as walkers, wheelchairs, various assisted devices versus when they don't.
And can you just describe whether in fact -- I mean for example, it's irrelevant whether I can walk 20 feet or not, because I don't. And so how are you handling those kind of issues? Are you looking at compensatory devices? And a second question is that Social Security doesn't pay for assisted technology that can allow people to work. So there is a Catch-22 there.
MS. ADLER: Right. Because this is a research project, we are looking at disability in a very broad way. And we end up with having the DDSs assess disability the way they would normally do it, using the information that they would normally do. But the survey is very broad, and we're looking at everything, including assistive devices.
So the ADL and IADL questions are from the disability survey, which include functioning with/without the assistive devices, and all those various combinations.
DR. IEZZONI: You mean the NHISD disability survey?
MS. ADLER: Yes.
DR. IEZZONI: No, that doesn't include whether -- maybe the ADLs include that, but the eight functional status questions do not ask you whether you can do things with or without.
MS. ADLER: Yes, the ADLs.
DR. IEZZONI: The ADLs and the IADLs do, but there are eight functional status questions like can you walk three blocks? Can you climb ten stairs? But they do not ask whether you can do it with a device or can't.
MS. ADLER: In the NSHA initial screener we have that.
DR. IEZZONI: You do?
MS. ADLER: Yes, we have that, and we have a list of common devices that we thought could help with any particular thing. And in addition to the NHISD, we have more activities. We have getting out of a chair. We have lifting heavier objectives. I think it's 50 pounds and 25 pounds, because those come out of the vocational guidelines. So they are expanded. And in the NHISD we had holding a pen or pencil, but in addition to that we have a few more grasping questions that we didn't have before.
So there is quite a bit on assisted devices in terms of functioning. And then we have quite a bit on assistive devices in terms of working.
DR. IEZZONI: Sorry, let me just finish up with this. I just have one more question about it. Michele, if your findings suggest that if people had access to assistive devices, that it would remove the barrier to work, is it possible to change the SSA laws or the regulations governing it to have those devices paid for?
MS. ADLER: I don't know. The purpose of the study is to really look at disability and what is happening now. The DI program was enacted in 1956. We haven't really had the data set to look at these issues since then. We never really had it. And we don't intend that this is a research study that will answer a lot of our methodological questions, which we really need help on, and then be put on a shelf.
So this is really an action study. And that's why we have the DDS involved. So there is going to be a lot of potential. And for work, if you are working now, what keeps you in the workplace. If you're not working, what would have kept you in the workplace? Or if you used to work, what did you have, but it broke down, and what did you need? That sort of thing.
DR. NEWACHECK: Michele, SSA is also embarking on a major new survey of disability in childhood. One of the reasons that the ICIDH is appealing is that it can be used across the age spectrum in many ways, with comparable kinds of indicators collected. I'm wondering, is there any attempt in this survey that you described today, the NSHA, to link some of the measures that are being collected there to what is being collected in SSA's childhood disability survey?
I know that there are differences. There is not a physical examination in the child one, but there certainly are going to be functional status questions, disability questions, a variety of other things. Is there any communication between those efforts and attempts to link measures and such?
MS. ADLER: Yes, there is communication. In fact, that study came out of the NHISD childhood disability part. And in looking at the development of the NSHA, we not only looked at the adult part of the NHISD, but also the child part, particularly in mental and behavioral issues. Because I thought that the child part of the NHISD was just extraordinary. Gloria Simpson and Bruce Stein did a wonderful job of putting those questions together, and we used that as the basis of the SSI study coming out and this one.
We have not gone back and reconvened and looked and seen where we are. But we had the same starting point. Once you get into this, after a while it kind of gets lost whose question came from where. It's all so incestuous. But the children's questions, because they affect age groups who are much younger, and where functioning is so much more difficult to measure, that is something that we have looked at.
DR. NEWACHECK: Just a comment. I was on the technical advisory panel for the SSA child disability survey, and the measures are coming from a variety of different sources, some from the NHISD, and many others from other surveys, and some being developed independently. But it seems to me there is a real opportunity here to try to link measures together that seems to be missing on SSA's part. There is no linkage also with the new national survey of children with special health care needs that is being conducted out of NCHS.
It seems like we really ought to be trying to draw together these surveys, and trying to link measures up.
MS. ADLER: Thanks for bringing that up. I asked them this question a month ago, and that was the answer they gave me. So we'll go back and try to get another viewpoint.
DR. IEZZONI: Don, I haven't forgotten about you, but I do want to see if I can get Michele, from her experience with looking at disability over a number of years, to talk to us about how you think about ICIDH. I know it's something that you thought about a little bit.
MS. ADLER: Yes, I certainly applaud the efforts of the ICIDH. We have had the conundrum here -- I think Ray Seltsar brought this up a while ago -- it would be really nice if you could define disability, and then figure out how to measure it. But we were in a situation -- we started this debate, what ten years ago? -- where we couldn't measure it. We couldn't wait until we defined it before we measured it.
So we started out with the NHISD and a lot of other things, the SIP(?) questions, the Census 2000 questions, and we're in an evolutionary process here. It would have been a lot easier for everybody if we could have sat down, defined, and then measured. But we are defining and measuring simultaneously, and as an iterative process.
In the end I think we are going to look back on it and say this probably was pretty good, because we were constantly learning things. And for example, your question on assistive devices brings out the point that disability is not medical. It is interaction with the environment. When you have the environment and assistive devices changing around you, and jobs changing around you, and you just stick with the medical perspective, you are really not going to succeed. Like who would have thought of the Internet five years ago, and what affect that has had on jobs and disabilities. So what it means to work.
I think we are all changing, and I think the key is we all have to be involved and change together as the world around us is changing. This may just be an area where we have to constantly be involved for the rest of our lives, and for whoever else follows it, for the rest of their lives, because they are so mixed up, and so involved in the real world, unlike other parts of health measurement.
And if we don't keep changing and talking to each other, and reacting to what is really going on, we are not going to be able to do our job. We are going to be dated like a computer; as soon as you buy it, it's outdated. So I don't know exactly. There are different ICIDHs. The NHISD one, we learned so much from that. And the redesigned NHIS, there are some questions that are different, because we didn't know. We didn't think of things.
And then the world is changing around us, and we just have got to keep together and keep at this. We are not going find out the definition. As soon as we find out the definition, the world is going to change, and it won't be the definition. So we are all going to have to keep doing this for a long time to come.
DR. IEZZONI: To what extent does disability determination now rely on ICD-9-CM diagnostic codes, if it does? I have no idea. Does it at all?
MS. ADLER: I don't know if it does.
MR. RUSH: It was about in the mid-eighties that we went to a variant of the ICIDH-9. And so we have a three digit number.
DR. IEZZONI: You mean the ICD-9?
MR. RUSH: ICD-9, excuse me. So it's a variant of it. So that you can go from the ICD to our system, but once you have the code, you cannot go back. So it was shrunk, because at that time there was the feeling that it was difficult to use, and to come up with accurate classifications in the state DDSs.
DR. IEZZONI: So is there any interest in taking from the blue book, the Social Security determination book, taking the various impairments and activity limitations that are described there, and categorizing them by ICIDH codes? Has that ever been something that anybody's been interested in?
MR. RUSH: There has been discussion, but that would be something that people from the Office of Disability would have to talk about. I'm not in the Office of Disability anymore.
DR. IEZZONI: Is there somebody else here who knows a little bit more about that from SSA?
MS. RIMES: Paul, you did a number of trainings for the whole SSA too.
DR. PLACEK: We did ICIDH 1980 training with SSA. We did workshop there and HCFA and NIDR and so on. But my perception is that there is an incredible amount of momentum in the SSA system. It's kind of creaking along day by day. Even if a better idea comes along, it's not going to change because it's a better idea. Because there is a very entrenched and working system at SSA. And that momentum is just absolutely incredible.
MR. RUSH: It's not necessarily entrenched, but it is -- remember there are 54 state DDSs. There are 14,000 disability adjudicators that exist. And so we have extensive PALMS(?). It's a system that is working for the purpose for which it is designed, which is to make disability determinations.
So that change is something which does take a while. If you look at a system, we have to publish in The Federal Register, any changes that we're going through. There is significant comment that comes. It is one that as far as changing quickly to new medical knowledge, it does take a while because the system. It's a legal system that is in place. And there are regulations that regulate that, that really come into play.
So it's not an easy system to change. That doesn't mean it doesn't change. But it is sort of like an oil tanker. It goes two miles before it starts to turn.
DR. IEZZONI: Right, that's inertia in the best physical sense of the word. That's how I meant it, from the laws of physics.
MR. SELTSAR: Ray Seltsar, currently retired. I am forced to say something about this, because ten years ago I chaired a task force which SSA asked the Public Health Service to convene to look at improving the medical criteria for disability determination.
The one recommendation that came out loud and clear after -- and this was I think, a fairly comprehensive type of task force which involved many, many people, and many organizations -- was a recommendation that the US adopt the ICIDH as a classification system which would enable the Social Security Administration to get its act together, because it was basing its disability determination largely on what was called listing of impairments, which is an ICD-oriented classification system.
But when the final determination was being made, it was based on function. And the ICD told them nothing about function. And some Supreme Court decisions came along which gave them real problems in using ICD codes. So that the old ICIDH, and even the better one now is a functional classification system which basically would have solved, and is in the position of solving some of the problems that the Social Security Administration has had through the years in trying to provide compensation for people who are disabled, but with a definition of disability being a functional term, and yet with the classification system they have in not being functional.
We finally have a classification system which would allow them to do this. I certainly encourage you to look very closely at it.
DR. IEZZONI: Can you give us some insight as to was it the kind of tanker analogy, the physics inertia analogy as to why this didn't happen?
MR. SELTSAR: Well, it didn't happen for a number of reasons. One of the reasons was that both the commissioner of Social Security and the assistant secretary of health who was responsible for getting this through, both left the government at the same time when they report came through, and nothing ever happened.
MS. ADLER: Not because of the report.
DR. IEZZONI: So it's an institutional memory problem?
MR. SELTSAR: No, it's just one of these things that happened. Many of the ideas of that report were resurrected in the reengineering plan of Social Security. And I think that what you are seeing now is a study which goes far beyond our expectations. We did design, incidently in that report, a set of research proposals to get Social Security and the Public Health Service working together to establish a scientific basis for disability determination. We were asking for the enormous --
DR. IEZZONI: Thank you. Don?
DR. LOLLAR: I just wanted to respond that the SSA listings, really as I perceive them, is a combination of emphasis on diagnostic codes, impairments, and functional activity limitations. In fact, I believe Bill Frye(?), who is now at Westat, has done the ICIDH mapping of listings that I think you are asking about.
DR. IEZZONI: I'm just asking whether Michele thinks that kind of thing would be useful.
MS. ADLER: It's been done, but it hasn't been implemented.
DR. IEZZONI: Would it be useful? What are the barriers to implementing it? What would be things that would need to be tweaked about it to make it useful?
MS. ADLER: That's certainly something we can look at. The listings are done nearly from a process point of view. We have all these DDS workers, and we have millions of claims coming in every year. As someone for example, has something on the listings, then their claim is adjudicated faster, and it can go on to the next thing.
So we find out nothing about secondary conditions. We find out nothing about work, because we don't get to that point. And in the truly process mode, you want to do things as quickly as possible so people can get these checks as quickly as possible. And that's really what it is about.
But the NSHA study is unique, and we are going to find out not only perhaps the disability that would have gotten them on the program, but other things, other aspects of their lives, and see how that fits into it. So this is a tremendous opportunity to break out of this process mode. I loved the oil tanker.
While the oil tanker is going down for two miles to turn around, that's our little window. Do you think you could get them to take longer to turn around? Because during that time we could be looking at these other things.
DR. LOLLAR: I just wanted to ask about the formula for making the decisions from the 100,000 down to the 10,000 who have a disability. You are doing something to, on the basis of telephone questionnaires, to make that. Is there a formula or weights, or is that done a priori? Is this a bell curve?
MS. ADLER: There is a formula. It's a very, very complicated formula.
DR. IEZZONI: Kind of like going to phase 2 in the NHISD, which was very complicated.
MS. ADLER: Right. And there are a number of consistencies and inconsistencies taken into account.
DR. LOLLAR: And has the Cobb County telephone survey already been completed?
MS. ADLER: Yes.
DR. LOLLAR: So now the Cobb County and the North Carolina telephone survey have been completed, and folks have been selected from former Rep. Gingrich's district.
MS. ADLER: Bob Barr's district. He is putting us on their Website. Are you in Cobb County?
DR. LOLLAR: No, another district.
DR. IEZZONI: It sounds like you should go by the truck anyway.
MR. HANDLER: Is that near Raleigh, North Carolina?
DR. LOLLAR: This is the Georgia one. There is another one in North Carolina.
MR. HANDLER: Where is the North Carolina one?
MS. ADLER: It's near the Virginia line off of Highway 85. Tom was there yesterday. Is it near anything?
MR. RUSH: It's in Grandville County, so Oxford and Henderson would be the closest areas. About an hour north of Raleigh.
DR. IEZZONI: Michele, we as taxpayers and as scientists are very interested in hearing about this study as it goes along. And I'm sure that we will continue to be updated on it. It sounds like a really amazing opportunity to learn.
MS. ADLER: And I would like to say that I guess between now and about September or October are the times that we'll be putting the questionnaires and instruments for the main study. So this is an important time to hear your comments, and for you to help us so that we can look at the kinds of things that you are interested in too. A year from now we will be out there, and it will be too late. But we really encourage your participation.
DR. IEZZONI: Paul, a final comment on this?
DR. PLACEK: I like your accessible truck, by the way. I have two questions. One, a global question, other than the Census question that you mentioned, and also about can you weight the sample of the national estimates. On the global question, you mentioned the Census question is six items and takes a minute. But are there any other kinds of global questions you are testing out? Those one or two questions that might capture whether a person has a disability or not, as you will reliably check in great detail later. And of course, can you wait to sample through a national estimate of any kind?
MS. ADLER: Well, on the weighting first, in the main study it will be weighted up. It will be national estimates. In the pilot, we can't do that. We have two urban, two rural, and two suburban communities in the East. The reason they are all in the East is so that we can see what it's like when this MEC travels up the road. We don't want to have a site in North Carolina, and the next one in Tucson. So many things can happen on the road.
As far as global questions, we are also looking at the perception of disability questions. But when within questions we are looking at -- like within the ADL set, and within the IADL set, if we can shrink that. I don't know if there are any global questions to test. If you can come up with some -- we are in the pilot, but we are also testing various things in cognitive labs along the way. So if you have something, let me know. There may not be a whole really tiny holy grail. It may have five or six questions in it.
DR. NEWACHECK: Just one final question. A big chunk of the disabled population is being left out here, those in institutions. Does SSA have any plans to do any surveys of the institutional population?
MS. ADLER: Well, in institutions there are -- the estimate from the 1990 Census, the numbers are going down. I think in the working age population it's about 700,000. And the number is going down every day. In the area sample that I have glossed over, it's an area probability sample which has a couple of reasons. One is to adjust for people without telephones. The other purpose of it is it will be supplemented with group homes. And we are defining group homes in the pilot now as having 10 or fewer residents. We may expand that in the main study.
But that's so that we will include people in group homes in the sample. If we did it just the traditional way, it's true we might get somebody in the area sample, and we might get their phone number, but we couldn't really count on it, so that's what we are doing. That's our approach.
DR. NEWACHECK: But there are no other separate surveys that are envisioned?
MS. ADLER: No. And in the NHISD we tried that. That was also the civilian nonstitutionalized population, but with the group homes. And we did an early cost estimate, and it would have doubled the cost of the study. It's the same thing here. So anyway, that's the basis for that decision.
DR. IEZZONI: Thank you very much for coming, and thank you to your colleagues from SSA for making the trip from Baltimore. We appreciate it.
MS. ADLER: Sure, good to be here.
DR. IEZZONI: All right great. Thanks.
Okay, does the committee feel the need to take a five minute break, or should we move right along with Paul and Gerry? We can't take longer, because we have lunch in an hour -- half hour.
DR. NEWACHECK: Just keep going.
DR. IEZZONI: Keep going? Okay.
Gerry, were you having technological challenges over there?
DR. HENDERSHOT: Yes, I am.
DR. IEZZONI: And did they get themselves worked out? Okay, I guess you're first up.
DR. HENDERSHOT: I will be making the presentation this morning. Paul and I will be sharing the presentation at our scheduled time tomorrow, or rescheduled time tomorrow.
I'm Gerry Hendershot. I work for the Division of Health Interview Statistics at the National Center for Health Statistics. I have been involved in work on the ICIDH for about 10 years. I was glad that Ray Seltsar was here and gave you a little background on the Public Health task force that he chaired. He invited me at that time to chair a work group on defining classifying disability. It was in connection with that work group activity that I became involved in the ICIDH, and have remained involved ever since. So I give thanks to Ray for having started me out on this career. It's been very interesting.
I want to do two things in the time available to me this morning. One is to do a Power Point presentation, which will be the introduction and background of the ICIDH that I wish you had had before this very interesting discussion we have had so far this morning. But this may fill in some of the gaps that you were feeling as you have tried to follow the earlier discussion.
The second thing I'm going to do is to invite you to participate in the part of the Beta-2 testing of the ICIDH, and give you some instructions on how you can do that, and hope that you will follow through and complete that task. It will also give you a way of learning better what the structure and the purposes of the ICIDH are.
So we are talking about the ICIDH-2. It will be called, when it's approved, if it's approved, the International Classification of Functioning and Disability, ICFD. I tried to use that language throughout this presentation. The way that works, this is the actual classification in its current form, Beta-2 form. You will see it says ICIDH-2 on the side. That is remind people where it came from. But its actual name is International Classification of Functioning and Disability.
This presentation, by the way, was prepared by WHO staff, and I have modified it and tried to update the language. Because it is in revision, the language is changing periodically. I'm trying to use the most current language.
It's part of the WHO family of international classifications related to health, the most important and best known of which is the ICD. But this is a second classification in that family, and as this graph indicates, a lot of other classifications spin off from that, which are part of the family. So it's got a provenance. It doesn't just come out of the blue.
Why do we need an ICIDH? Because there have been changes in health issues from concern with acute conditions, to concern with chronic conditions. There has been a change in a focus of interest from disease to the consequences of disease. And there is a need for an international common language for the consequences of disease. Finally, a new kind of classification framework for thinking about disability was needed to serve the needs of persons with disabilities.
One of the characteristics of the provision process is that it has attempted successfully by most estimates in involving people with disabilities, and advocates for people with disabilities in the revision process. That has been very important to the effort.
Diagnosis alone fails to predict. This point has been made earlier. It fails to predict service needs linked to hospitalization, level of care, and so on. And WHO here had provided some very graphic citations to document those assertions. But diagnosis plus disability can predict much better in the same areas, and again, they have provided some citations to back up those assertions.
I didn't bring with me copies of this Power Point presentation, but I will make them available for the record.
So to make a complete assessment of health and illness, we need not only diagnosis of the health condition, disease, disorder, injury, and so on, which we have through the ICD, and which has been very useful and influential, we also need an assessment of functioning. And using the ICIDH-2 or ICFD terminology, we need assessment of impairments, activity limitations, and participation restrictions.
Another way of thinking about the issues involved here is to make distinction between what people refer to as the medical model and the social model of disability. You have the medical model on the left of the screen in which disability is regarded as a personal problem, a problem that the individual has, and it's appropriate to address that problem with medical care.
That requires some adjustment or change or treatment and response by the person themselves. And this is an area in which health care policy is appropriate to consider at a higher level of generality.
And that is contrasted with the so-called social model in which disability is viewed as not a personal problem, but a social problem, which is responded to by integrating the person into the biological, psychological, and social milieu in a successful way. In this view, the disability environmental intervention of some kind is an appropriate way to address the problem. And in this model, law and social policy, rather than health care policy are the appropriate societal responses to the problem.
So we've had these two different kinds of models of disability and some in between. And what the ICIDH-2 or the ICFD tries to do is to bridge the gap or the different perspectives of those two models, and say it is both personal and social. It is both medical care and other kinds of integration. It's both personal adjustment and environmental intervention. And it's both health care policy and law and social policy.
So this is a classification that attempts to cover both the medical and the social and political aspects of disability in one comprehensive theoretical framework and classification. It is expected to have applications, and in indeed in its first version did have applications and statistics of various kinds of health and medical research as a clinical tool, social policy tool, and in the field of education. So it's intended to serve a wide variety of actual applications on the ground, as it were.
This is an attempt to represent the conceptual framework in a schematic way. At one of the meetings there was a long debate about whether or not we should try to show a picture of the conceptual framework or not, because pictures can be easily misinterpreted. And the picture that was in the first version of the ICIDH was actually misinterpreted rather widely to the detriment of the success of that classification.
So it all begins as it were, with a health condition, disorder, or disease, and sets in process a sequence of events if you want to think of it causally, in which first of all, the body is affected, then the person is affected, and then their participation in society is affected. Those are the three main dimensions of the ICIDH framework -- the body function and structure at the body level, the activities at the person level, and social participation at the societal level.
And the new version of the ICIDH attempts to be neutral with respect to whether activities are positive or negative. But when we are applying it in the case of disability, we are usually thinking about the negative aspect of body function and structure, which is called impairment. And the negative aspect of activities is activity limitation, and the negative aspect of participation is participation restriction.
The framework also incorporates the effects of the context, which includes both the environment, social and physical, and also characteristics of the individual person. So the arrows in this picture are drawn to go every which way. And that is to get away from the notion that there is one causal sequence involved in these processes, which was the way the ICIDH 1980 was misinterpreted. It doesn't actually say that, but it was misinterpreted to mean that. In the ICFD the emphasis is on the interaction of these various components of the total system.
Key concepts of functioning disability are function structure at the body level, activities at the person level, and participation at the societal level. And as I have said, we have impairments at the body level, activity limitations at the activities level, and participation restrictions at the participation level.
The lower case items below each of those broad dimensions of disability, in the case of function, structure, severity, and localization, these are what I will later call qualifiers. We will get a code, and then you'll have a qualifier to the code. The qualifiers of function and structure are severity and localization, for activities, severity and assistance, and for participation, severity.
This is an attempt to illustrate briefly how the different dimensions of disability might come into play in thinking about attention deficit disorder. The impairment here would be poor attention, problems in concentration, increased arousal. At the activity limitations dimension the activities might be doing homework or waiting a turn. And a participation restriction, being defined socially as a problem kid, and exclusion from class activities.
Some features of the classification, health-related experiences. We are not concerned with all kinds of experience. These are health-related experiences. It is descriptive, not analytical, if those terms mean something to you. This is not science as it were, it's classification.
All people are included, because disability is a universal human experience. And as I said earlier, the attempt was to think about activities for instance, in a positive sense, as well as a negative sense. Although in application to disability issues, it will be the negative sense that you are usually thinking about it.
What is classified is not persons, but the attributes and experiences of persons, or the situations and circumstances of persons. We are not talking about classifying people. And this is one thing that the disability advocates and persons with disabilities contributed to the revision process. We are not classifying people.
Body function, body structure, activities, participation, those four dimensions are conceived as distinct, but parallel classifications. One of the big problems in understanding the whole system is I think an assumption that you have to classify everybody in all of the dimensions of the classification. That is not the case. Each one is seen as an independent classification system, which could be applied for a particular purpose without using the other systems. But they are parallel in structure and are related to each other at a conceptual level.
Some things about appearance, they may be temporary or permanent, progressive, regressive, or static. They may be intermittent or continuous. They are not the same thing as etiology. They are not the same thing as health conditions. They are consequences of health conditions. And having impairment does not indicate illness or sickness in the traditional sense of those terms.
Activities are associated with every day life, and can range from simple to complex. It's the actual performance of a task or activity, not the potential. The difficulties which are mentioned in the classification system can be qualitative difficulties or quantitative difficulties. And in the area of activities the qualifiers are difficulty and assistance.
Participation is sort of at the frontier of disability within this conceptual framework. It's the lived experience of people with a health condition, impairment, or activity limitation. And we are talking here about an ecological or environmental interaction model.
Taxonomic principles -- this is a classification system. It is not a measurement device. It's a classification system. And certain taxonomic principles were incorporated into the design and development of the classification. The order is generally from simple to complex within the dimensions.
Items within the classification categories within the dimensions are grouped by similarity and essential properties, and the overall structure is hierarchical. It starts with large categories, and within those looks at smaller or narrower categories, and within those even narrower categories.
The WHO official who leads this effort is Beterin Ustin(?), who is Turkish. He loves metaphors. He loves to talk about this as tree, branch, leaf structure. And this meant to kind of illustrate what is going on here. The term at the top says disablement. We have struggled with an overall term to apply to the whole classification. Disablement was the last term. It's now disability. So we are using disability to refer to the overall several dimensions.
But under that you have impairments, activities, and participation. Then you have domains within that, and then facets within that, and finally categories at the lower level. So it's all hierarchical. Depending on the detail of data you have, you can classify people in broad categories, narrower categories, or very fine categories. All of the situations should be encompassed by this hierarchical structure.
The classification manual itself -- this is the classification manual, and I just wanted to describe how it is structured. It has an introductory chapter which is important to read, because that sets out the conceptual framework. Then there are one and two level classifications, that is the broadest category, which is a single digit, and then somewhat more detailed categories, which get you into two digits, and then the detailed classification.
There is a different classification for each dimension, here called body functions, body structure, this B&S, activities, participation, and environment, A, P, and E respectively. So each detailed classification code is preceded by one of those letters to tell you which dimension you are in.
Then within a dimension there are chapters from one to I don't know what the largest number is. But this divided the content of that chapter into broad categories. Then within those chapters there are blocks or major categories which group together things which are related to one another. And then finally you get down to detailed codes and qualifiers.
The detailed code consists of an letter indicating the dimension, and then a number, which indicates a coding classification or category. It has a label. The illustration I'm using here was A410 is walking activities. That's followed by a brief definition, which in this case is "coordinated actions of advancing on foot, step-by-step, in a manner in which at least one foot is always on the ground." I don't know where that definition came from, but that is the ICIDH-2 definition of walking activity.
And then there are a number of inclusions and sometimes exclusions which are examples of things that you would include or would not include within this category. The inclusions in the case of walking activities -- this is not the whole list, but it's a partial list -- strolling or sauntering. My wife says I saunter too much. She refuses to walk with me anymore, because she doesn't saunter.
You can be walking forwards or backwards on smooth or rough surfaces, with the help of a cane or prostheses or not. So this is an attempt to give some guidance to the person classifying the information, to make sure they are in the right place. The code is structured like this, as I said. It is preceded by a letter which tells you what kind of consequence or what dimension you are in, and then a number indicating a general ICFD code. Then a more detailed number. Then a decimal point.
After the decimal point are qualifiers. There may be one qualifier or two qualifiers. They give you some additional information, and we'll talk more about that in a minute.
So that's the structure of the classification. One thing I have given you, this is selected pages from the classification, including the title page. Then you see the first page of the introduction. Then you see the one digit classifications for body function, body structure, activities, participation, and finally environmental factors.
Then you see the first page of the activities dimension. That tells you what the qualifiers are for the activities dimension. There are two of them. The first one is what they call a uniform qualifier that is used in other dimensions as well, which gives you the amount of difficulty the person has with whatever activity it is.
The second qualifier is the assistance that they receive. You look at that little paragraph in between the first and second qualifier, if only one qualifier is used, it implies the level of difficulty without the use of assistive devices or personal help. If the first qualifier is used with the second qualifier, then the first qualifier now implies that this is a level of difficulty using an assistive device or personal help.
This is an attempt to find a way to code with or without the use of assistance of personal care or an assistive device. So you have those options in the coding structure, depending on the information you have, and the purposes of your classification.
Then the last page in that handout is the beginning of one of the chapters in the activities dimension, Chapter 4 on activities of moving around. And that is where you have the walking and related activities, which was one group. You see the walking activities, and that's where I got the illustration that was on the slide earlier of the definition, and some of the inclusions for A410, walking activities. And then you can go into finer detail within A410 to look at particular walking tasks.
This is the manual, and this is the main product of the revision activity, but it's not anticipated to be the only product. This is what in this slide called the main volume. It is anticipated that there will also be clinical descriptions and assessment guidelines, something appropriate for use in clinical settings.
I didn't make copies of this, but this is the current version of the clinical version of the ICIDH-2. It's a checklist. It's a reduce list of selected categories in each of the four dimensions. So the simplest way to use this would be just to check in a column opposite a category in whichever dimension you are looking, and that would meet the requirement for the clinical use of the classification. This would be simplest kind of application that is currently available.
And there might be standard measures. These are not measures, this is classification, a point that's been made earlier, but different measures could be developed for us in research, and then we could also develop questionnaires for surveys based on the classification. There may be special adaptations needed for special population groups, or applications such as children or lay and community reporting. So that's a very quick overview of the structure and purposes and potential applications of the ICIDH-2.
The second document I gave you says on the top of it, "NCVHS Subcommittee on Populations, ICDF Beta-2 Test Study 3." This is something which I hope you will find the time to do between now and tomorrow morning. What I have done is give you a case. It came out of The Washington Post a while ago, a couple months ago, a man called Gregory Curry. But that's your case, and I have circled parts of it that are applicable to reaching some coding decisions for this case.
Then I've given you the title page of the ICD, and a couple of pages from the ICD where I have indicated -- you find out in reading the case that this man had diabetes, and he has also had a leg amputated. So I have shown you where in the ICD you might find codes for those things.
Then we go into the ICIDH. I've given you the title page of that. Then I have given you Form A. This is a form that is actually being used in a participation test of the ICIDH-2 called the Study 3, which is coding cases and then recording the results of those codings. So what I'm going to ask you to do is to record the results of the coding of this case on that form.
Then I have made it really easy for you, because I have included only the pages from the manual which contain the codes that I suggest would apply in this case. And I have also indicated by an arrow, where you go next. So you start with body structure, and I have indicated you go to Chapter 7. I have given you the first page of Chapter 7.
If you turn over to the second page of Chapter 7, I have indicated what you might code in terms of body structure. It's not going to be this easy all the way through, Lisa, so pay attention.
DR. IEZZONI: I'm just thinking that you are biasing the audience a little bit here.
DR. HENDERSHOT: Then there is the activities dimension. I have indicated which chapter to go to. You get the first page of the activities dimension, which tells you what the qualifiers are. Then I have referred you to Chapter 4 within the activities dimension, and the walking activities within that, and indicated what code you might assign for this case.
Then we go to participation, same thing. I tell you what chapter to go to. The chapter is work and employment, and I have indicated what codes might apply there. And finally, environmental factors, and pointed out what the qualifiers, and what chapter of the environmental dimension you might go to, to find a code.
So I have indicated what the codes are. You've got the form. You write in that code. I haven't told you what the qualifiers are. So you have to look at the case, look at what the appropriate qualifiers are, and make a coding decision about what qualifiers should be included.
DR. IEZZONI: In ICD training, Gerry, there is not always a right answer. There is something called the art of coding. Now are you assuming that there is going to be a right answer here, or is there going to be an art of coding here?
DR. HENDERSHOT: Well, in this illustration -- I shouldn't call it an actual coding case -- I have made some coding decisions myself. But you are absolutely right --
DR. IEZZONI: Because that gets to the issue of reliability.
DR. HENDERSHOT: Right. And in the actual application of Study 3, Fran Shavary(?) is here, who works with me in HIS and has lead responsibility for designing and analyzing the Study 3 activities that we are going to be doing in the United States, and Paul is going to be talking more about that -- Paul or I will be talking more about that tomorrow.
So we are actually going to be doing this, but in the actual Study 3 cases that we'll do, people will just be given a coding manual or a form of the coding manual, and will make new coding of cases, and then we will analyze it later to look at reliability, and find out, are these reliable.
You have to remember, the people who will be doing this in virtually all cases will have very little training or experience in using this classification. So we're starting with an assumption that there will be fairly high unreliability. It took ICD a lot of years to get to the point where you could code things reliably, and still, it's not 100 percent. So this classification is at a much earlier stage in its development, so people are not trained. They don't have experience, so we are going to have a long way to go to get to reliability.
DR. IEZZONI: Very good. Thank you, Gerry. Yes, it probably would have been better if we flipped a few presentations around, but we are glad to have had this introduction.
Paul, Kathy, Elizabeth, any questions at this point?
DR. NEWACHECK: First, Gerry, I think the slides were great. I wonder if we could get a copy of those from you?
DR. HENDERSHOT: I don't have them today, but I'll make sure that Patrice gets them.
DR. NEWACHECK: They help lay out the conceptual framework in a way that we haven't seen in some of the handouts we have received so far. So that would be really helpful.
But I wanted to ask you, in terms of operationalizing this concept for administrative records, which is the subcommittee's interest, do you have thoughts about that? We asked Don about that, and he gave us his ideas about how you might do that in terms of having a couple of different fields, that sort of thing. Do you have any thoughts about that?
DR. HENDERSHOT: Well, I agree with Don that the dimension in this classification which is most appropriate for the uses you have in mind is the activities dimension. So you would be looking at activities limitations. And there are seven or eight chapters in the activities dimension, and the simplest thing you could do is just put a 1-7 if there are seven chapters, and just a checklist kind of thing; yes, this person has a limitation or not. That's would be the simplest kind of thing to do, and I think that would be useful.
I think that's what Don was suggesting as a minimum kind of thing. You could do lots more, but I think that's the simplest thing you could do, and I think that would be useful.
DR. NEWACHECK: But not to have all eight fields on administrative records, but rather just if there was something that showed up in one of those fields, correct?
DR. IEZZONI: Yes, because, Don -- you're still here. Can you come back to the mike? I think that you were suggesting for two codes, having complete information around the code, not just a binary yes/no at the more global level. You were talking about having two fields, where you would have the most detail.
DR. NEWACHECK: The two most significant activity limitations.
DR. LOLLAR: Right, with the idea that it really isn't too much to ask a health professional to be able to remember that communication is number two, and learning is one.
DR. NEWACHECK: So that's different. That's more the ICD model where you list the primary diagnosis and the secondary, in that concept. Whereas, Gerry, what you were suggesting is perhaps a checklist where you might have all eight categories and a yes/no or some severity indicator attached. So there would be two different models.
Can I ask both of you guys though whether or not think that this is too narrow of a conception? That is, just sticking to activities. Obviously, the ICFD has a lot more than just activities. And if we are really going to implement the system, would it be myopic to focus just on this activities classification, or is it a good starting point?
DR. HENDERSHOT: I would say it's a good starting point, but I think as Don was saying earlier, there are lots of instruments in use which implicitly or explicitly are looking at what the ICIDH calls activity limitations. So there is a lot of measurement being done already that could be classified to this classification, so it's the easiest place to start.
But the real contribution I think of the ICIDH-2 or ICD to our thinking about disability is the participation dimension and the environmental dimension. Sort of the long-term pay off of adopting this conceptual framework is in those dimensions of the classification. I think the case which I gave you from The Washington Post is a good illustration of that. If you just did an activity limitation or ICD codes, you are really missing the point of that case.
The participation and environment codes I think pick it up pretty well. So you would have a single code that would tell you something about this case that is useful to know in the environmental area or in the participation. The two of them would be better. If you had the qualifiers, that would be better yet. But just the ICD codes or the impairment and activity limitation codes don't tell you about that case, what's really the most important. The participation and environment things add that, and I think do it pretty well.
DR. LOLLAR: To the extent that conceptually there are two competing notions about disability. ICIDH talks about disability as a universal phenomenon. There is another framework, however, which is disability as a minority status. And I think what you get into with the participation in the environment more than anything is the minority status emphasis where you are talking about whether or not people with limitations or impairments have the opportunity to participate in society equally with everyone else.
So consistent with the Healthy People 2010 objectives where we talk about eliminating disparities, the issue of being able to identify people with disabilities in whatever mechanism we choose to do that, and it would be nice if it were consistent, allows you then to look at outcomes such as participation, and to look at environmental factors.
To the extent that we try to include those in administrative records, I think that's a wonderful idea. Certainly, we could go on about that. I guess it is myopic to just look at activity limitations at this point. It is substantially, however, I think more realistic and practical to just be that myopic for right now. It doesn't mean that the subcommittee should not recommend further stuff, but I would suggest that as a starting point.
DR. NEWACHECK: Maybe we can put that on the table for committee discussion later.
DR. IEZZONI: Yes, that's a good idea.
Yes, and then I have a question. Are there any other questions? I just want to make sure we get everybody before lunch.
MS. GREENBERG: As I think Don alluded to, one of the real issues right now in the countries that he mentioned, including those in the North American collaborating center, the US and Canada is the boundaries between activity and participation. And we are going to be doing some studies around that.
But there actually is some what are considered complex activities, are really a form of participation, such as going to school, going to work. There are actually some of those activities are included in this activity dimension. And they are seen more as participation, because they do involve society, and this is being debated, et cetera, but I won't go into that at this point.
But I think part of the reason that they are included in activities, my belief at least, is that if you were only going to use the activities dimension, which I these can stand alone to some degree, as Gerry said, but you get a richer view of the overall conceptual model if you collect all of them.
If you would at least capture some of those aspects of participating in society -- going to school, going to work, using public transportation, et cetera. It is a little muddy from a classification point of view, because you can put things into two places, an activity and in participation. And actually when we have had some discussions with WHO, they have taken the position, well, if in the US and you ask about would you use a US version? But if in the US you chose not to use those aspects of the activity dimension, fine, and only use them in the participation.
So I think that it's true that activities without those complex activities are more limited if you wanted to get a better view of how a person really participates in society. And I guess there are different options that you do or not.
DR. LOLLAR: If you start doing that in an administrative data set, then you get people trying to decide which is most important to put. And you are again, mixing constructs that I think make it much more difficult to really know what it is you are getting. And the other thing that you don't get is if you put work into the person level, then you are confusing the relationship of the environment -- or because of the environmental modifications.
And when you put a participation like work into an activity, then you diminish the potential for looking at the environmental influence.
DR. IEZZONI: But Marjorie seems to have a much better understanding of what exactly is in there than the rest of us do. So maybe it would be good to see some examples of those codes that you thought are kind of on the boundary, and could go into either pot.
DR. LOLLAR: Well, that's Chapter 8.
DR. IEZZONI: Let me just ask Gerry one final question. We have got to at some point -- Susan and Carolyn -- involve the privacy committee in some of what we are doing. I'm actually quite happy to see that they are now calling this thing the International Classification of Functioning and Disability, because back in 1997 I think they were going to call it something having to do with abilities and participation.
There was some kind of language in there that obviously came from some community that didn't speak the same language that I think the average person speaks, and it was getting extremely cumbersome. So I'm kind of relieved to see that they have just kind of said, okay, let's just call it classification of functioning and disability.
But you alluded to earlier, Gerry, that "the disability advocates community" or people with disabilities have really helped hone some of your perceptions of how to classify certain things. And do you have any insight from the experience working with them, or observing them, how they might feel about adding an ICIDH type of activities code for example, or your checklist of seven dimensions to a routine encounter record?
Do you think that they would view it as giving sufficient information that it was worth breaching their privacy? Do you think they would view it as a breach of privacy? How do you think they would view it? And we should probably try for the July meeting to hear from some of these folks themselves.
DR. HENDERSHOT: That's hard to answer. There are two levels I think, to consider that. One is the institutional level or organizational level. There are a lot of disability-related organizations, some of whom would welcome the inclusion of that information on the administrative forms, and some of whom probably will not.
DR. IEZZONI: Would you be able to identify for us, so we can think about bringing in some people for July to speak to us on both sides?
DR. HENDERSHOT: I think we can do that. But I would say that those who would be opposed to it at this point probably have don't know much about the classification and its purposes, and the role of people with disabilities in developing it. So I think we could work with them to resolve the problems that they might have with it.
The other level would be the responses of people with disabilities on whom this administrative information is being collected. And I don't know what the response would be. I was surprised at the response -- at least the reported response in the newspaper to the questions about disability in the long form of the Census. It may have been a very small proportion of people, I don't know, but some people apparently felt offended by those questions. So I'm sure people would be offended by this application, but I have no idea.
DR. IEZZONI: Does anybody have that Washington Post article that describes that? I think, Aaron, you mentioned it. Because I think it would be good to hear, Gerry, what is at least being said about people's discomfort with that on the Census.
MR. HANDLER: I know -- I forget what her name is. She's like the communications officer of the Census Bureau.
DR. IEZZONI: I think we have to anticipate that there might be issues around privacy. So involving the privacy subcommittee to help us think about something in July might be something that we should think about.
Marjorie, do you think that's a good idea? Okay.
Gerry, this was very, very helpful. We are going to start tomorrow morning at eight. I know Paul lives on the Eastern shore of Maryland. So we'll hear hopefully from Paul, but maybe not. But thank you.
DR. PLACEK: For those who won't be here in the morning, I have three handouts.
DR. IEZZONI: That you can make available this afternoon?
DR. PLACEK: Yes.
DR. IEZZONI: Okay. Great, well, why don't we take an hour for lunch and come back at 1:30 p.m. then? Thank you, Gerry.
[Whereupon, the meeting was recessed for lunch at 12:35 p.m., to reconvene at 1:30 p.m.]
DR. IEZZONI: Thank you. Let's reconvene for the afternoon session. We have three speakers who I think will give us some interesting comments, from what I hear, on ICIDH.
Dr. Holm, could you just introduce yourself really briefly.
DR. HOLM: I am a faculty member in the School of Health and Rehabilitation Sciences at the University of Pittsburgh. I am an occupational therapist. And I do clinical research, and also teach some graduate students.
I think I have been asked to speak here today based on the premise that unless we learn from history, we are doomed to repeat our mistakes. My presentation today is based on an analysis that compares and contrasts four functional assessment instruments. And when we did this, it revealed usage of dissimilar terms, definitions, and measurement scales.
Health care personnel who work in subacute units, rehabilitation hospitals, home health care, and nursing facilities have an ethical obligation to understand the terms, definitions, and scales that are used in functional assessment tools so that when they rate a patient's or resident's functional status, they can do so accurately and expeditiously.
And this is because both in research and in clinical practice functional status is often viewed as a marker of overall health, as an indicator of the ability to live independently, an indicator of the need for assistance. It is linked to the level of care and resource use that is needed for quality care, both in hospitals and in long-term care facilities. And it also may be used as a predictor of morbidity and mortality.
I realize this is one way arrows. I just made them that way, because that's what I could figure out. When function is assessed, it can be assessed at many levels. Using the ICIDH-2 model a functional assessment could focus on body function and impairments in function or structure such as impaired strength and range of motion, or on activities and activity limitations such as managing finances, or limitations in bathing, participation or restricted participation in home management work or community activities.
The analysis that I'm talking about today focuses on that portion of each instrument used to rate the activities of daily living or activity limitations.
In the order of their development, the functional independence measure was developed to bring uniform terminology to medical rehabilitation, and demonstrate that rehabilitation benefits could be measured. The minimum data set was developed to provide a comprehensive, standardized assessment with common nomenclature, as well as to identify the resources that would be needed to provide quality care in long-term care facilities.
The outcome and assessment information set was developed to measure ongoing quality in home health settings, and the MDS PAC(?) for post-acute care is designed to identify the resources needed to provide quality care during short stay, subacute, and rehabilitation admissions. Because these instruments may be used to qualify patients for services, their accurate completion is critical to optimal health care.
Patients' functional status could be rated on all four instruments over a continuum of care, beginning with the MDS PAC during a five day stay on a subacute unit, followed by the FIM after a transfer to the rehabilitation unit. And seven days later, for example, when a patient is unable to meet the three hour rule for rehabilitation and is sent home, the home health team will administer the OASIS. Finally, one week later when the elderly wife is unable to provide the needed care at home, the staff in the long-term care facility will use the MDS to rate functional status. This all occurs within 19 days.
The terms used to name daily living tasks on the four functional assessments have been grouped here under the nine constructs on the screen. Under each construct in the order of their use along the patient's continuum of care, you can see that there are eight different terms that are used to rate moving across surfaces among the four tools.
There are nine different terms used for rating moving from place to place; two terms for consuming food; four terms for cleaning the body, and you will note that bathing is used on all four. That is the only common term among all four of these instruments at this point in time. Four terms for maintaining continence; five terms associated with wearing clothing; two terms for communicating using a phone; four terms for using medications.
And there are six categories, five different ones associated with managing the home. You will note that for the last three, using the phone, the medications, and managing the home, these are not on the MDS, nor are they on the FIM. For the MDS it is probably because that is in a long-term care facility, where individuals won't be managing their home environment. But I'm unsure why they aren't on the FIM, because rehabilitation is to help people return to the community.
In addition to the differing terminology, the scales that are used to rate levels of function also differ. With no activity limitation being rated as a zero on three of the instruments, and with a seven on the FIM. The OASIS is unique in that each item has criteria for rating that item, but the range of the ratings vary. And in some instances the subtasks included in the criteria also vary within each item, making progress very difficult to document.
The implications for patients and staff relate to the potential for inaccuracy of functional status due to differing terms for the same construct, which I have just shown you. Also due to differing exclusions and inclusions among the same items. For example, toilet use on the MDS includes the transfer to the toilet. But if I'm writing what I think is the same term on the MDS PAC, toilet use and toilet transfer are two separate items.
In addition, the numerical scale ranges vary between the tools, and also within the tools, especially the OASIS. The differences in the direction of the scales, for example, for three of the tools no activity limitation is the lowest rating, whereas for the other one it is the highest rating.
Also, differences in scale terminology. For example, if one uses moderate extensive assistance on the MDS PAC, it includes a definition having to do with weight bearing or full performance by a staff. Whereas, moderate assistance on the FIM, there is a notation that a helper is required who provides more than just touching.
Also, functional status ratings on the FIM are reduced if assistive devices are used, as well as on the OASIS. However, on the OASIS it depends on each individual item, and it varies by item. Neither the MDS nor the MDS PAC penalize functional status when an assistive device is used.
The variability among the activity/activity limitation sections of the functional assessment tools have implications for both the patient whose performance is being rated, as well as for the staff who are required to rate the functional status. Patients or residents, as I mentioned before, could be rated on all four instruments during their continuum of care. And when status is re-rated with the next admission, differing nomenclatures and scaling can cause confusion.
Patients can also be rated on more than one instrument simultaneously by the same staff. For example, I conduct research in a hospital that has a rehabilitation unit, and a hospital-based skilled nursing facility, and a home health service. Inpatients on the hospital-based skilled nursing facility unit are assessed simultaneously, with both the FIM, which the whole unit uses, as well as the MDS, because it's a hospital-based SNIF(?).
And some patients are then followed home by the same therapy staff, who are then required to use the OASIS. And finally, the same patient may have greater or fewer activity limitations based on the specific items in the instrument alone.
Conceptual confusion can also impact whether or not a patient or a resident is admitted. For example, some MDS nurses from skilled nursing facilities now go into the hospital and screen inpatients, and then put together sort of a shortened RUGS(?) so that they will be able to determine what the resource needs of the patient are. And if they are too great, they deny admission.
It can also impact the functional status and progress between assessments, thus either triggering an early discharge, or an extended stay. It can also impact the level of reimbursement, because functional status is a part of the RUGS grouper for the MDS data. It can also impact discharge disposition for assisted living centers for example, that have specific levels of criteria for independence between they are allowed to be admitted.
It can also impact the effectiveness rating of a facility. If you ever have heard some of the comments from physiatrists who use the FIM, they send the data into Buffalo so that they can compare how their effectiveness ratings are in their facility compared to other facilities that also use the FIM.
And likewise, it can impact the quality of care in a facility such as a long-term care facility, and both the MDS and the MDS PAC have quality indicators. Also, it can impact compliance with federal and state regulation.
As your committee continues its discussion of functional assessment and the ICIDH, the conceptual confusion within the ICIDH should be also be considered. For example, dressing is included under activity, but not under participation and personal maintenance. And how does one differentiate moving around under activity limitation from mobility under participation based on their overlapping definitions?
Or does seeing/hearing functions under impairments differ significantly from seeing/hearing under activities? Or how does one coincide classification of sensory functions such as seeing and hearing under the same rubric as economic skills, work acquisition skills, and activities related to excretion?
The four instruments currently used to assess functional status at the activity level use observation of performance as the method of assessment, with various well defined time frames for the performance observation. What methods of assessment are appropriate for each ICIDH level? And especially for function at the societal level, which can be culture bound?
And finally, I hope that the conceptual confusion that I have raised will help you as you consider how conceptual confusion within the ICIDH-2 might interact with differential methods of assessment appropriate for each ICIDH level of function over a person's continuum of care.
Thank you.
DR. IEZZONI: Dr. Holm, that was very helpful. That was a very, very clear presentation. Carolyn, I want to say it was two years ago in March that you arranged that meeting for us. Carolyn worked very hard, almost lost her job, and she arranged a meeting for us at HCFA where we heard from the different kinks within the organization that were proselytizing along these different types of instruments.
And we heard what you just described to us, but much less elegantly, and much less succinctly put. And it seemed that people recognized that there was a problem, threw their hands in the air and said, there is nothing we can do about it. So as an academic I often also find myself doing that exact thing, identifying problems, and then when people ask me for solutions, I say, well.
Do you have any suggestions? Knowing what is happening, do you have any suggestions, number one, for the issue along the SNIF and the OASIS and the MDS PAC continuum we're seeing in HCFA? And then maybe as a second order question, whether the ICIDH would just add so much to the confusion that you would not advocate thinking about it further?
DR. HOLM: I'm not anti-ICIDH. In fact, the World Health Organization model I think is very helpful for conceptualizing the consequences of function, and that it can go both ways. I have difficulty, however, in the new ICIDH and previously with some of the categories. And again, it's some of as I mentioned, some of the conceptual confusion. I really cannot sort it out neatly and cleanly. But I think that obviously as we heard earlier, that there has been a lot of work on this.
What I have difficulty with in the sort of federally funded instruments is that along the same line that the terminology keeps changing. And it would be nice if they shifted it once. And when you have like the qualifiers that they are talking about in the ICIDH, you have a set of qualifiers that are clear and well defined, then if there is nothing wrong with those, can they be carried over to the next?
I have difficulty implementing some of the systems such as the OASIS, which are criterion-referenced. And I use criterion-referenced tools, however, it is difficult when each one of those levels has a different term in it. For example, with telephoning one level talks about answering it and dialing it. The next one talks about looking up like the numbers, but doesn't talk about answering it. And the next one talks about conversation.
But what if my patient improves in one, and it is not listed at the next level? It shows they have made no progress. If they have made no progress, they may need to be discharged. So I have difficulty with that.
I guess with federal funding I would like to know that there is some communication from one to the next to the next. And that as improvements are made, those are brought forward in the previous tools. Or that there is one that covers the whole continuum of care.
DR. IEZZONI: I guess the question our subcommittee was thinking about pursuing, but I don't know -- Kathy, maybe you will -- whether it was just shear exhaustion or being overwhelmed that made us kind of think, okay, let's focus. But this whole notion of coming up with a consistent data model over the continuum of care was one that is very challenging. Do you feel that somebody has done that or somebody is doing that?
DR. HOLM: I think that the MDS PAC was originally thought of as at least starting to go there. I don't think it ended up there, and I don't know why. At least in some of the early conversations -- I don't know that, but I know that when we went to a TEP(?) that they were saying to them at the time, and I think that there were other constraints.
MS. RIMES: I think Lisa is talking more generically too across the continuum of care, beyond long-term care.
DR. NEWACHECK: I think we were also talking about the continuum of the age span too, to incorporate children as well. And none of these federally funded tools are really designed for children.
MS. RIMES: They cannot be. Because we tend not to think in terms of anything beyond --
DR. NEWACHECK: Right, whereas the ICIDH does attempt to look at the whole age spectrum.
DR. IEZZONI: Dr. Meyers, did you have a comment?
DR. MEYERS: I know you didn't ask me, but I'll just make one observation. Apropos of FIM and kids, there is WEFIM(?). There is a version of FIM for kids, which is controversial actually. Dr. Simeonsson can talk about that.
In the world of spinal cord injury and other major disabling conditions where I work, there is an additional problem. And I agree, that was a quite fine and comprehensive review. I have just completed one that wasn't half so good. But there is another concern, and that is floor and ceiling effects. And there is an instrument called QIF, the quadriplegic index of function, simply because FIM wasn't sensitive to the kinds of clinically and operationally important changes that characterize rehabilitation for people who are quadriplegic.
FIM would indicate that they were making no progress at all, nor is there in the current state-of-the-art, any reason to expect that their FIM scores are going to change substantially.
MR. HANDLER: I have some additional problems to add to the collection. Under the category managing the home, most of the items listed, except for transportation, are things that usually are done by women, not by men. So a man asked these items might say no, because he usually doesn't do them anyway.
DR. HOLM: Not in this generation.
MR. HANDLER: If they are by themselves, I guess they are forced to.
DR. HOLM: It is a cohort effect. I do research in that area, and that is very true that it is women with this generation, very much.
MR. HANDLER: And then my father-in-law with Parkinson's disease has trouble getting dressed. So what he has learned to do is if he has to button a shirt, he keep it buttoned hanging up, and then he puts it on like a t-shirt. He has trouble tying his laces, so he wears sneakers with velcro or slippers. So he doesn't have trouble, but he watches what he wears.
DR. IEZZONI: Yes, I think Dr. Holm as an OT, would understand that completely.
Are there other questions or comments from the committee? That was really super. Are you able to stick around? Okay, why don't we see if we can hear from all the panelists, with some questions right afterwards. And then maybe we'll have global questions at the end for everybody.
Dr. Simeonsson, could you introduce yourself.
DR. SIMEONSSON: Thank you for inviting me to join this group today, and to share some ideas about children. I want to begin by indicating that we are talking about a line of research that we have been conducting with the support from the CDC for the last two years.
I also want to clarify that when the letter was sent to me, I wrestled with what I should be presenting, because in that letter it said here are six or seven questions, and alternatively the words "functional assessment, assessment of function, functional status, health status" all came up, and I wasn't sure if it all meant the same thing, because in different spheres of work, some of those words do mean different things. As a matter of fact, functional assessment and they meant functional analysis of behavior, which obviously is not what we are talking about.
But in any case, what I share with you today is kind of an integration of some of the work I have been doing, and please consider this preliminary. So I prepared some materials for you that you can keep, but don't have the final word, so to speak.
Also, we will be operating from the standpoint today of looking at proxy reports for children; in this case, the caregivers. We do have a good deal of our research which has involved children whenever possible, describing their own functional status. But today I'll be restricting the work to reports by family and caregivers.
The purpose will be to address the questions of interest that I think NCVHS has asked me to look at with reference to functional assessment of children, and then present a rationale for a research program and some representative findings that we have been dealing with. I hope it's the kind of things that you are interested in. Otherwise, we'll try another file.
Then I'll take a shot at identifying some issues. I don't know if there indications, but I always feel that if I have a chance to advocate for children, that's what I want to do. So we'll certainly say some things about children.
As I understood the issues, as I felt compelled to hit those seven points, so let me just say that this will be the last time you will see this part. The type of information being collected -- our work is involved with research in surveillance. The data collected on specific subpopulations or groups of children and youth with or without disabilities.
Decision basis for data collection, and I understood that to mean why did you look at these things rather than something else. In this case our interest has been to differentiate the various dimensions of human functioning and disablement with particular reference to children.
What is the relationship of these kinds of measurements and functional assessment to quality of care and outcomes? And in our work we look at adaptation and quality of life of children and families. I will not be talking about that today, but that's the framework in which we are operating.
What are some logistical problems that we have experienced? And I think those relate to future plans that we are entertaining to look at. Certainly issues of severity are a significant problem; not about rating it, but the issue of severity in the child or the person with a disability. The proxy aspect always comes up. The role of development, and the role of development in environments.
And what are some possible recommendations? Certainly, the importance of a common language and measurement across several systems for children. I think children, again, represent a very unique group. Allan and I were talking about it over lunch, and he was reminding of the fact that we had a classification of children, we would probably need about four of them, because there are four different groups.
Certainly, when we look at the systems that serve them, everybody has their own way of defining and measuring. This creates an enormous burden in terms of effort for those involved trying to provide administrative data. So I'll come back to that later.
The research program that we have been involved in attempts to examine the changing paradigm of human function and disability through a modeling of the person-environment interaction, and a dimensional measurement and classification. I'll indicate why we chose to look at it that way.
The program of research focuses on functional assessment of children and youth with and without disabilities. And in this context we have conducted a series of complementary studies which vary in methodologies and samples, but draw on common assessment framework of health status, functional characteristics, secondary conditions, activities, participation, environment, and quality of life. In most cases we have tried to get some aspect of that, but some studies look at some of those, and not others.
Finding a basis for functional assessment from our standpoint is first of all to differentiate the person, the environment, and the interaction elements of the person-environment interaction model. Everyone agrees to this broad framework whether it's called a biopsychosocial model or the person-environment interaction, or the social model. And then to identify the measurement and classification within a dimensional framework. That's where we draw on the ICIDH.
I know you are very familiar with this, the traditional paradigm of disability, in which the problem was inherent in the person, and limitations were presumed to be a direct reflection of the person's limitations. There are a variety of different models. This is the one we have chosen. I like to have visual things. Don Lollar and I make a good team, because he doesn't have graphics, and I can't talk without graphics.
But I wanted to illustrate this in the sense that I think there is a difference between a model of a taxonomy, which is an attempt to say how are these different dimensions related, as opposed to a model of human function. This is certainly not novel with me, but I wanted to emphasize the fact that if we look at a person's interaction in environment, there is always two components to this interaction. And the person can act upon it and react to an environment. And the environment can be a source of stimulation, as well as feedback.
Obviously, if a person has physical or mental impairments, the ability to act upon and react to that environment may be limited, thereby resulting in a limited interaction. And again, if the person has a sensory limitation, or by some other means has limited access, the environment may not provide an adequate source of stimulation.
So from our standpoint, it is that aspect of the interaction which really is what we are measuring and looking at in functional assessment. That is, what does a person do? I think in the ICIDH the word "performance" is in there, but it often gets buried in the activities. It's actually the performance of activities which is a central part.
And the debate has gone on about is there a difference between participation and activity? The Dutch would say simple and complex. We might think that it's really a range of activities which are embedded. So a very complex activity is made up of a variety of smaller activities, and more minute activities. We won't debate that this moment, but this is a framework that we are looking at.
So trying to then tease apart these elements within a dimensional view, the ICIDH, which is still the official document, by the way, so we will try to show you how we might code it to the ICIDH and then ICD -- we had trouble with that one -- ICIDH-2, Beta-2. But certainly what we think in terms of children, we think about the fact how does the body or mind function?
How does a child perform daily life activities? In some cases, those activities are very different from those of adults, in other cases they are very, very similar, they just have a different form. So whether it's dressing or eating or toileting, they are on their way towards greater independence.
How is a child involved in roles and situations? In that context, roles is not exactly a favorite word right now in the ICIDH-2, but I think it is an important word, because it really captures the idea that children have roles. We all have roles. We have the role of a student. You have the role of playmate, family member, and so forth and so on. And the extent to which that involvement occurs is certainly an important measure.
The environment, as Don spoke about this morning, is very crucial both measuring in its own right, which there has been a great deal of interest. But also in terms of how it mediates participation in activities. Going back to our interactive model, of course over any period of time a person changes the environment, and the environment changes the person. And that's something that also we can measure in terms of outcomes.
This is the old model. I just saw an article the other day in which the Dutch began by reviewing the current ICIDH, and saying that the original model was a dramatic breakthrough. It really was, because it really dimensionalized human function. Up until that point, it was all collapsed into one. And I think we still owe a great deal of debt to the original formulation.
The current model, as you know, looks like this, and I won't belabor that, except to say that our focus today is largely on what is in that middle part of the activities, the activity limitation, although to some extent there is some slippage over to body and structural function depending on the particular approach to be used.
Let me just illustrate briefly then some of the things we have been doing, and only in terms of both illustrating how activities may be assessed, and the strategies we use to do that. These are just lifted out of a larger research study, so they really don't make that much sense by themselves. But I wanted to illustrate what kind of areas we looked at, how they might be coded to the ICIDH, and where some of those remaining issues might be.
We recently completed a family survey of health status and functional limitations of children without disabilities in four North Carolina counties. Actually, they couldn't have disabilities we did a random probability sample. And we are interested in two counties for reasons of the fact that they have better health services and schools.
Secondly, the other three deal with children with disabilities that are identified. Two of them are population-based, and one is -- the reasons for eligibility, a reason for encounter, but eligibility is the educational framework in which this research was funded, the national early intervention longitudinal study.
The third is a coding of functional characteristics in a national survey of students with disabilities in the school environment. And finally, a coding of secondary conditions in a longitudinal study of children.
The first method, so you just briefly have a context for this, is a population-based telephone survey of families in four counties in North Carolina and was a probability sample which is limited to those who are known to have school-aged children in their family. Survey items encompassed health status, functional characteristics, behavioral style. We felt very fortunate, it was almost like a 40 minute phone interview, and you have a ton of data.
We have about 110 from about each of four counties. And our interest here for the purposes of today's presentation is to code the health and functional status of these. Some of these items have a one-to-one correspondence with NIH's D(?) items. Others were created to capture the particular interest of our study.
But in this case families were asked what are the child's major difficulties? You can see here that they range from 9.3 for expressive communication to following instructions for 34 percent. In most cases we have here listed the ICIDH-2 codes. I think that as you can imagine the reason why we have following instructions and problem solving very high, and writing and that is that this encompasses school-age populations.
What are the child's behavior difficulties? Reporting being depressed, making friends, getting along, coping with stress. Those can be coded to the ICIDH-2. And let me clarify that in most cases you will see a 730 or 720, or in some cases just a 700 or something like that. I want to emphasize here that these are estimates of the closest candidate for that. In some cases there is a very exact number, in which case we list that, like 8301. But in other cases that's the approximate area.
As a matter of fact, it was interesting in the field trials we will have an opportunity to say how comfortable we feel with that fit. I think this is important that at this point some of the categories, others are really quite a stretch. They somehow touch on the issue, but they're not really a very good fit.
One area that we have found very important in our work over many years for lack of another word is behavioral style. Behavioral style is kind of a concept that some people like to call temperament. That immediately puts it into one frame of reference. Behavioral style really just simply talks about the fact that a child might be temporized by the style which varies along with different dimensions of behavior. The original contributions came from Stan Chaz(?), but there have been many different variations since then.
In this case, some of these dimensions that characterize behavioral style could in fact be coded to the ICIDH; others are not. But the basic premise here is that a child who presents with functional limitations, and the nature of that functional limitation, to the extent to which that may or may not become a problem, therefore being of concern to a caregiver, being of concern to a health care professional, or being of concern to a school is really mediated by the extent to which that child's behavior style might be mediating that particular behavior.
So in this case for example, a child who has difficulty communicating, but if they have a very high activity level, and are very irregular in their pattern or very intense, that just magnifies the level. So this I think -- these are bureau of SS next items. Do you have a limitation? Do you have trouble learning? Do you consider yourself to be disabled? In this case, 9.5 percent do. Interestingly enough, when we asked how many children were -- obviously at the end of the survey we have this involved in special education.
There is not a one-to-one correspondence actually with the number of children in special education. It is slightly less than 9.5 percent, which is probably a little less than we would find in the population in general. However, parents may not always know exactly what services a child is receiving for special education.
Child health status, overall health, days lost. Do they require medication? Twenty percent. And then what was interesting to us and I think we are all aware of the fact that fitness issues are of significant concern. In this case about a fourth of the families are saying the child has fitness.
The physical education -- and we haven't crosswalked this yet, as a matter of fact, we haven't crossed any of this with child care statistics pertaining to functional limitations. But as we have found out in our other research in clinical work is that children with disabilities are often the first to be denied, or not have access to physical education when in fact they are the ones in need of it.
The national early intervention longitudinal study follows certain families from the time they enter early intervention to at least kindergarten. This is a national program funded by the US Office of Education. It follows about 200,000 children throughout the United States. We are working with Stanford Research Institute as participants in this study, developing and analyzing child behaviors.
Our interest here was to code the reasons for eligibility. As in many other programs in the Office of Education, a child has to demonstrate eligibility in order to be served in that program. And this true, and I'll show you later on, for special education services, as well as for early intervention.
Children might become eligible for a reason which is defined by the school providing the intervention systems. But there is enormous variability within those labels. So two children with different labels and two diagnoses may be much more similar that two children with the same diagnosis. There is a lot of noise in that system, so we were interested in looking at that.
Just to give you an idea of children coming into an early intervention. They can come in on the basis of developmental delay, which is the default option, which means somebody has to make a determination the child is developmentally delayed, which is an important issue that is currently not really addressed in the ICIDH.
Secondly, a proxy for that can be that the child has a diagnosed condition. So if you demonstrate a developmental delay, that is you are not walking -- you should be walking or making vocalizations and so forth and so on -- then if you have a diagnosed condition such as Downs syndrome or anything else, you are eligible. Or you might be determined to be at risk for developmental delay, because you had spent three weeks in an incu(?) or because of socio-economic factors, maternal age, any current environmental factors. Or for that matter, you might be at risk for delay if you don't receive early intervention.
The national stratified sample we did across the US and so forth. I can give that detail if you want later on. But based on this, more than half, about 60 percent, came through the gate of developmental delay, about a third under the rubric of diagnosed conditions, and 13 percent at risk.
The stimulus is that on the referral form that was made by the early intervention agency, they indicated one of three criteria for entering the system. Then they described why the child came in through that gate. They said this child was made eligible on the basis of developmental delay. But on the line to say why was the child developmentally delayed, they might say, well, the child had Downs syndrome.
So for example in this case they would have a descriptor for one condition, which was actually fitting the criteria for another. So this kind of variability is throughout. So we took the descriptors that were provided, not the criteria themselves, and tried to code those to the ICIDH, and in this case because ICIDH was not sufficient to do the job, you will see on the left-hand they coded to the ICD-9, and in some areas we also coded to DSM.
But where possible we looked at the crosswalk to the original ICIDH to B1 and B2. And as you can see by global delay, which is an important phenomenon to capture, a large group of the children at this point in time is not readily codable to the ICIDH. Prenatal condition, congenital factors and so forth obviously are not going to be specifically appropriate for the ICIDH.
As you will note there are not some areas in which we cannot currently code children's recurring problems to the ICIDH. It's not a problem of the ICIDH. It's just simply saying when kids are coming into a system, in this case a government supported system of services, we should have a way to code them that would capture their characteristics. In this case, they are coming in because of a functional problem, which is developmentally delayed. That should be codable.
What we then took is we set up a coding schema which had about 22 categories, which is a combination of ICD-9, ICIDH, and the DSM. And then based on the criteria that we applied using the three groupings of developmental delay, we recategorized them so there would be a correspondence. If there was a functional limitation, their child actually was assigned to an inferred eligibility. As you can see here, the numbers at risk rise somewhat, and the developmental delays rise somewhat, and the number of diagnosed conditions drops.
The national survey of school environments, this is another study we have conducted. It was also then a probability sample of teachers of students with disabilities. They were asked to identify a student in their environment who had a disability. And we asked them a range of questions about their environment, about their functional status, and so forth.
For this purpose today, I just wanted to go to the bottom, which is to explore the classification alternatives to categorization. Special education, separate from mental health, separate from health, has its own categorical way of identifying children. And it is a combination or a variety of things. The current version of the Individuals with Disabilities Education Act defines eligibility in terms of children with disabilities means children with, and then there is about 12 categories.
If you pick one of those, and then who by reason of their reason of their own need special education related services and you are in. Between three and five may at a state's discretion include children who have experienced developmental delays in one or more areas. Actually, children three to five, this is one is functional assessment, because we are talking about areas of function. We said communication, mobility, and so forth. Whereas, children with disabilities for COP(?) really is a mixture of etiology, impairments at the body function and structure level.
I thought you might find it interesting for some of you who are not familiar with this area, that the original designation of special education made available as a right to all children began in 1975 with the Education for Handicapped Children's Act. The designation here, mental retardation, hard of hearing, blind, visually impaired, and so forth.
As you can see, with 1991 the culture changed. The word "disability" came in. But some of the words still remain the same, even though it was IDEA. In 1997, some of the categories were retained, others were dropped. Autism, as you can see, came in 1991 and continued on. Other groups have varied in terms of eligibility. In this case, for example attention deficit disorder, which is probably one of the most rapid epidemics we have, hyperactivity disorder in many cases comes in with other health impaired. There is not a separate category.
So what we were interested in doing is looking at instead of these categories, which are very arbitrary, and which actually are not only arbitrary in terms of their designation, but are in fact operationalized differently from one state to the other. So the prevalence or the head count if you will of children in special education services varies dramatically from state to state by category as a function not only of the obvious demographic make-up of the state, but also how the state chooses to operationalize the choice.
So we wanted to see if we could code those. And as you can see here, we look at the old ICIDH. Most of them are coded to the eye, and as it is in ICIDH-2, these are really body function structure, even though we are ostensively looking at functional limitations in children.
If we get to other areas like learning, communication, and behavior, then you are moving into the area of activity and functional limitation. There should be an item there for motor. I'm sorry, actually everything should be shifted up. This dropped down, and so actually motor, learning, communication. We don't have anything for multiple, which seems not codable at this point.
The last study I just call your attention to as a way of thinking about coding for functional status of children is a prospective study of four age groups of children, youth with spinal bifida. We enrolled about 60 in each group, and we followed them over several years. Both parents and children were respondents on measures of health and functional status, secondary, conditions, and quality of life. And we have come to code those secondary conditions in particular to the ICIDH.
This is where we have gotten some really very useful information from the children. As long as they are old enough to respond, 10 or 11, we have gotten some very useful data from them in terms of their own perspective, and also to look at the issue of proxy.
Secondary conditions is a term that maybe is varied, so we want to share with you how we operationalized it as physical, social, and psychological complications, sequela, or consequences secondary to a primary condition or impairment. In this case, spinal bifida. The primary condition is a necessary, but not sufficient cause. And of course this is inferred in the sense that not every child who has the same lesion level, has the same complications. And of course some complications wax and wane with development, and that's what we have been looking at.
Just to give you an idea here of what that may look like. These are illustrative. Pains can break down, latex allergy, weight, and balance are some of the more frequent. And as you can see also here, that they vary in terms of developmental age with weight and balance, which are particularly key issues for the middle ear. We have coded these. I didn't have them on there. But these are codable to the ICIDH-2 to some degree on the body function.
In terms of functional limitations, activity limitations, these are coded. Self-confidence, there is not a place for that in the ICIDH-2, but endurance, learning, attention, problem-solving are codable to selected chapters in the activities area. Again, as you can see, there is variable that is a function of different age groups, and they are looking at this longitudinally as well. This is cross-sectional.
On the issue of participation, what used to be called handicap, obviously family issues and the isolation exclusion issues are certainly important ones, defining 30-40 percent of young people. And these again are codable to the current participation dimension of the ICIDH.
So we then pretty much conclude here. I think that just as an overview there are frequent functional limitations in these. They certainly appear with some frequency in the areas of communication and learning behavior. We obviously also have indications of significant health status issues. And the issue is, is that measured, depending on its manifestation, either at the body function level, or at the functional activity level.
Similar patterns, but higher frequencies are found in children with disabilities. We have been exploring of course also the data suggesting to us that there are probably a reduced set of items which encompass some of these dimensions with the inclusion of a behavioral style, because this is really again, what I think mediates a lot of these characteristics, and raises the level of both intensity and visibility to the functional and developmental problems of children.
What are some of the issues that merit attention? The issue of proxy. For other age groups it's one kind of issue. For children, particularly young children with severe disabilities it's one that we really have to look at very carefully. And obviously whenever possible, we need to find ways of asking children themselves. There are those, including ourselves, who are trying to develop simple measures that children can respond to.
But then the issue is are they responding to the same thing that we are asking the parents, and can you have equivalent forms? The alternative is of course to observe behavior, but some of those behaviors may be in low frequency and be difficult to capture. And the more you go down that road, the more you almost end up having to set up structured tasks which then become the very equivalent to developmental assessment, but this is a key issue.
Children who are nonverbal obviously create other challenges as well. The measure of assistive technology's efficacy and its place with children, I think Don made that case this morning. I think it's very important with children, because often times the initial fittings if you will with prosthetic devices and other more simple kinds of assistive technology becomes available to children, I think there is no question that the future holds even more dramatic opportunities for assistive technology, and how to look at that relative to functional assessment.
We have already raised and continue to raise the issue of delay. Is delay different from deficit? We certainly would argue that it is. Somehow we have to capture that, and I think it is an issue of whether or not you make it a qualifier, or if you build it into the items. But certainly just as we have problems in maturation, problems in growth, there are children who are characterized by variability in those dimensions.
It's not, if you will, under the frame the disability, although it is a limitation. But how to capture that, and certainly for many young children it is a very important concept. What are the characteristics of the environments which mediate functional capability?
I have some recommendations to make. Certainly I think there is a sort of an emerging consensus of a conceptual framework. I don't mean the ICIDH-2 per se, but I do mean the fact that I think there is consensus that there is a dimensional approach, and those dimensions are broadly defined by something at the body level, something at the performance level, something at the engagement level, and then whatever the context might be.
And I think to do that, to adhere to that in some way can help really improve the quality of our measures. And that is already happening. Increasingly, there are a lot of measurement tools which either are being developed, and they are framed within some sort of dimensional framework, or as I would say, they are being retrofitted. The ICIDH is having an effect in that area.
These can help you define the basis and the dimensions for eligibility. And in this case I think obviously different dimensions will be used for different purposes. I will reiterate what Don said this morning. I think the context which I'm very familiar with, I think in this case all children in all societies have to go through is the school portal.
That one really is defined by a functional limitation, that is, meeting the demands of the school environment, the learning environment, the social environment, the cultural environment. I think having a common framework would be very useful.
So I will leave you with that. Thank you very much.
DR. IEZZONI: Thank you very much for a thorough presentation. I'm turning to Paul. Paul, you are the person with the most expertise here about children.
DR. NEWACHECK: I do have a question, Rune. You have been working in this area for a long time, and also with ICIDH for a long time. Can you give us your overall assessment of the value of ICIDH to us as it now stands, and perhaps the next iteration that is being envisioned here in terms of its utility for measuring functional status of children, which is really what the committee is -- we are really focused more on the functional status of it, although it might be broader later.
And also, do you have any sort of sense about the feasibility of using an ICIDH to tie functional status indicator in administrative records? That is, encounter forms, that sort of thing.
DR. SIMEONSSON: I know you know this. You realize that I'm not a person who is not neutral in this matter.
DR. IEZZONI: The rest of us don't know you, so be explicit.
DR. SIMEONSSON: I will be. By that I meant that I want to be both the strongest supporter and the harshest critic of the ICIDH, because I'm very committed to the issue of having a usable tool for children. I think that a lot more needs to be done, but I would go back the final premise that certainly my colleagues in different areas of services, health and education and so forth, the underlying concept of dimensional approach is inherently attractive, and makes a lot of sense. I think that some of the tools that are in wide use and growing use have been the PD, things of that nature. I think they are reflecting that aspect.
For administrative purposes I will reiterate my point I was trying to make at the end, which was that we waste a lot of time in my opinion, and an enormous amount of effort in the transition between the services, because we don't have a common language. And even if we could have just I think the dimensions of the activities, the codes of the activities dimension coded at the first level, just at least so we know the ball park, I think that would be a start. And I think that could be done quite readily.
The work that we are doing now with this national early intervention longitudinal study, I mean I think we feel like this is a very productive way, and once you have an established system that you can then provide. They are going to code something. Why not code something that has a common framework, and that can be linked at some point to a taxonomy? I think that would be the goal.
We certainly would hope that some of the limitations of the ICIDH with particular reference to children will be corrected in the next couple of years. And our goal is to have a separate version for children.
DR. NEWACHECK: But using the same conventions.
DR. SIMEONSSON: Using the same broad framework. So some of you have seen the DSM PC for pediatric context, which builds on the relevant codes of the DSM, so there is a consistency with the main volume, but you only use those that relate to children, and then you expand on them with particular reference to children's behaviors, children's context.
For example, one of the earliest items in there, and I still have trouble making case. I'm sorry, if you don't mind, but things like consciousness. Those of you who have worked with young children, things like level of arousal, being able to state organization. These are crucial issues for young children, and they are very important for children with disabilities.
But in the adult thinking these have to do with stupor and coma, and that's not what we are talking about. And the mediating role of state organization is a very important marker. So it's things like that, that probably will not find their way into the main volume, simply because they are perceived to be too unique, but play a very significant role in a child version.
DR. NEWACHECK: So where does that child version stand now? Is it under development?
DR. SIMEONSSON: No, the process is that the main volume will be completed, and then the children's version will be -- the work of the children's task force is already underway, but the children's version would be developed over the next probably 2-3 years.
DR. NEWACHECK: Are there any other versions being developed for other than children?
DR. SIMEONSSON: Not that I know of. Because by definition it's supposed to be a life span, and inclusive of human functioning. So in that sense that's where we go.
DR. IEZZONI: I'm not sure I understand this last exchange, because DSM I always think of as the Diagnostic and Statistic Manual in psychiatry. Is that what you are talking about?
DR. SIMEONSSON: Yes.
DR. IEZZONI: Okay, so that's for psychiatry. I guess I'm wondering -- or mental health, behavioral health disorders -- is there any analogous process going on for the ICIDH development? I got a little lost in there.
DR. SIMEONSSON: The DSM has a version that was designed for primary care for pediatricians. It was geared towards children and youth.
DR. IEZZONI: But it's still behavioral health?
DR. SIMEONSSON: Yes, it's still a mental health and psychiatric taxonomy, but it actually is quite attractive in many ways, because it allows you to look at children's mental health problems prior to the level of a formal diagnosis. So it allows you to say this is a concern.
So pediatricians, for example -- the whole point to this was that pediatricians would see kids in their practice who they would be very reluctant to make a referral to the psychiatrist for a diagnosis, but at the same time they want to call attention to the fact that this child is manifesting some problems. And I was only using that as an example of the fact that in the same way that there is a correspondence between the DSM PC for children and youth in the area of mental health, we would hope to have an analogous ICIDH for children out of the main volume.
DR. IEZZONI: Okay, so my question is, is there a process underway for that to happen?
DR. SIMEONSSON: Yes, there is a process underway for that, and there is a task force.
DR. IEZZONI: Because you mentioned the issue of role. And you said that the word "role" has some kind of overt lay to it that since I'm not a pediatric researcher, I don't really understand. But I would think that there would be cross-cultural and cross-country -- at least certain age groups, different perceptions of relative roles in childhood. And so I guess how is that going to be worked out as this task force is looking at this?
DR. SIMEONSSON: Part of this is probably the activities that go on as something evolves. But in the original ICIDH handicap was defined by survival roles. And there was a lot of reaction to survival roles. So the roles got thrown out with the survival. So now it's the issue of involvement in life situations, which is a little awkward. But people sort of get the general idea, which is participation is about being involved and engaged in situations.
Roles is a perfectly fine word, but roles really has to do with that participation area, whereas the activity dimension is the performance part. And keeping those relatively separated is part of the challenge.
MR. HANDLER: I have a politically charged question, and you can plead the fifth amendment if you want to. For about three months ad nauseam we have heard about Elian Gonzalez. Both sides say they want what's best for the child. Is there an objective, functional assessment that could be given to evaluate the current condition of that child? And then could that be used to predict if he goes to Cuba, what would happen, if he stay in Miami, what would happen?
DR. IEZZONI: You don't need to answer that.
DR. SIMEONSSON: I won't answer that, but let me answer it in a round about way. I think that we certainly frame our work in the larger context of the UN declaration of rights for the child, which is the fact that the child has the rights to access. In this case rights to access means right to access to a family, to a family life, to school, and so forth and so on.
In that sense there is a broader philosophical frame. And I think the work of the ICIDH-2 is also framed in the standardization of opportunity. And Allan, you might have some comments on that as well. So I think that, no, I'm not sure that somebody could open a book and say, here it is, and that's why we can document that. They have a team of psychologists and psychiatrists.
But at the same time the broader premise, what are the dimensions of human function and disability, and can we talk about what is better or poorer functioning, then I think the dimensions are captured in the ICIDH, should be the basis for defining what is a reasonable life for any child.
DR. IEZZONI: That was actually beautifully done. I very impressed with how you brought the ICIDH back into that.
MR. HANDLER: I just think it's a shame that with all of this ability, an exam like that wasn't done, to my knowledge. That's just my point of view.
DR. IEZZONI: Okay, thank you. Yes, you are through, but if you could stick around there might be a few questions for you at the end.
So Dr. Meyers -- some of us might remember that Dr. Meyers actually did appear at Boston Logan Airport that January morning when we had 12 inches of snow here in DC. And he was told that the airport in Washington was open. The airport in Washington was totally closed. So now we have been graced by a reappearance.
DR. MEYERS: That's very kind of you to say. Thank you. We were canceled once by snow, large amounts of snow. As you may or may not know, in Boston on Tuesday we had snow. There was none forecast for today, and indeed there was none today. But we have a place in Vermont, and indeed, there was snow in Vermont this morning.
DR. IEZZONI: There wasn't much. I didn't any snow. I live in Newton.
DR. MEYERS: I said Vermont. None in Boston, but there was in Vermont. That was another distraction, yet another reason for not attending these discussions, but indeed I am here.
Like a good professor I am text rich and graphic poor. I have just a couple of overheads. I'm Allan Meyers. I'm a professor and health services researcher at Boston University School of Public Health and Boston Medical Center, obviously in Boston. For about 20 years I have had professional and personal interests in impairment and disability, and for the past 7 or 8 years these have been my preeminent professional concerns.
Currently, I'm the research director of a hospital-based center for the treatment of people with spinal cord injury. I'm also a consultant to the Massachusetts Department Health's Office on Health and Disability, conducting a statewide longitudinal study of second conditions among adults, comparable to some of the work that Rune is doing.
And I continue to work very closely with Boston's Community Medical Group, a model program that provides comprehensive prepaid managed care on a risk-based, capitated basis to people with severe physical impairments, something that many advocates and indeed many critics say can't be done.
From all of these perspectives, I'm very glad that you are holding these sessions to discuss the inclusion of information about functional status in standardized data sets. I'm flattered that I was invited to join you. Candidly, I'm not entirely sure why I was invited to join you; flattered, nonetheless.
And I'm very pleased by the subcommittee's title, which refers to populations, and which is sufficiently ambiguous -- at the risk of over-interpretation -- to address two fundamental concerns that I think underlie and motivate my remarks today. The first of those, people who are affected by many types of impairments or disabilities have many of the same characteristics of other special populations, for example, racial and ethnic minorities, people of color, women, and therefore require special methodologic attention and sensitivity as we try to measure or characterize their health status or health care needs.
Secondly, this is sort of the reciprocal, at the same time that impairment and disability are parts of the experiences of general, that is normal populations, as they are part of the life experiences of normal individuals, and therefore, we must take care not to sequester people with impairments and disabilities, and to make our research methods, our research instruments sufficiently cosmopolitan, sufficiently flexible that they are available and accessible to as many such people as we possibly can.
I am impressed by your agenda of questions, but I find that many of them address clinical or administrative settings. As I'm neither a clinician nor an administrator, I'm afraid I can respond directly in an informed way to such questions. What I would like to do instead is to use my research experience and my experience working closely with clinicians and administrators in clinical and administrative settings to respond to the spirit of most of your questions and the letter of a few by addressing three related themes.
And the second and last of my overheads. The three themes, and you will be either frustrated or very pleased to know that I'll make only tangential reference to ICIDH, although I'm very interested in ICIDH. But methodologic and philosophic risks that are associated with heavy reliance on proxy assessments, which is may sound like a counterpoint. I think it is not a counterpoint to Dr. Simeonsson's presentation. I'll come back to that in a moment.
Secondly, the need to enable our research instruments, in the spirit of enabling America, so that they are operationally, conceptually, and philosophically useful to as wide a range of people as possible. And my contention is that currently our research instruments are disabled, or handicapped to use the old language of ICIDH.
And thirdly, the need to expand our functional assessments in two ways to address salutatarian and felicitarian dimensions -- I like to invent words. I will define the words that I invent. Actually, I invented only one of those. And to put function in social and environmental contexts. That will be the interface with ICIDH.
One of my students told me last week that defining your own words or creating your own words is diagnostic of schizophrenia. Well, you form your own conclusions. Only one of the words is one that I created.
In a sense in my remarks I know that I'm preaching to the choir, a choir that includes Gerry Hendershot, Paul Placek, Don Lollar, who was here before, Rune Simeonsson, who already have done a great deal to address these themes. Nonetheless, I know that sometimes even the choir appreciates a good sermon, so let me proceed.
Let me talk first about proxies. I understand that there is a considerable administrative appeal to using proxy responses. I think though on the other hand that we have considerable or enough evidence of both the nature and the magnitude of the bias that is associated with proxy responses, particularly responses for people with impairments that I think we should avoid that as much as we possibly can.
Just an aside for a moment, to return to one of the comments that Dr. Simeonsson made, a conversation that we had last weekend, and that we continued over lunch today. A lot of the reason that we use proxy responses in research with kids are reasons that have not to do with the capacity of kids to speak on their on behalf, to articulate on their own behalf, but rather have to do with research requirements -- institutional review board requirements, which are very much legitimate requirements.
Working in a clinical setting, many if not all of those requirements are removed. And it seems to me the artificial in the sense of legal and ethical restraints, restrictions on the use of direct responses that we have in research settings, would not be present in clinical settings, and that's even a stronger argument I think for letting children speak on their own behalf.
More broadly, I think that we have not yet tested the ability for people with disabilities to speak or to gesture, because many are not able to speak on their own behalf. We seem to have operated from the assumption that people cannot, and therefore follow the more expedient direction of using proxy responses.
My preference, and the preference that we have begun to test with some research colleagues is to begin with the assumption, the null hypothesis that everyone can. And only insofar as we find that it's an excessive personal, ethical burden, then we rely on the proxy. Insofar as we rely on proxies, I think we lose data. We lose the capacity of people -- we compromise the capacity of people to speak for themselves.
Let me move to my second point, enablement. Once we accept, or indeed if you accept the need for direct rather than proxy measures of functional status, health measures, it follows that we have to make significant mechanical, methodologic, and conceptual changes in order to make the accessible to people with disabilities.
In the spirit of the ADA, and only somewhat ironically, and in deference to a seminal publication in our field, I have characterized this process as the enablement of research instruments. This is a subject on which I have expanded in other settings. I have provided some detail in my written comments.
Let me just highlight a few areas where in contrast to enabling, and again using the language of the older ICIDH, we have handicapped our research methods. Sampling methods that systematically exclude people with disabilities, and these are probably less relevant in administrative settings, much more relevant in research settings, modes of administration that don't allow people to complete research instruments. And I'm thinking now in the context of people who are tetraplegic or tremulous.
I, too have a father-in-law with advanced Parkinson's disease, so all of your remarks were hitting very close to home. Any effort to use self-scoring questionnaires is an effort that makes it difficult, if not impossible for such people to participate in the research process. Again, it causes a reliance on proxies. That in turn introduces a bias, the nature and direction of hard, probably impossible to measure.
And then finally, research instruments whose content and concepts are so offensive or so alien to the experiences of people with disabilities, that even if they have the functional capacity to participate in the research process, they don't have the inclination to do so. And again, there is a text that I elaborate in considerable detail. I have a particular concern with SF-36, which at least in the adult population is emerging as the industry standard.
A short form SF-36 makes in reference to a domain of mobility, makes reference to walking in a horizontal plane, climbing stairs in a vertical plane. I think again, there is no villainy involved. There was just a confusion, a kind of intellectual shorthand that confused the domains of ambulation with the domains of mobility.
That was a very interesting presentation. I want to come back if we have time to ICIDH-2, because I'm afraid in the walking and related activities -- and until you presented this, Gerry, I really hadn't been aware of this -- a similar kind of confusion prevails.
At any rate, this is a very important measure. It's the industry standard measure. The language of walking distances or climbing stairs is language that in our experience was unavailable in the best case, and in the worst case, blatantly offensive to people with spinal cord injuries, not limited to those who are tetraplegic; those who are paraplegic as well. It was an offensive term.
There were also concerns about the order of administration. So if you were not able to walk, and not able to climb, you in fact have this rubbed in your face three times by the, I think, unnecessary order of administration. And what we have done in recognizing that there is a methodologic compromise to doing so, when we use SF-36 in our studies of independently living adults with mobility impairments we say going instead of walking, and we say going up instead of climbing.
And we change the order of administration so that if you tell us you are unable to meet the smaller challenge, we don't ask you about the subsequent larger challenges as well. To me, to us that's a process that we call enabling. To some of my colleagues, that's methodologic compromise, but it's a compromise that I think in the interests of intellectual truth, because we say we are talking about the health of populations, and in the interests of compatibility and continued good relations with the populations that we study, it's a compromise that we have been prepared to make.
And then finally, salutatude and felicitude. Felicitude is a neologism. That's my invention, and it was an invention, a term I invented in writing a paper for a friend and colleague, Dr. Alan Crocker, who is certainly well known in the field of childhood disability, and also somebody who is very interested in positive, as well as healthy outcomes.
What I mean by salutatude and felicitude, those of you who are as old as I know that salutatude comes from Aaron Etanovski(?), and came to me actually from his junior colleague and my senior colleague, the last Saul Levin. Once we establish the utility of direct measures, once we make measures that acceptable and available to people, it is important that we measure affirmative, as well as negative experiences.
What possible relevance does this have to functional status? Well, traditionally, and there were a couple of very fine reviews this morning, both before I came and after of our traditional measures of activity of daily living. All of the measures have been based on the implicit assumption that function was an intrinsic property. It was a property of the organism. Don Lollar uses the term the naked person. I guess that's a Southern metaphor. In the North we say people taken out of any sort of a context.
It is very important to put the measure of function in that context. And when we put it in context, and I think this is more to the points that I wanted to make, function includes such domains as work, leisure, and sexual expression, which take place in highly complicated and elusive settings, and which are subject to considerable environment constraints.
I'm not suggesting that the measurement of ADL and IADL, the traditional functional measures of feeding, bathing, grooming, and elimination. Obviously, these are important. I don't want to belabor that point. But we can't have our functional assessments, and with those measures. That probably ought to be where they begin.
In conclusion, let me repeat, because I said it earlier, that along side the benefits of enabled instruments, and we feel these benefits are considerable, instruments that measure salutatude, felicitude, there are going to be costs -- logistic costs, methodologic, and fiscal costs. We appreciate that.
And without knowing the costs, because we haven't begun to quantify them yet, I am reluctant to say that the benefits will outweigh the costs in any kind of economic sense. However, I am quite comfortable saying that the benefits of justice, equity, and fuller participation have considerable and perhaps incalculable values in their own right. And that the benefits of having a more truly accurate picture of our population has benefits that are similarly great.
Again, I would like to thank you for the invitation. Sorry about the snow, and I wish you ever success in your task.
DR. IEZZONI: Thank you. I have had discussions with John Ware on this exact topic. Because I say, John, I never know how to answer these questions. And I wonder, Dr. Meyers, whether you have published your work anywhere, because it's not just people who have spinal cord injury who might react that way. Whether you have published it, and whether there are some efforts through the Medical Outcomes Institute or whatever to disseminate your enabled versions of the SF-36?
DR. MEYERS: A couple of questions, and then I'd like to send one back to you, if you don't mind. We are publishing this paper, or a version of the enabling the instruments paper in a special collection of the Archives of Physical Medicine and Rehabilitation. It will be coming out in August I believe. It's a collection edited by the senior editor, Elaina Andresen(?) at St. Louis University. Don Lollar and I are junior editors, and there will be an expanded chapter in there on enabling of instruments. So yes, it's coming out.
The Medical Outcomes Trust -- this is politically charged. They are aware of our concerns. I have not discussed this directly with Ware. I have discussed this with his colleagues, beginning actually when Saul was alive. And the reactions were mixed. For some people it's how could we do that? Others' reactions were why would you want to measure the health of somebody who was in a wheelchair?
What I learned only last weekend that really infuriates me, there is a new version of SF-36 that has changed the language, including the language of some of the mobility questions, but left intact the words of walking and climbing. I'm astounded.
Again, this is not villainy. It's oversight. It's shortsighted, myopia, with all deference to people who are visually impaired. It's a kind of characterologic, intellectual myopia.
If I could put the question, how do you answer those questions when they are posed?
DR. IEZZONI: Well, the irony of this is that I had this conversation with John Ware in a taxi as we were both going to be on the technical advisory committee to Westat, the Michele Adler study that was discussed this morning. And John kind of implied he had never thought about it before. But I can't quite believe that. I cannot answer those questions.
DR. MEYERS: What we have found is -- and our survey, we describe it briefly in the testimony. It has been described elsewhere, and there are references. We are dealing with young adults for the most part, who are connected with independent living centers in Massachusetts, many of whom are highly politicized.
And what we found was we started off by saying, you know when the questionnaires talks about statements about he and mankind, we say what we really mean is he or she. It's the generic, the white, Christian gentleman. Well, we mean people who uphold standards of civilization.
So we said walk means whatever the usual interviewer instruction. And for some respondents that was acceptable. And then they would say, well, do you mean with or without my wheelchair? For some respondents the response was furious. I won't repeat it, because we are on the Internet. And after about a 100 such either perplexed or furious responses, what the interviewer said to me is we're not going to do this anymore. If you want to ask them offensive questions, you ask them.
So least I appear to be a hero, my sense was one of kind of methodologic purism, until it became very clear that either we had to stop using SF-36, don't use the mobility questions, which again is absurd, because that would suggest that people who use wheelchairs aren't mobile. And on Monday we will see wheelchair marathoners who cross the finish line in the Boston Marathon an hour or so ahead of the ambulatory competitors. Preaching to the choir again, so I won't go through that.
Or we could give preemptive responses, and just assume that everybody in a wheelchair is giving the most restrictive, and attribute the appropriate score. And none of those was a tenable or acceptable alternative.
DR. IEZZONI: Let me just point out that after this presentation we are going to hear from folks from the National Center for Health Statistics about how they include functional status in their national surveys. And there is a question, do you have any difficulty walking a quarter of a mile, that is, two or three city blocks? And in our analyses of the 1994, 1995 NHISD phase 1, we found that 0.5 percent of wheelchair users said that they had no difficulty with walking. They don't walk.
DR. MEYERS: And you know from earlier NHIS -- I would be very interested in seeing that analysis, actually.
DR. IEZZONI: So I actually experienced this with a qualitative interview that I was doing in somebody's home where I asked him about his trouble walking. And he looked at me and he said, I don't have any trouble, I don't even try. So it's kind of this performance versus --
MR. HANDLER: Maybe the people in a wheelchair change the word "walking" to "moving."
DR. MEYERS: We don't know what they do. Some people in conversations, some people who use wheelchairs will talk about walking. They mean wheeling. But I think part of the problem is that ambulation and mobility are overlapping domains to be sure, but they are distinct domains, and some of us have forgotten that.
DR. IEZZONI: And one of the problems with the current NHIS I think, at least as they ask functional status information, is that they ask a global question about uses of assistive technology. And so you don't really know whether somebody is using a wheelchair or walker or what other kind of device. So you can't kind of sort this out at all. I guess it's probably a length of the questionnaire issue for them.
Okay, I've been dominating. Does anybody have any questions for Dr. Meyers or the panel? I really enjoyed your presentation. I will say something at the risk of sounding harsh. And that is that I'm not sure how many people will read the Archives of Physical Medicine and Rehabilitation. I think that it would be important to get this kind of finding out there in the broader literature, so people who design surveys and think about questioning people who have serious functional impairments, and may have disabilities, understand these issues, and can learn from how you have dealt with them.
DR. MEYERS: That's not harsh at all. And actually, we've been speaking with the editors of the Archives, and Don Lollar will have funds available for much wider dissemination, either mailings of the original document, or reprinting in a separate format.
MS. RIMES: It was also in the proceedings of for the National Vital and Health Statistics Committee.
DR. MEYERS: That's right. Actually, that was the first public presentation of these ideas.
MS. RIMES: That also, I think, play fairly well.
DR. MEYERS: Thank you.
DR. IEZZONI: When was that?
DR. MEYERS: The August 1999 conference. I was one of the speakers at that conference.
DR. IEZZONI: We are going to take a brief break before the next speakers, and I'd like to thank our panel, which gave us a lot to think about. Thank you very much.
[Brief recess.]
DR. IEZZONI: So this afternoon we are going to be hearing about the National Center for Health Statistics surveys, and how they are rethinking about how they look at functional status. Can I have the two speakers each introduce yourselves, and which one of you is going to go first?
MS. MADANS: I'm going to go first and third.
DR. IEZZONI: All right. I've been told by Jennifer that she has long versions and short versions of what she is going to speak about. She will pause at various points and ask us if we would like more in-depth information, and if we do, she is happy to do it. But if not, we'll just move on.
MS. MADANS: I'm Jennifer Madans. I'm the Associate Director for Science at NCHS, and this is Elsie.
MS. PAMUK: Elsie Pamuk. I'm the Acting Director of the Division of Epidemiology at NCHS.
MS. MADANS: And the third person who couldn't be here, but did leave me some information is not an NCHS person, but someone who is working with us in a new endeavor. I'll mention that in a minute, Nancy Mackowitz(?), who is at the Joint Program for Survey Methodology is doing some survey development work with us in this area. I just wanted to mention some of the things that she was doing.
As you can see, neither Elsie nor I work for the National Health Interview Survey Division. And so it's somewhat odd for us to be doing this since -- not everyone at NCHS speaks with one voice on anything having to do with our survey work. So in light of full disclosure, I will give you my impression of what we are doing on the HIS, and on some of the other surveys, but I'll definitely focus on the HIS.
And I thought about this a lot over maybe the last five years, because I did chair the committee that did the redesign that went into effect in 1997. And so I do take responsibility for taking off your favorite questions, Paul. I did it. But I do think that we have a better survey than we had before, and it is better to position us for the future.
And by that I also want to say that we now, or I feel pretty strongly that all our surveys are kind of works in progress. That we haven't hit the ultimate survey mechanism yet. And so any suggestions you have for changes, we certainly are taking them under advisement, and will not wait another 30 or 40 years before we do design again.
I'm sorry I missed this morning's presentations on ICIDH, but having thought about the way HIS approaches measures of health writ broad, and all aspect of it, in a lot of ways, the traditional way that HIS has approached functioning is I think in very much keeping with that approach of a continuum of interrelated sectors that in some senses exist on their own, but also you have to have all the arrows. Another way that is not so -- and even the redesign is not so compliant with that approach to measuring functioning, disability, whatever you want to call it.
The other full disclosure here I misused the word. I have still not figured out the difference between activity and function, and so I'll mess them up, and I apologize for that.
HIS very much has always looked at health in terms of its impact. It's much more, I think, a sociologic definition of health than a biologic or physiologic definition of health. I think that the founding fathers did not believe that an interview situation was appropriate for finding out people's true health status, if there is such a thing, or in having them report about conditions.
Instead, they felt that an interview situation was much more appropriate for finding out whatever you actually do have, how does it impact your life? And measuring that impact through the interview. And so you get lots of measures like restricted activity days or disability days or bed days or work days, which we are looking at the impact of health on the society. Those variables were not measured at an individual level. They weren't meant to identify health status of people. It was more groups.
How many days of work are lost due to ill health? And the fact that at a given level of health, different people will stay home, and others won't, didn't really affect health. They were looking at that kind of global impact.
And then the more person level measures, our famous limitations of activity measures were again trying to get at that impact of whatever objective health status the person was experiencing. And I think that is very much in keeping with the idea of participation, and getting away from a biologic model to a more sociologic model.
But it has a lot of limitations, because it does incorporate into the development of those questions, I don't want to say reporting biases, but they turn out to be reporting biases. I'll talk about that a little bit more in how some of people's sociologic characteristics affect how they will report even the same behavior. And that became very clear in some of the analyses that have been done over time, and I think that's an area where we have a lot of work to do.
The redesign a lot of ways was a great improvement. I think the one area where we didn't improve hardly at all was in the area of limitation of activity, role functioning, some of the functioning questions. I will say that in my opinion, and this is all my opinion, we are probably as good as anyone else. We're at state-of-the-art. I don't think there is another set out there that are better than the set of questions we do. But ours is still crummy.
When we tried to do the redesign, we really tried to improve on those questions. We had a variety of workshop and consultancies, and just a lot of working back at the office trying to figure out ways of getting around some of the methodologic problems with these questions, and were unsuccessful. At the end of the day we couldn't figure out anything better to put in there, so we left a lot of what we had for a variety of reasons. One was we had to measure the year 2000 objectives in a consistent way, so by default some of those items stayed in.
But really, there was no overwhelming reason to change that. But at that time there was definitely a commitment to continue that research to improve those questions, to expand on them. And it's very nice that we now have a mechanism through the Center for Excellence on Health Statistics, which Nancy Mackowitz is working on to do some more methodologic work.
So I think at this point we need to do the research to improve these questions. We're not where we want to be. We acknowledge that, but there is no quick fix. We have not been able to come up with something that we can do quickly to make these much, much better. So with that introduction, I'll just kind of generally -- we had to think about we measure health status, and actually just as from a quality of life presentation.
All of this is basically on the 1997. I'm not going to go back to the old way. But starting with symptoms, which wasn't in the HIS for a long time. There were very few symptomology questions, to measuring conditions in a very different way, to functioning measures, and then to our global measure of self-rated help. I think that's one way of thinking of the continuum as it is used in HIS.
When you talk about functioning, we get at the functioning in a variety of different ways which I actually think is appropriate. So there is not section that is only functioning, and that's the only time you get functioning. So for example, when we do the conditions now, we ask about various aspects of conditions -- the diagnosis, onset, some symptoms, medications. And for some conditions we will do a follow-on and ask some functioning questions as an outcome of that condition.
And at this point we are only doing it for injuries and asthma. I honestly do not remember why those two. I think it was a mistake. There were meant to be more or less.
These questions are not meant to get estimates of functioning. They are meant to get estimates of the severity of the sequela of the condition, and it's one way of getting at the functioning aspects, but not the only way.
Similarly with hearing and vision questions, which have been on HIS for a very long time, these performance batteries are kind of taking some of the diagnoses and use of equipment, and then these are questions -- can you read? Can you recognize sounds in a noisy room and understand the conversation? So one way of getting at the functioning measures is through the root of conditions, but it's not the only way, and is not meant to be used independently in my mind. It's very much a part of this little component.
If you want a list of the functioning measures that we need now, this would be my list. Again, we go back to the disability days approach. We still ask about work and school lost days. Again, here more so to look at a societal impact. We know that on an individual level the thing that affects whether or not you miss work is whether or not you have sick leave. And that's really not all that interesting to us from a health point of view. But from an economic point of view in terms of lost productivity, that's why we ask these questions. They are not individual measures. They get some kind of societal impact.
The 12 month bed days actually had been used as a real measure, like we really want to count the days. We know that that doesn't give you a great estimate of days. This is really used on a personal level, as a dichotomy. If you've got a lot of days, you're probably sicker than if you have few days. It's a severity indicator.
I will say that in the redesign we spent a lot of time looking for severity indicators. If you look at the literature, people tend to count conditions, and people who have more conditions are sicker than people who have fewer conditions, and that's not necessarily so. But there are very few measures of severity and impact. So in the redesign we were trying to find ways of characterizing health status through conditions without counting them.
Self-assessed health status I mentioned previously. We do do the ADLs and the IADLs. I have never been a big fan of those, because I don't quite know how to interpret them. And I think they, along with a lot of the other measures that we have, and something I'll talk about in a minute, the context for those, that interaction between person and environment is very difficult to tease out in the ADLs and IADLs as we ask them. So they are very frustrating to analyze, and I think that's a big problem.
The naja(?) functioning items, which you were discussing before about the walking, climbing, bending, stooping, HIS asks those specifically without using any equipment. And I think this was for us, something that has been going on for a long time, not so much on the HIS, but in the survey that Kaiser Circle, the HIS. There is the HIS core that is done every year, and I did assume everyone knew about the HIS and didn't do the normal what is the HIS stuff.
But the stuff that is done every year is what we normally think of as the HIS. The HIS has lots of supplements, and of course there was that big disability supplement. And things change in supplements. The questions come in and out, and they vary a little bit.
It also has follow-up surveys. For example, the LSOA deals with things very differently than you will have on the core. So it's kind of hard to talk about the HIS and how it deals with functioning. But this is pretty much on the core HIS.
The importance, and I think it's extremely important and is at the heart of why we have so many definitions, so many estimates of disability, and such confusion in sometimes analyzing our data is that we don't clearly differentiate environment and what Don calls the naked person. I don't know what else to call that.
But it's important to do that, especially in the case of some of the other NCHS surveys that are longitudinal, where you want to control on environmental effects. And also for some of these questions we also use the NHANES survey where we do exams. And you want to be able to take the self-report and match it back to whatever exams you have been able to do.
So one approach, and an important direction for us to take in the methodologic work on the HIS is to try to do a better job of separating that out so you can look at changes in both, and in the interaction. Especially because our data is used a lot for policy purposes, if you are looking at ways to intervene, you want to make sure that you are not losing an intervention point by not making services available at a given point in that continuum, because you don't know you need to. You haven't asked the question that way.
These questions are asked very explicitly without using anything. What we have done on the HANES-1 follow-up and the LSOA, and I think is the appropriate way to go, is to then ask now using what you normally use, what is your level of functioning. And that was done in HANES-1 follow-up, and Lois Ruby(?) I think has done some very interesting analyses of that. I don't remember if the LSOA does that. It probably does that.
The issue there is time. It's probably not something we can do every year, but it should be something that we do periodically. It does increase the amount of time on the survey, and while HIS is big, it's not that big.
DR. IEZZONI: Jennifer, can I ask a question about that? I know that time is a premium, but if you have skip patterns built in, it is going to be a minority of people who are going to report serious problems. So couldn't you have a skip pattern such that if somebody reports a serious problem, you then ask them about it?
MS. MADANS: Yes, and taking your comment before about how we ask the question about the use of equipment, but we don't find out what kind. There may be a way we can work around that. That for certain kinds of equipment, where we have enough numbers. That's the other thing, we want to be able to analyze what we get. I think that is something we should look at. Is there a way where we can imbed in at least some parts of it, major issues where we can ask that follow-on question, because I think it is important.
MS. PAMUK: Can I add something about that? To bring it down to a much more mundane and pragmatic level, when we were working on the redesign, we had greater hopes for the capiazation(?) which was going on at the same time. And for a lot of reasons, it did not pan out as well as we had thought. A number of those questions were specifically constructed with the idea that at some point we would be able to incorporate look-up functions and make them much more efficient, and be able to do exactly the kind of that you are talking about. What has really bogged us down has been technical issues.
MS. MADANS: We did decide to do the questionnaire content redesign at the same time we switch to CAPY(?), which made a lot of sense, because you could do things, but CAPY turned out to be even more complicated than those of us who thought it was going to be complicated to do. And Elsie makes a good point. We are switching the authoring system in the next couple of years to a more up-to-date one. And hopefully, that will make it a lot easier to change some of these skip patterns. But that's a good point.
I think we have gone further in these kind of questions than in the limitation of activity questions in terms of making them more appropriate. These questions are in the sample adult. So it's only asked of one adult in the household. There is a family core where questions are asked of everyone. We have less flexibility there, because it's proxy respondent; just one responds for the whole household.
These are we get self-reports, and I think in the adult sample, child. And there are some more questions on the child, is where we can get more bang for the buck in terms of expanding this battery.
Role functioning is something I really wanted to spend a lot more time on in the redesign, because the roles we were dealing with and the limitation activity questions, as you'll hear, are kind of limited. Kids go to school. Adults work, and if you are over 65 you do ADLs. That was it. And that is certainly not comprehensive. We were totally unsuccessful in doing that, trying to get good questions that were workable and wouldn't take forever, and that's an area that we are still working on.
Cognitive functioning was another one that we had big hopes for, and I think we're down to one memory question. Again, it's very difficult to do in the kind of survey that we do. It's a very large, population-based, very production oriented, always in the field. No time to think, because you are in the field for the next one.
This is just to give you a real quick example of for the NAGI type of questions we get use of equipment, but as you say, it's not specific; by yourself, without anything, how much difficulty do you have. So we are trying to get some kind of wider range than just do you are or don't you.
We do try to get some attribution, not by each activity, but in general. This is extremely hard to do. It takes even if you give categories, and they are gross categories, you get a lot of write-ins. Then you find out that they told you something that is not really a condition. And then you have to go back and recode everything. So it's a mixed blessing. And there is some onset type information, which for some things I think is very important.
Then I just want to say something briefly about the problems that I have with the limitations questions, and then Elsie is going to talk about how we changed them, and what effect it had. And some of the issues I think are because of the implied standards. These are all questions that say because of health problems, defined mental, social, physical, are you limited in.
I think the problem is the activity that we say. Are you limited in play? Are you limited in work? Well, it kind of depends what kind of work you do. So it is not a standard question. When you try to compare people, and we are interested in looking at disparities by a variety of sociodemographic characteristics, what they do is related to those characteristics.
So you get these very strange differentials, because there is no true standard that you are asking people to compare against. It's just work. And it's just not specific enough.
We also don't have a standard for what is limited. And we know that how people determine whether or not they are limited is due to a lot of different things, some personal characteristics of theirs -- race, age, gender, religion. All kinds of things will determine how you value that, or how you put a value on that.
We also know that societal things have a big effect. Some of these measures vary very much with unemployment rates and other kinds of societal movements that you wouldn't think would have an effect, but they provide a context, well, I thought I was limited, because I couldn't get a job. Unemployment is really low. Now it's easier for me to get a job. I'm not limited anymore. You see this variability, which is telling you something, but it's not telling you what we thought the questions were telling you, which is something about a person.
So we have a real problem with the lack of standardization and the lack of a comparison. It's an implied comparison. Everyone is bringing their own comparison to it. And it is affected by other characteristics that we happen to be interested in.
The other thing with these questions is they are totally confounded with environment. I assume and I don't know that people are answering this as I work now. The work I want to do in my current situation as the world is now; not could you. And without that explicit environment context, it's very hard to look at change over time in these kinds of areas.
It's very hard to understand where you could intervene to improve functioning so that while I think we are now very cognizant of this change and approach to measuring disability, and through the ICIDH with participation very much being on the interaction, these particular questions do not deal with that directly. It's all implied. It's all in there. And it makes them very hard to interpret, so they become very frustrating.
As we move forward in these in terms of the methodology, I know that I have some things that Nancy is working on, doing more to separate those. But of course as soon as you start separating, you start adding questions and time and all of those things. So we need to find a parsimonious way of doing that without having 40 minutes of an hour survey in this one area.
We also get health insurance and immunization and utilization and everything else. But there has to be a way. We just haven't found it yet.
Do you want me to stop, or should I do the HANES stuff? I think the only deal on the HANES stuff is we get --
DR. IEZZONI: Why don't you quickly just go high points of HANES?
MS. MADANS: The HANES questions, especially in HANES forever, it started in 1999, and it will be in the field forever, very closely follows the HIS questionnaire. We are trying to keep them very close, even though people keep trying to make them different. But we are trying to at least get the self-report things pretty comparable.
In HANES-3 there are a lot more functioning exam questions. That's what HANES gives you that you'll never get from HIS, are some objective measures. These are the exam components that were related back to the functioning data -- timed walk, chair, stand, mobility measures. Those were done in the MEC, mobile exam center, but also some of them were done in the home. In HANES 99+ I think we're actually down to two. And then there are some related components. Questionnaire components are very similar.
There has been some work done on comparing the self-report to what you get in the objective measures. And as you might imagine, they don't always match up. For all those reasons that we know, that were talked about why those questions are not as precise and explicit in the self-report as you would like them to be. So we know there is a big disconnect.
However, having both on one survey tells you a lot, and I think can lead you towards some improvement in the self-report, but also kind of keep you honest. Sometimes I think we believe our own data a little too much. That it's really telling what we set out to find out. This kind of stuff kind of makes it clear that it's really not doing that.
They do a little bit more on the cognitive, because they can do it in the MEC. It's done on the MEC. It's still ADL; the physical activity is very similar, and then the general health questions. So there is a nice crosswalk between the HANES and the HIS. HIS has a lot more on utilization and that side. HANES has more on the exam side. So you can kind of do some interesting methodologies by kind of crosswalking the two.
So with that, I'll stop.
DR. IEZZONI: Thank you. Are there any quick questions for Jennifer on points of clarification, areas that you were confused about? No. Yes?
MR. HANDLER: Anytime you have a suggestion, it costs money. But is there any consideration to oversample so that the people you interview, you know in advance have a disability? And then you'll go into in-depth questions, because your population is the disabled population, not the general population.
MS. MADANS: Well, oversampling implies that you know where those people are. And that's almost not true of any group. Now we can do screening. And we do do screening for other things. We are doing a phone survey where we are screening for children with special health care needs. It is possible. We screen now, because we oversample by race and ethnicity. We would need to have a screener that worked, and then would just not ask any more questions of folks who screened out.
It costs about a third of an interview, a third to a half of what an interview costs just to do the screening, because you have to go to the household. So it is expensive. And you have to have a good screener. And it's very hard to get good screeners. But it is doable.
DR. IEZZONI: Well, the Social Security Administration told us about a $40 million project this morning.
MS. MADANS: We know about that.
DR. IEZZONI: They are doing a screener. Let's wait and see how theirs works.
MR. HANDLER: I know my agency -- I'm not obligating anyone at my agency to spend any money -- but at Indian Health Service we serve 1.3 million people. And when they come to us for care, the first thing they do is get a check up. So we have diagnostic information, and it's automated on a patient record. So you could use at least the Indian population. Maybe there could be joint activities between Indian Health and NCHS.
MS. MADANS: Just on the screening issue, we did the disability supplement. The screener was 40 minutes, as you recall.
DR. IEZZONI: Yes, it was very complicated.
MS. MADANS: I don't know if it had to be that long.
DR. IEZZONI: Okay, Elsie.
MS. PAMUK: Well, thank you.
Gerry left. There is a reason that Gerry left. As he can tell you, I have two versions of this presentation, long and very long. And I will very much try and keep it to the sort of thing that will inform this discussion most. It's a little hard for me. I'm a real data drone, and this is really data. So you can feel comfortable just giving me a speed it along sign if I start getting bogged down in details.
But Jennifer thought that some of the work that we have started doing on looking at the way the HIS activity limitation questions have been answered would help inform this process, and take it out of the realm of the purely conceptual, and what you would ideally like to have, and the information you would like to get to addressing some of the very pragmatic issues of what it's possible to get.
And this only addresses that in a very oblique way, but what I would like to say at the beginning is this is a start for us. One of the things that we feel very strongly about is that the redesign of the HIS, by separating it out into this section which is the family core very much like in purpose, the old family core, but then adding on the new sample adult questionnaire, and the new sample child questionnaire will allow a kind of crosswalking that will let us examine what some of the problems are in the way that we are asking these questions.
So again, I don't know how many of you are very familiar with this, but when the redesign of the HIS was being conceptualized, we were very much torn between two diverse purposes. On one hand there were people who didn't want us to change anything, because we are the only source of trend data for the United States. That has a great deal of weight and responsibility to it, and it's something that we really felt we needed to address. We couldn't just ignore it.
On the other hand was everybody who was saying you're going about this the wrong way. You've been going about it the wrong way, and you need to correct that. So we came up with what we thought at the time was a very ingenious solution, which was that we would do our best in the family core to maintain as much of the trend as possible, and at the same time whenever we could, we would get rid of some of the more obvious flaws. Then we would radically change the way that we were doing things, switching for example from condition to symptomology in the sample adult and the sample child questionnaires.
So what we have set up now is kind of an unanticipated experiment in that what we really tried to do in the family core was maintain the sense of the questions the way they had been asked in the past, and at the same time try and do a bit of refining to address some real obvious kinds of needs.
What I'm going to show you here is sort of what happened with that. So this only focuses on the activity limitations of the family core. And what I'm going to compare is the old HIS. I'm going to show you data from 1995 and 1996, just so that you can see how closely 1995 and 1996 track. And then show you 1997.
The point I want to make is this. In the old HIS, rightly or wrongly, every activity limitation question was there basically to slot individuals into one of these four categories. And this whole set of categories was conceptualized to actually convey a continuum of disability from very severe to no limitation at all.
And if you look at the first three categories combined, you come up with a grouping that is any activity limitation due to a chronic health condition. And you get there by a series of questions, and then a residual question saying are your activities limited in any way because of a chronic health condition? And we tried to keep that basic paradigm intact in the redesigned HIS, so that in theory we could get again any activity limitation due to a chronic health condition, and it should track over time.
And that's what the next slide shows you. And in fact what you see is that in 1997 we had a bit of a drop in any activity limitation. It's not very much. Essentially, it went from 14.4 percent of the population, which is where it had been virtually throughout the entire previous reign of those questions, from 1982 all the way up to 1996, down to 12.7 percent. That's not much.
And if you consider everything that changed in the course of the redesign, maybe what we should have done is just stop right there and say, oh, we did it and go home. Because what happens is unfortunately that minimal kind of change is produced by counterbalancing some more extreme changes in the different age groups.
And essentially what you see happened is that we actually show increases in activity limitation in preschool kids, a very slight increase in school age kids, a substantial increase in the elderly, persons 70 and over, and then a substantial decline in working age adults.
So what we decided to do then was to decompose the questions that got you into that category of any activity limitation due to a chronic health condition. But basically, I felt I needed to remind people that this notion of being unable to do your major activity, or limited in your major activity, or any other kind of limitation derives from age-specific activities.
So that for under five years old it is play. For school age children it is school. For working age adults it was either work or keeping house, depending on they said their major activity was. And for the elderly it was performing ADLs and IADLs, personal care needs and routine needs.
So in the old HIS what you find is that these questions incorporated a set of imbedded hierarchies. So people were first asked about their ability to do the age appropriate activity at all. Are you able at all to perform personal care activities? And if they said yes, then they were asked if they were limited. And if they said no, they weren't limited, then they were asked this residual question about any other activity limitation.
We did ask about ADLs and IADLs for people under the age of 70, but only if they had reported some activity limitation in the age appropriate activities. That's the old HIS, and that was the paradigm. In the 1997 redesign we actually made an attempt to collect the information like it was collected in the past, but with some changes in wording that we were felt were necessary to bring the questions more in line with other assessment instruments.
Then we thought the place where we had the greatest latitude was in to take away some of these imbedded restrictions so that it was no longer assume obvious that if you were over 70, you didn't want to work, or that was irrelevant to you. So we started asking work limitation of everybody.
And basically, what this does is -- we did not ask people ahead of time what their usual activity was. We didn't presume in many cases. In children we still left some of those presumptions in what their usual activity was. The net effect of all of this is the addition of questions for certain age groups that are there by virtue of the fact that we didn't have the age restrictions on them that we had before.
And some changes in wording. So that for all children under the age of 18, through a series of negotiations with ASPE, a question on receipt of early intervention services or special education was put in. For everybody aged three and over, we ask ADLs; everybody 18 and over we ask IADLs. For 18 and over everybody got asked about work. And everybody got two additional questions that weren't there before, which was a question on difficulty of walking, and this global question Jennifer alluded to that we're not terribly happy with on cognitive function.
I won't go over all of these, but these are the new questions. That's basically the way that -- those are things that were added for essentially the additions of those age groups.
So what we tried to do then was basically disentangle this by looking at the questions that were very similar, and comparing them between 1995, 1996, and 1997, and try and tease out the effect of what is happening in those questions where there were minimal changes in wordings, and then what the effect of all the additional questions was.
So for children less than five years of age, preschool aged children, basically what we did was look at the play activity questions, which were essentially pretty much the same 1997, and pre-1997. There was a change in order, and for all of these the initial phraseology was changed so that instead of saying impairment or health problem, it says due to a physical, mental, or emotional problem.
So just comparing those two play limitation questions, unable basically, no real change. Be sure and look at the percent scale over on the side. These are very small differences. So there is no real change in unable to play. No real change in limited to play. All of the change comes in the other category. It's basically all due to the additional questions that we are asking.
For 15-17, this was a bit more difficult to untangle, because we really changed the essence of these questions. It's the one area where we went from more questions to fewer questions. Prior to 1997, there were questions that asked essentially about special education, but in a series designed to get at whether a kid was going to school at all, were they in special classes. If they weren't in special classes, did they need special classes, and so on. That was replaced by this single question on receipt of special education services.
Overall in 1997, there is a slight increase in the number of kids with any limitation. And basically you see that what is happening there is a shift in other limitation to special education. The change in the way the question was worded got more kids into the category of being limited, almost defined by their receipt of special education than had been the case before with those education questions sort of serialized the way they were.
But overall, when you add those two things together, you really are not seeing much of a change. It's an upward trend that is in concert with what was going on between 1995 and 1996.
And then over to the side I also show you the information for ADLs for this age group, to show that in fact that also shows an increase. So that in my estimation anyway, this is pretty good evidence that basically these questions are conceptually very similar. It's just that the reason for the limitation or the specific way that it is defined is receipt of special education. It's somehow getting more by the change in the question than it had in the past.
Then I'm going to skip to 70 and over. Again, in the past, this was a global question on activities of daily living, and a global question on instrumental activities of daily living. Really, the only thing that changed in the way that we asked that was again, the introductory physical, mental, or emotional, and the specificity of the ADL question. But that's basically it. So for 70 and over what really constitutes the true additional questions and the expanding of the work questions, so that they were asked whether or not they were limited in work.
And in contrast to that huge jump that you saw in any limitation, if you just take out the work question from 70 and over, you see a substantial decrease. You see basically very little change in ADLs and IADLs. And that other category shows a substantial decline. And if we had left the work question in there, it would have shown a substantial increase.
So basically, what that is telling me is that when in the past, we were asking elderly people are your activities limited in any way? Some of them were thinking, yes, I can't work anymore. And now that we are asking it specifically, maybe more of them are saying sure, yes, if you are asking me if I can work, I can't work. And they probably wouldn't have said they had any other limitation if they hadn't been asked specifically about work.
And I really don't know what to make of this, except that it's pretty clear evidence in my mind at least that what's been happening is that in the past in that residual question is we're getting a lot of fairly non-specific information that really depends on the person's notion of how limited they are or not. And when we do ask elderly specifically about their ability to work, we are going to see an increase. And we include that in the definition of limited. We are going to see an increase.
I'm not going to go through this in any great detail. There was a fair amount of change in the working age adults. And specifically that we don't ask about usual activity. We don't ask them to say their usual activity first anymore. We just come right out and say can you work? And if you can, are you limited in the kind or amount of work you can do?
But certainly conceptually, since everybody was asked that work question even prior to 1997, it's just that if people their major activity was keeping house or going to school, they were asked about that as well. But they were always asked about the work question. You would expect us to pick up the same level. At least there is not to be a major shift. And in fact, what we see is that totally unexplained decline.
And it is not concentrated in any one of those questions. We see the decline in people saying they are unable to work. We see the decline in people saying that they are limited in the kind or amount of work they can do. And we see the decline in the other category, although percentage-wise of course the decline in the other category is most substantial. It's about a 43 percent decline.
So I don't know what's going on here. I'm certainly open for suggestions. Jennifer and I have talked about this at length. We know that this is a question that is very sensitive to external circumstances like the overall employment environment. Don Lollar actually suggested to me that by having the phrase "emotion, physical, or mental" we may actually by sort of highlighting the emotional or mental, we may actually be getting people more unwilling to say --
DR. IEZZONI: I was going to actually suggest that. I have seen elderly people -- the SF-36 does the same thing. They ask questions, because of physical and mental problems, do you have a difficulty with X? And I have actually had elderly patients say to me, I don't want to say that I have any mental health problems, so of course not. So I think that it would be good, if you had cognitive interviewing potential, it would be good to see if you could tease that out.
I was actually going to suggest that earlier as an issue, because that is such a stigmatized condition. And especially since these are face-to-face interviews.
MS. PAMUK: Right. Well, the other thing that is going on now -- this is again, something I'm not really clear about. For this same age group, you aren't seeing any significant change in ADLs. But there was a concomitant decline in reported IADLs between 1996 and 1997.
DR. IEZZONI: The issue is that ADL problems are so rare.
MS. PAMUK: Right.
DR. IEZZONI: When you look at your Y axis there, it's so spread out compared to the work limitations issue. So I think it might play out quite differently for those two questions, because they are at the extremes on the continuum.
MS. PAMUK: Right, but I guess I was sort of hoping to find no effect on IADLs either, and then I would be much more inclined to sort of look at. But it's really not clear. And I do want to kind of end it here, because I think one of the things that for me is very exciting, and I'd like to encourage people to do is there is such a wealth of information on the sample adult now to sort of look at the interrelationships between what is reported on the sample adult questionnaire about work, type of work, that sort of thing, their symptomology, their conditions, and what is reported on the family core.
It takes a while to accumulate enough cases. And the reason that we haven't really looked at it to date is this is one year; 1998 data is coming close to being available. So we are starting now to build up enough sample size, and the sample adult and the sample child to make it a fruitful area to sort of cross that information.
We have already started doing a little work on looking at attention deficit hyperactivity disorder, and what is reported. That condition is reported on the sample child, and then what is reported for that child in the family core questionnaire in terms of activity limitations. So it's a nice area to spend some time looking at what's available as the sample size starts to accumulate and accrue.
Thank you.
DR. IEZZONI: That was very helpful. Let me ask the committee members if there are any questions.
DR. NEWACHECK: I wonder if Elsie and Jennifer if you could comment on what the reaction has been in the user community to these changes in the redesign in terms of the functional status.
MS. PAMUK: Can I respond to that, first of all, because when we did the roll out last August right after the public health conference, I think we were all in sort of a little bit of fear and trepidation about what the reaction was going to be to the changes. And at the end of the day when it was opened up to the audience, about the only reaction we got was why can't you add this question, this question, this question, and this question.
So I think to some extent inevitably because it is like trying to build a house and use just a hammer, everybody's perception of the HIS is that it is just not enough. And to some extent, that's still the major reaction regardless. I have seen less reaction, and most of it has been positive to the changes.
People are very pleased with the locus of the HIS being on the individual now. That is one thing they are very pleased with. I think in general people who know the potential of CATY like Lisa, are still, as we are all, very frustrated that we haven't been able to utilize that mechanism as effectively as we would like to. And we really hope that there can be some real significant improvements brought about.
One of the things I didn't mention here is that for this set of questions we do ask causal condition. I spent last year recoding that data, and trying to deal with the verbatim answers we got, which were never designed to be gotten that way. At some point, we really hope to have a look up function that will allow this to be a very rich database, even for fairly rare conditions, because it accumulates over time.
We haven't been able to really incorporate the state-of-the-art technology to the extent we would like to. I think it's going to get much better when we do that.
DR. IEZZONI: I have just been conferring with Marjorie, and we're kind of running a little shy on time. Aaron, it has to be really quick.
MR. HANDLER: Yes, it is. From the tone of your presentation it seems you think there may be something wrong with 1997 data, but what we're not sure what. Well, maybe people are healthier. Maybe they are exercising more. Maybe they are dieting more. Maybe they are not drinking and smoking as much. Maybe they are going to an HMO. Maybe they are taking medications. Maybe there are operations that are being done now because they more money to spend. Maybe there is a positive reason why there is a difference.
MS. PAMUK: I would love to believe that there is a reason that can produce a 4 percent decline in work disability among working age adults in the course of a year. I'm trained as a demographer. Anytime I see a year to year change like that, the first thing that hits my mind is data problems.
If we see the same thing for 1998, which in my mind would be a continuing decline, then I will be ecstatic, unless it turns out that it is all due to the prescription of prozac.
MS. MADANS: I don't know if we would say the 1997 data are bad. They are better. I think the thing on the other, when other is a mush, you don't get great data. When you say it's more or this or that, you get better data. Unfortunately, we were unsuccessful in maintaining the time trend in the one place we wanted to maintain the time trend. So that was unsuccessful, but I actually think the questions are better.
DR. IEZZONI: Great. I really admire what you guys have been doing digging through this, making it easier for those us who use the databases to figure out what to do.
Jennifer, you wanted to present findings from another --
MS. MADANS: I'll do this just really quickly. I think that if the committee is interested in this kind of work, Nancy will be more than happy to come back. And I think after they have done some of their research, you might want to have them do that.
I'm just going to put this on to give credit where credit is due, to the folks who are doing this research. These are people that joined programs for serving methodology which is at the University of Maryland and the University of Michigan, and the CEHS is the Center for Excellence for Health Statistics, which is an extramural program that we at NCHS are funding. We have three of them, one at this program. It's an infrastructure center based program, and one of their projects is on methodological advances in disability and functioning.
She gave me a whole bunch of really nice slides, which I'm not going to show you, but she can come back and show you. If you think we're hard core data people, these are really hard core data people. Just the tension that we in the measurement world are having is that we have an extremely complex concept. The terms are ambiguous at the best. There are lots of questions in terms of context, and if we want to get at this environment interaction.
So it's a very difficult measurement problem. What we hear from some of our colleagues is we would really like you get down -- my ongoing conversation with Donald Lollar, three questions please for all surveys, for all time. These two sides of this are not working well together. And so we gave them this problem to solve.
What they are looking at is the wording, which we know has lots of problems. What context the questions are in. Whether it is self versus proxy. The mode of data collection, because for money we have gone very much to telephone mode when we can. Whether these questions are in a health survey or let's say in something like FIB(?), which is not a health survey, but is focused on something else. And who is the sponsor, does that have any effect?
Starting with the decennial census question, which got all kinds of publicity for reasons I do not understand. And looking at that question, and from a methodologist's point of view the first thing you would say is too complex. And it's not really clear what the attribution is. And the terminology is ambiguous. And you have a discrete answer, a yes or no.
For those of you who weren't lucky enough to get the long form, this is the lead in, and I think there are six different activities that are asked about learning, remembering, and concentrating. The one on this slide, and then there are some others.
What they are going to test are alternatives to the wording to see what difference it makes. So an alternative one, rather than ask that question, they are going to say how much difficulty do you have? They have switched the topic here, so it's dressing, bathing, or getting around. And you have rather than yes/no, you have the difficulty question, so you can get levels.
And then there is a follow-up question to that which asks folks to describe how they are able to do this, and this gets at the special equipment. They are very much interested, and I didn't show that slide, of the interaction between the person and the environment.
A second alternative is another set of statements. Again, not a yes/no, multifaceted. This is the design, and this is what they are going to start doing. They have 800 pairs of completed interviews. They go and randomly assign these questions. Then they'll go back and reinterview. And sometimes they have the subject and they will interview a proxy. Sometimes they reinterview, so they have this nice little matrix, which I won't show you, of how they distribute these 800 pairs so that they can have the experimental designs to look at the things that I showed you to begin with.
They also have another part of this which I won't get into, which are these little vignettes. They think that the sex of the person, the kind of disabilities, all of those things will have an effect. So they have an vignette, and then they change those one at a time. Then they ask someone to read it, and to rank that person, not themselves.
We had thought about using vignettes actually for the redesign of the HIS, and couldn't come up with them. So I'm glad that they are working on that. The last I heard from Nancy was that they were changing the vignettes, so I'm not going to show you those.
Then what they are supposed to be looking at is consistency across interviews, because when they did some early testing on the 1990 Census questions, they got terrible reliability test, retest. So they want to look at that again, the self-proxy. And do you get better consistency, reliability, more comparability with their alternative questions than with the original question. I think they are going to start that research relatively soon. And it's going to be done by telephone, again, for the money process.
We hope to be able to fund them to do some additional work along this area, particularly looking at ways now that we may not be limited to the way we have asked limitations for the next 10 years. But building on the work that Elsie has done, looking at our own sample adults, and how does that work, so we build a research program around that.
The limitation question is really the only role functioning questions we have. As I said, I think we have done better in the sample adult where we get at some of the activities on this side of the continuum, not all the way at the right of the continuum. But we don't have anything else all the way at the right. And clearly, we have some problems with those limitation questions. Can we include that role functioning for all ages? That I think is going to be the focus of the work they are doing.
And also looking towards can we come down with fewer than 45 questions? Maybe not three, but a smaller number that can be used on other surveys that aren't health surveys.
DR. IEZZONI: It sounds like very exciting work. We should definitely hear back from her once things are a little bit further along.
DR. NEWACHECK: That was very helpful. I appreciate it.
DR. IEZZONI: This was very, very helpful even though it wasn't exactly on ICIDH. I think it's certainly material that we needed to hear, and we thank you two for presenting it to us.
DR. IEZZONI: On the agenda it says that supposed to recap the day and discuss. Do I see any groundswell of enthusiasm for doing that?
Yes?
MS. COLTIN: I do have a question, and I'm not sure whether this is the right point to ask to it. But when we are talking about capturing functional status in administrative data, it would be provider reported information. And do you have the capacity to link some of the provider surveys at all? Do the samples overlap in any way? No. So we have provider recorded functional status here and there in a few of the provider surveys.
MS. MADANS: On our provider surveys they are totally anonymous. Now there is a move afoot in the center to take those provider surveys and use them as sampling frames. So we did this once with the nursing home survey. We took a sample of nursing home folks and then created a follow-up survey from that. So we went back and got other information.
The problem with it is there are all kinds of consent issues, which have to be worked out. I don't think that they are impossible to work out, but you have to work them out.
DR. IEZZONI: Now, Jennifer, is MEPS doing any of that, because don't they --
MS. MADANS: MEPS gets the records, but they don't get everyone's record, because it's too expensive. The other way is to forward like MEPS does, and it's just very expensive. So their main focus is to get the money. So they want to get the --
DR. IEZZONI: The utilization stuff.
MS. MADANS: So they do differentially sample records, depending on where they need to get the actual cost data. But I think it's certainly something they could do if it is on the record. The other thing that we can do is link to the Medicare records. Now that doesn't give you anything on the under 65. And we would also like to link to some of the nursing home, the OASIS data if we could.
So if it's on the form, if it's on the billing record, and we do link to Medicare records now, we can match it back. And we do do that all the time on the longitudinal surveys.
MS. COLTIN: But that's precisely the problem that we are trying to address, is that it's not on the administrative record.
MS. MADANS: I understand. But you're going to get it on their first, and then we can do it.
DR. IEZZONI: Jennifer, I didn't realize that you link to Medicare. You link to Medicare claims files? Who gets those data?
MS. MADANS: For the longitudinal surveys we have the ability to link all of them, and we will probably link the whole HANES survey. We don't link all of the HIS surveys because they are just too big, so we link the longitudinal surveys. So the one that --
DR. IEZZONI: The aging survey?
MS. MADANS: Yes, the aging survey. That's public use.
DR. IEZZONI: Actually, this spurred another question that I had, and that is that even though MEPS and NHIS currently use the same sampling frame, the phrasing of your questions is different. And that actually is a problem for investigators like me, who right now have a grant from ARC to look at disability-related issues using NHISD and using MEPS. And we are going to be using NCBS. But I didn't have any pretense that NCBS would be similar.
I have been surprised by the fact that the phrasing of some of the NHIS questions and the MEPS questions around functioning are different, especially around mental health. You probably know that MEPS doesn't even pretend to ask mental health questions. So is there any move afoot to try to get the questions more consistent across those two surveys that are drawn from the same sampling frame?
MS. MADANS: Limited. When they were joined as a sampling frame, a lot of the questionnaire development work had already been done for MEPS. And we weren't going to start from scratch. The problem with doing that, I think that what we and MEPS staff -- and we do talk to each other, contrary to popular belief -- have to hear is that there is a reason to do that.
I know you know this, but it probably bears repeating that any person who is designing a survey, because of time and money constraints, their one focus is to maximize those questions for their survey. And forget about that there are other reasons. And we need to be reminded of that. That if they change their questions a little, and we change our questions a little, it may not maximize either one, but it maximizes joint. And I think we need to hear that from you. Left to our own devices, we won't do it.
DR. IEZZONI: I'll give you an example. The hearing questions and the vision questions are different. They are the same number of questions, it's just the phrasing of them is somewhat different. So we in fact are finding different rates of hearing impairments and vision impairments across the two data sets. Some of the differences are pretty big.
MS. MADANS: Now we should look at the cross of those. I mean now that data has not been public yet, but it will be at some point, either in a public use file or through the data center. I think the first step is to look at what one is picking up versus the other. If you tell us to change our hearing questions, and we screw up the time trend again, on my epithet will be the person who destroyed the four year time trend.
The first thing Paul said to me when he saw me after the 1997 data, is you dropped my question. That was first -- not hello. You dropped my question.
DR. IEZZONI: It may be that MEPS should change their phrases.
MS. MADANS: I agree, and I think that's exactly what you should tell them.
DR. IEZZONI: They don't have four years.
MS. MADANS: No, I think it's something we need to work towards. And it most interesting that we get different estimates, because it is the same population, and we can look at that, and we should.
DR. IEZZONI: But the questions are phrased slightly different.
MS. COLTIN: And you should include NCVS in that.
MS. MADANS: We will once we get the data from HCFA. There was harmonization with NCVS and MEPS, because they were going to use the NCVS. We are still working out some of these integration issues. But at least we are talking about them.
DR. IEZZONI: Yes, I think especially for disability, where there are kind of societal issues on trying to track it and understand it better. That spending tax dollars to collect data sets that can be used together is better. So I guess if our committee can make a statement to that effect. Maybe we should think about writing a letter about that to the Data Council. A short letter.
MS. MADANS: I think you should write it to Ross and to Jane Gentleman, because they can have a more direct effect on changing the questions than working through the Data Council. Not that the Data Council is not a place you should write, but I think you might do better.
MS. COLTIN: I also think it's helpful to be able to point out very specific questions that you need to be aligned, rather than the general problem. We did actually succeed in getting the wording changed on the pneumonia vaccine question. It is now the same in all three surveys.
DR. IEZZONI: Well, I think the issue about the four year time trend is a very valid concern. And since MEPS is shorter, it does suggest the potential direction of change, but we won't go there right now.
Does the committee -- Elizabeth, you're not going to be here tomorrow. Do you have any parting words of wisdom for us?
MS. WARD: No. I think I was sitting here thinking sort of where do we want to go next, because I don't work in this area. Other committee do, or are more constantly aware of what are the questions raised.
DR. IEZZONI: Well, basically the latter presentations this afternoon have not been to use the school phrase, on task, because they haven't been specifically about ICIDH. I think that we do have a pretty clear feeling for the ICIDH part about where we are going to be going. Which is in July we will hopefully be hearing more about what happened during the June training. We'll be hearing about this a little bit tomorrow morning from Paul and Gerry.
We should also probably try to bring in some of the disability advocate groups. The privacy issues need to be raised. And so I think we will probably be trying for the July meeting to get a little more richness. I think we were going to try also for July to maybe have some other provider groups come in like the physicians, nurses, other people who might be asked to collect these data, and comment to us on what that might be like. And then in the fall we would hearing actually the results of the pilot study that is going to be done.
DR. NEWACHECK: Lisa, do you think it's premature -- most of what we have been hearing today, and in previous sessions has been more informational presentations. Do you think it's premature at this point to start to put together a work plan for this? That is, where we are going to go with it. Or do we still want to get more information before making a decision? Because we have talked about whether we want to even pursue this project at all in some ways in the longer term anyway.
DR. IEZZONI: Paul, I guess I was acting on the optimistic notion that by the end of -- is the meeting, Patrice, in October or September? It's in October. In October, we would be at a point that we could recommend whether pilot studies or further work should be done on ICIDH. That we're going to be very, very focused on ICIDH. So our work plan is very ICIDH-centric at this point.
So I think in terms of work plan, the major thing is to make sure that Susan and other staff folks know who we want to hear from. So at the end of the October we can make a specific statement about what we think we have learned about ICIDH, and whether in fact it is useful for Don Lollar to help fund and maybe ARC help fund or whoever help fund pilot studies in this area, or whether we think the barriers are just too great, or what we think we've learned.
So I certainly would very much welcome other committee members suggesting who we should hear from, because we kind of got slightly diverted midstream by snowstorms and things happening in January, and not having the second day to kind of sit around as a group.
MS. COLTIN: Not really being well versed in this field, I don't know who, but I know what I would like to hear about. So that if others knew who was doing work on a particular topic, that would be helpful. Because ICIDH or ICFD as it's going to be is a classification system, one of the things that was mentioned today is that there might be measured created from it, particularly looking at for instance the qualifiers, and changes in the values of the qualifiers.
Is anyone doing work in that area? Because it would seem to me that one of the uses to which ICIDH information on administrative data might be good is to support medical effectiveness studies, to support quality evaluations. In order for these data to be able to be used for those purposes, you want to know if you can look at changes over time. So is anyone looking at that? What do we know about it? And this is at the person level, not at the population level.
The second would be to be able to compare the ability of ICIDH to look at changes over time with alternatives like the person-specific samples. I don't if any of those other than the longitudinal study on aging are over time on the same individuals, but to the extent that questions that are asked of the person can tell us somebody about changes over time versus observations that are made by professionals, and changes in those scores can tell us something over time.
DR. IEZZONI: That reminds me that there was one other group of people that we did want to hear from in the summer, and that is the international people who are involved with this. Because I think, Kathy, from what it sounds like, the international people have been doing a lot more thinking and work with this in terms of studying it and assessing it than we have in our country. So I think we should not just say has anybody in the United States done this kind of work, but has anybody worldwide done this kind of work, and try to find out about it from everybody.
DR. NEWACHECK: Yes, on that topic I think it would be helpful to hear if any other countries are doing what we are trying to do. I'm not talking about the measurement issues now, but trying to incorporate ICIDH into administrative records in some way. That is, are there parallel efforts going on in Canada or Britain or places like that, that we might learn from in terms of feasibility and such.
DR. IEZZONI: We had the gentleman from Canada at our February breakout. I'm blocking on his name.
MS. GREENBERG: Michael Wolfson(?)?
DR. IEZZONI: Michael Wolfson, yes, exactly. And he was very helpful. I'm actually hopeful that maybe we can have him back for the summer, because I thought he was really good.
MS. GREENBERG: I think in Canada they are looking at incorporating or using ICIDH in rehabilitation data. And we could possibly get someone to talk about that. But I don't know that they have experience using it. I think they are looking towards using it.
MS. COLTIN: The other thing is that brought to mind something that Don Lollar had suggested this morning about if you were going to pilot it, if you decided that it looked worthwhile, and you wanted to gradually introduce it, that there were a number of ways you could do that, either by age group or disease category or type of provider. And you think they are basically doing it by type of setting. I think we might want to talk about the relative merits of different approaches to that.
DR. IEZZONI: If we do in October decide that we should propose pilot studies, I think it would be good for us to be as specific as possible. And maybe have Don back, because it's obvious from what he said that he has thought about it a lot, and thought about the ways that one might do that.
MS. WARD: I was going to say for me, that's the part that I am most interested in, how practical is this going to be. Because that's always my concern, that we are designing some ideals, that there will be such an inertia problem, we will be back to the oil tanker issue, and all the multiple sets in long-term, that it just seems hopeless.
DR. IEZZONI: That's why I at least have said from the very beginning of this initiative that I don't expect a recommendation to implement anything to come out of what we are doing.
DR. NEWACHECK: But I think you are raising a really good point, Elizabeth, but I do think it is part of our job, is to push the envelope a little bit, and to get people to kind of start thinking a little bit ahead. Even if this isn't ready for prime time, we still want to encourage people to think more broadly about it.
MR. HANDLER: [Remarks off mike.] We're talking about modifying a HCFA form, someone, somewhere is going to prepare the information for the OMB clearance package, with the proper justification, with all of the background information to justify why an item like this is being added.
MS. WARD: It's already part of the standard.
DR. IEZZONI: Right, that's what I was going to say. There is a placeholder for it on the electronic data transmission standard, not for ICIDH specifically, but for functional status.
Again, I think that I would very much welcome suggestions as Kathy has made for the types of people that we would like to hear from. Because I feel kind of badly that people fly in for these meetings, and we should try to have a full two days kind of packed for things. And so in July maybe we can have a dense program.
DR. NEWACHECK: I wanted to commend the chair of the subcommittee and the staff, because I think these have been really interesting sessions. We have learned a lot.
DR. IEZZONI: I think today has been a really good one. It has been, and I thank our speakers. I think it's been a great day.
DR. PLACEK: We are inviting over 200 people to our last two weeks in May rollout, including studies 1, 2, 3. Those are the professional associations. So we may be able to do that in July, rather than wait until October.
DR. IEZZONI: That would be great. That would be super.
Well, I think one of the things that I would like to have as the final kind of accomplishment of this is that if we polled the Committee on Vital and Health Statistics, they would all know what this is.
DR. NEWACHECK: And they would know that ICIDH is really the international classification of functioning and disability.
DR. IEZZONI: Disabilities and functions, functions and disabilities. We've got to start practicing having this acronym roll trippingly off our tongues.
Any more comments? I do thank you. I think it's been a productive day.
So we will get together again tomorrow morning at eight. And those of you can be here on time will be, and those of you won't, won't.
[Whereupon, the meeting was recessed at 4:55 p.m., to reconvene the following day, Friday, April 14, 2000, at 8:00 a.m.]