[This Transcript is Unedited]
Room 705A, Hubert H. Humphrey Building
200 Independence Avenue, SW
Washington, D.C. 20201
DR. IEZZONI: Our subcommittee has dwindled even further today. Kathy Coltin was unfortunately unable to get a flight out of Boston. I referred yesterday to some of the problematic weather, and so she sends her apologies. It's going to be Andy and me holding down the fort today. Nonetheless, we will have complete transcripts of this. Every word will be recorded, so every word will be able to be read by the subcommittee members. So I hope people will continue to realize that we're really happy to have you here, and the Internet too.
MS. GREENBERG: I think that Kathy said that she was planning to tune into the Internet.
DR. IEZZONI: Great. Well, yesterday was terrific. We really, really benefitted from the folks that presented yesterday, and got into a really interesting dialogue. So today is our second day, and we have a variety of terrific folks who are going to speak to us today.
I would like to start with the usual way of going around and introducing ourselves briefly. My name is Lisa Iezzoni. I'm professor of medicine at Harvard Medical School, and I'm chairing the subcommittee.
DR. QUEEN: I'm Susan Queen from HRSA, and staff to the subcommittee.
DR. KRAMER: Andy Kramer. I'm professor of medicine from the University of Colorado, and a member of the subcommittee.
MR. HANDLER: Aaron Handler. I'm in the statistics office of the Indian Health Service. I'm support staff to the subcommittee.
DR. RUGGIERI: My name is Alex Ruggieri. I'm a research fellow at the Mayo Clinic, Rochester.
DR. STARK: I'm Susy Stark. I'm representing the American Occupational Therapy Association.
DR. GUCCIONE: Andrew Guccione, American Physical Therapy Association.
DR. THREATS: Travis Threats, American Speech, Language, and Hearing Association.
DR. USTUN: Bedirhan Ustun, World Health Organization. I am responsible for the classification of the SMS survey.
MR. MC LAUGHLIN: I'm Chris McLaughlin from the American Psychological Association, where I'm director of professional development demonstration projects in the APA Practice Directorate.
DR. PLACEK: I'm Paul Placek from the National Center for Health Statistics.
DR. HENDERSHOT: I'm Gerry Hendershot from the National Center for Health Statistics, and staff to the subcommittee.
MR. HITCHCOCK: I'm Dale Hitchcock with the Office of the Assistant Secretary for Planning and Evaluation, and I'm staff to the subcommittee.
MS. GREENBERG: Marjorie Greenberg from the National Center for Health Statistics, CDC, and executive secretary of the committee.
DR. WOLFSON: Michael Wolfson from Statistics Canada.
DR. SELTSER: I'm Ray Seltser.
MR. GREEN: Tim Green from the Medicare Payment Advisory Commission.
MS. ZINESTER(?): Marie Zinester, American Speech, Language, Hearing Association.
MS. HANSON: Michelle Hanson, an intern at the National Center for Health Statistics.
MR. GRISS: Bob Griss from the Center on Disability and Health.
MS. LUX: Jayne Lux, American Psychological Association.
MS. HAWLEY: Judy Hawley, Minnesota Chapter of the American Physical Therapy Association.
DR. THREATS: Susan Colbert Threats, Washington University School of Medicine, St. Louis, Missouri.
MS. UPCHURCH: Patrice Upchurch, NCHS staff.
DR. IEZZONI: Okay, terrific. Now we are very lucky, because Dr. Ustun just happened to be in town, and we happened to be able to snag him for a little bit of time this morning. We heard yesterday, and I want to make sure that it is true, that you are the person who is largely responsible for the classification that we are talking about. We know that other people around the room have also contributed to it.
So we had a chance to talk briefly before the meeting. So Ustun is going to have some prepared comments. And we should feel free to question him further when he is done, because we have him here for just a brief amount of time.
Thank you.
DR. USTUN: Thank you, Dr. Iezzoni.
Yes, I am responsible, or rather I suffer from ICIDH myself. And I'm very glad to be here. I think this point is a milestone that we have reached in the revision process that a nation is seriously considering using a classification system for reporting the function status on administrative records. So I think that is a level that we want to reach.
It is very timely from the point of view that the World Health Organization is now going to approve ICIDH by 2001 May World Health Assembly, with an executive board resolution to gather unified data from the member states, to report on the health systems of the countries.
So I wanted to focus on the proof of the pudding, not in the making of it. The proof of the pudding is in its eating, so that it will live up to the standards or the goals that we have set for tool that we are developing. And what I'm going to say to you is summarized in this slide. So I'm going to tell you what I'm going to tell at the end, at the beginning, that ICIDH has been revised over the last 10 years using a large international participation. The number of countries that have commented on it is more than 80 now.
At least 40-42 are actively engaged in carrying out either translations or field trials with that. It has been extensively field tested. You might have heard that there was an alpha draft, beta-1 draft, beta-2 draft. But this analogy borrowed from the computer technology, we have gone through three successive field testing traces in that.
We really wanted to put the horses before the cart, because the first ICIDH was approved by the World Health Assembly for field trials purposes. So it has got not very much of a binding power for its use in the member states. So since it was for field trials purposes in 1975, people felt free to use it or not to use it. So this time we wanted to have the field trials first, and approval will come for good.
The second point I want to emphasize is that we really wanted to have the revision process being guided by the scientific principles, such as taxonomic guidelines, the internal logic guidelines that satisfy international institutes. And practical utility was the real touchstone. We wanted to see how feasible the use of classification in real life, ease of its use, and also ethical guidelines so that it cannot be used as a labeling tool for people with disabilities.
Of course, we are at the last phase, focusing on the reliability of its use, because it will be used by many people. So the test, re-test, or used by different users in a standard way, in a reliable way is important.
And of course validity is an important thing, because we really want to see whether it predicts something, or it explains or accounts for that the health care expenditure or length of stay in health care facilities, or receipt of disability benefits, or whatever. And I think some of the former consultants from WHO reported to you on that yesterday. Jayne Lux gave you some of the literature reviews that we have made on that.
But most importantly, the World Health Organization is going to use it as a framework to compare data from different countries. So cross-country comparability is an important tool. So we have engaged in a cross-cultural applicability research that really lays the foundation of that.
And of course the last point is when the rubber hits the road, will it really roll on, and will it be really useful as to identify the clinical outcomes for service provision. And for example, the World Health Organization has just released the report on health systems of different countries, which includes a summary measure of population health, which is important. I think yesterday Dr. Wilson has referred to that possible use of ICIDH.
And currently we are not only reporting the life expectancy, but also adjusting it for disability. That is the time spent with disability during a life span of an individual.
And of course no tool is perfect. When it's being used, there are different uses and different users. So there will be a need for application guidelines, training tools, and other types of quality assurance. So it won't come as a single end product, but there will be multiple products, and a lineage of products, or a family of products.
So these are the messages that I want to give at the end of this presentation. I hope I will put this into the 30 minutes or less that I am allowed.
The image that we want to create is WHO wants to come up with a set of family of international classifications. In that, I would like you to look into the middle block. There is a first sibling of the family in that this is called ICD-10. It was first created to report causes of death, but later it has also been adapted to use morbidity as well.
But that reports the disease or diagnosis. We want to supplement it with what happens after the disease process, what follows that in terms of consequences, impact on the person's life, and the participation in society. So we want to look beyond the disease into the person's life as to in terms of personal activities or participation. So couple it with the different classifications.
With that, we really want to create with the ICIDH, a scientific basis to understand and study the consequences of health conditions, and establish some sort of a common language, a rosetta stone if you will, that will translate the findings of different functional assessment tools to improve communications, and allow us to compare data across countries, health care disciplines, services, and times.
Now of course we really wanted to have it as state-of-the-art computerized too, but it requires a systematic coding scheme so that we can gather data.
And of course different users of ICIDH could be envisaged. Basically, we think that it will be used in the health care sector, but it might be used in education, labor, and other sectors as well. But in the context of health, it might be used to identify health care needs, matching treatments or interventions to those needs, measuring outcomes and their effectiveness, and also it can be used to study financing, setting priorities, or allocating resources, because different types of disabled user functional status might need different resources.
And you are all familiar with the key constructs of ICIDH. It's got a third part type classification. In the first dimension, you look into the body functions and structures. In the second dimension we will look into the person as a whole, and the activities of the person. Formerly, this level was called abilities, or disabilities, but we changed it to the word "activity," because activity is actually what you can observe. Ability is some sort of an inferred construct. We really wanted to have some sort of a real live focus in that dimension. So that is the change.
And participation was the former handicap, and it became the H-word in this country. So we changed it into a positive concept, but not only for political correctness purposes, we also changed the nature of the classification. We can then identify the areas of participation which is restricted. So that you can build specific interventions to those areas.
For example, if it is in the areas of the self-care in participation, for example the toilets for disabled could be put in public places, and that is an issue for participation. So in that sense, we really wanted to identify the specific areas of domains of life in the social participation level.
And as you have seen, the name from the original classification which was impairment, disabled, handicap is changing to body functions, activities, and participation, to neutral terminology. And this also allows a possible user of the classification, for those who want to measure the positive aspects of these dimensions.
But the negative aspects are still there, which is called impairment, activity limitations, and participation restrictions. We use different terms than limitation or restriction to differentiate between AMP in terms of the different kinds of negative consequences.
And for those of you who are visually oriented, I am using this set of items. As you know, the activity limitation or disabled symbol has been person in wheelchair. So we take the impairment dimension as what is wrong in the level of body. If you take the participation or this former handicap dimension, what is imposed onto the individual by the society as some sort of hindrance, and by rules, regulations, social attitudes, norms, or environmental barriers, physical or social.
So the terminology, these words have been used interchangeably, and simply giving you the Humpty-Dumpty theory of terminology from the, Through the Looking Glass, the Humpty-Dumpty says, "When I use a word, it means just what I choose to mean, neither more nor less." So many people for whom I have introduced the ICIDH with their spontaneous ideology of understanding the disability, they came to have an understanding of the ICIDH almost immediately.
I think these have some sort of terminological restricted means that we use, and throughout the classification and its introduction we have given operational definitions as what these terms mean, rather than their daily usage in their terminology. Some of the confusion comes from that, but some of the confusion comes from the power of disability. Even in this country I think there are 47 different federal definitions of disability. So I will come back to that at the end.
In earnest, the ICIDH revision in fact started in 1990, but I was responsible for the last five years, and since 1995, November. The first product was a revised ICIDH with a forward in 1993. The basic change in those four years was basically to delete homosexuality, and revise the forward to that. It was a list of impairments, and a list of disabilities, and a list of handicaps, which were only 7 in number as handicaps. And so really reviewed the concepts and data which was collected by ICIDH-1, as I refer to the 1980 version.
And then we drafted a classification with some definitions, because a classification such as list does not mean much more than a laundry list, and people misuse this by having them misunderstand arriving at the consensus like that. Plus, we drafted some taxonomy rules, so that how these categories could be defined. And what we call as a word operationalization, we tried to everything by some sort of measurable constructs that lends itself to questionnaires, surveys, or clinical inquiries, whatever could really be measured by way of questioning or testing the performance and such like.
And we drafted ICIDH, and there are different ways of skinning a cat, as you say, so there were options for example, for measuring memories, for classifying memory. You can have it by names, pictures, or digressive(?) memory, or mental memory. Have a short-term. We drafted these options, and there were numerous options for example we tested in terms of putting a theme in activities or in body function, and collected the pros and cons of each option.
And later in the course from alpha version to beta-1 to beta-2 these options have been reduced in number. And we have tried to identify how people could use it. And these have been tested widely. We haven't reported the whole test results. I think that is some sort of a health criticism on WHO's part, because of the extensive bureaucracy involved, in the changes in WHO's procedures.
But mainly, these tests were carried by different centers. And as I told you, there are 42 country and different NGOs, including disabled people's NGOs and governmental organizations, intergovernmental organizations, scientists, clinicians, or people with disabilities. I think in the last version of ICIDH we put it open up the Web. And the over the two years it is on the Internet, not only in English, but five other languages. Each person can access the Internet, and comment the totality, or individual item level of ICIDH. So we are benefitting from that kind of a public input to the ICIDH as well.
And by 2001, in January 16, the executive board will discuss the issue. And by May we are expecting a WHO resolution to approve the ICIDH as a framework for reporting the non-fatal health outcomes in general for summary measures of population health.
The work on ICIDH will not end there, because we really plan to go on with applications. And there are several projects of making an application guide, or a coding tool, or several other tools for different uses and users.
What has changed in terms of ICIDH is a conceptual transformation which I hinted to you while introducing the concept. And we really wanted to be very user friendly, and took the user need as our front guiding principle. And to tell you the truth, the changed need came from people with disabilities. I think in 1985 in the United Nations Assembly, and that was the decade of the people with disabilities, they proposed that ICIDH should be revised. So we are following that order of the United Nations Assembly.
But from that advocacy ground, we wanted to go into the science and find the principles. So can we measure the services that are provided by a society to the people with disabilities of every kind? And can we also put the classification framework into the summary measures of population health to gather evidence to inform policy?
And it was formerly called consequences of disease. That was the classification. And we really create some sort of a, if you will, theoretical framework that the causology could be studied scientifically, rather than preimposed. So if you can have the disease or health conditions on the one hand, and the associations on the other hand, and study the result in different models, rather than imposing and predetermining in a primary way.
In other words, the field had several polarizations, and you can imagine that. There are two different basic models of disabilities, one a medical model, if you will, and a social model. That was one of the greatest polarizations that we faced in the first phase. And of course when you are trying to make an international classification, there is a tension between a global classification and local versions of it, because the international world is somewhat different.
And plus, there have been two different versions of ideologies among people with disabilities that they wanted to adhere to minority model. People with disabilities as a minority, which have different groups with needs and rights, plus a universal model that everybody might become disabled or is disabled in one way or another. So that was another tension that we faced.
Of course if you look into the literature, there are different models in this country, 90 models in Canada, care(?) back model in Japan. We had a model in Germany. So everybody who is writing about disability, has come up with different types of conceptualizations. We usually have to have some sort of models to see whether it can have a common language.
What we could come up with is this multiple web in a sense. At the beginning, and since it is in the context of health, we take the health condition. By health condition, it might be a disease. It may be a disorder. Even some people would like to put pregnancy, aging, stress into it. We are testing that it be used there.
But the basic lineage of the three part type classification is here, body functions, activities, and participation. Formerly this was a unique directional progression from body impairments to disabilities to handicaps, whereas now we have some sort of bilateral arrows to show that sometimes when you put in these islands for example, they may lose their interpersonal skills. If you put somebody's arm or leg into a cast, the muscles will atrophy. So the physiology might go in the reverse direction. It is a possibility. In order to allow that, the arrows are indicating both directions.
And one of the interesting additions to the classification is that of the environmental factors. It was the environment that we are living that might impact on our participation, because of the cultural and environmental barriers. But also on activities and body functions such as the quality of the air, or the noise and the amount that we are in, and such like.
Plus we also put the personal factors, although we don't have the classifications. But they have an important explanatory power in terms of seeing the person's involvement with the health services and their health outcomes.
I had mentioned to you about the two models of disability. One, on the one hand is the medical model. I summarized it on the right on here. That medical model of disability disease is a personal problem which requires medical care, which is delivered by individual treatment by professionals, which aims at personal adjustment, and tries to change the behavior of the person. It is a matter of care and health care policy which aims at that patient.
Versus the social model of disability, which in a sense goes to the other extreme. It sees this as a social problem. It's a matter of biopsychosocial integration which requires social action, individual and collective responsibility to manipulate the environment, and change the attitudes of people. And it's a matter of human rights. It's a matter of political struggle for social change.
Both are right if you see in their own right, from where they were angled. And I think we take a sympathetic or conciliatory type of approach to bridge these two models, because we need both to relieve the impairments go into the participation, or remove the barriers for participation restrictions. So that was one of the debates that we had.
But the other foundation was in the philosophy of the changes that we are focusing on not only disability, but human functioning as a whole. I think that is an untapped use of ICIDH, and we hope that that will open up new dimensions for use.
Plus, we are saying that we are not all using this classification for the minority of people who are labeled themselves or by others as people with disabilities, but also for everybody in the population. We are going for a universal model.
Plus, as I already indicated, this is not merely a medical or social model, but an integrated model. And it's an interactive model, as you have seen, not a progressive linear model that you sort of fatalistically progress from impairment to handicaps.
One important thing is that it really is etiologically neutral. It doesn't take the health condition and progressively arrive at a disability pattern that has attached to that. But it's some sort of parity. By that what I mean is for example, the inability to work, or not being able to go to work next week might be because I am depressed, and I cannot go to work.
It might be because I have some sort of a heart attack that I can't go to work. It might be because I have an injury or a traffic accident that I can't go to work. So ICIDH is etiologically neutral to all the causes, but you can associate them with the health conditions as such.
I think it is an inclusive model. It involves environment and person, and does not focus on Western concepts, but tries to seek cultural applicability that is basic human functioning, that is universal in a sense, across cultures.
And despite the fact that we have some sort of a jury behind it, we tried to have an operational classification that spans itself over the life of an individual, not only an adult, because also children. We have a specific password that works on children, because of the development phase in children posed some problems about the applicability. And if you wish, we can go into that in the discussion.
I think you have already been shown this metaphorical forest of ICIDH. But the basic thing is that we have some sort of three basic dimensions. And each dimension has it's domains and passes. And at the end we have some sort of a uniform qualifier now, which is much better than the former ICIDH. This qualifier has some sort of positive and negative ends. It is a five point scale between no difficulty or none at all, to a complete or total one, and there is a midpoint which is called moderate on each side, and there is a mild and severe cut point.
This is a suggestion to code these dimensions in a uniform way, so it lends itself an easy key to coding all the dimensions. If you note in modern psychometrics in item response theory, you can calibrate your instruments with these five points in a reasonable way. For example, the Dutch data has been rash(?) analyzed. They arrived at these five points in mobility dimensions, or all disability dimensions.
So this will be the end product. I think that at the time of the approval, the acronym will change from ICIDH to whatever it is called, International Classification of Functioning and Disability. And we have two different versions. One is a full version with some sort of possibly nearly 10,000 categories, because with 4 digits you can address nearly 9,999. And there are big pictures guides, with likely a short version, who have 99 categories on each dimension.
This is mainly intended for statistical reporting, but clinicians might need more detailed versions with an assessment guide. And researchers might need even more precise operational guidelines or criteria for research, and we are working on those things.
There are some specialty adaptations needed. Neurologists want to develop one for neurology, to have everything about the brain and the nervous system. And children and youth will come up with their own version, for us to see from the mother version if you will, how it would be adapted to the children. But we won't stop there, and we'll have some dedicated assessment tools, which I will refer to that if I have any time left.
Now I would like to tell a little bit about the revision structure. In this solar system with the WHO in the center, because it tried to coordinate it, not because we are egocentric, and we worked in consultation with other WHO clusters and groups and regional offices, and there is a second layer of collaborating centers with whom we collaborate heavily in terms of coordinating the field trials in different language areas, or other geographic areas. For example, the NORDIC(?) collaborating center, headed by Marjorie Greenberg, is coordinating the field trials, and all the coordination of work on the ICIDH revision.
On the outer circle we have the 191 member states of WHO. And if you can see with this system, it is not always easy to put everybody in the orbit. The French has this Descartes(?) type of a dichotomy, saying that mind is not body. It's different, and you can't put mental functions into impairments. They are activities as a matter of fact. We had some way to go with that.
And they have a system which is in the social welfare system, rather than in the health system for people with disabilities. So we had to deal with different forces, and simply giving a sketch of what other countries are facing.
In the Netherlands they are mainly involved in taxonomy, because there is an international standard for classifying all classifications in the Netherlands, but it's an interesting thing. And they want to use ICIDH at a given moment, not at as a process. So that is there in a timeless version of ICIDH.
U.S. concerns were mainly around handicap, plus the environmental factors. And we have a task force which, thanks to a grant that we have from CDC, now we are classifying all the environmental factors on that.
As you have heard from Canada, there is the concern about the AMP distinction, which is also shared with some other centers. But they have a local model, Kevec(?) model which is about the handicapped creation process, and which has been presented to you yesterday.
Australia has concerns about people with disabilities. They have a group that was basically formed by people with disabilities, so they are very keen about the ethical guidelines, and use in summary measures of population health.
The Japanese collaborating center is very much interested about the subjective dimension, the feeling component. How well do you feel around these things. And the U.K. center is very much concerned about morbidity coding, and how it differs from ICIDH coding, because some of the morbidity codings and impairments are overlapping, and that is a concern, whether that overlap is to be allowed.
The Spanish network is basically concerned about the cultural sensitivity and applicability into Spanish speaking cultures. And we were successfully bringing together 12 different Spanish countries to agree on a translation. It was quite an achievement, I would think.
And the Finnish center is very much into the international standardization of this, ISO standards about terminology and its harmonization.
So you can see that there are different gravitational forces that take this away. I had asked Dr. Iezzoni whether I should tap into one of these, namely the AP distinction that Dr. Wolfson yesterday presented to you. And I think the basic problem is that how do you see these three dimensions? Are they a continuum, a spectrum as I depicted here?
Every depiction has it faults, but if we say that anything that is in B, should not be in A, and anything that is in A, should not be in B. That is the underlying axiom, if you want to have distinction clearly as such. So if you put a vision into B, it can't be in A.
However, the current model that we have been operating under is the following one. For example, as I am talking now, my vocal cords and my brain, each areas are producing these words and such like. And as a person I am talking to you, engaging in a conversation or in a presentation. That's my person level activity.
And I am societally participating in a social event, i.e., being a witness in this testimony and such like. And I am in the United States of America speaking English to you in this environment in this hearing. So that is the same level of thing.
Now you can put it into any activity. It might be urination, and it might be the body functions, how are you behaving in your personal privacy, in your toilet. Plus in the society there are no toilets for the disabled and such like.
So I will give you an example about driving, since it has been brought yesterday. This is a real life case. I won't tell who he is, and which country he comes from. He is 27 years old. We was born as a thalidomide baby, and he has got some difficulty with fine movements, but he can drive, and does drive a standard car without any modification. However, the law in his country prohibits him from driving.
So you can call this thing in the ICIDH, and these are the codes. Because of the shortness of time, I won't go into this thing. But the thing is that the person has activity. They are using transportation as a driver. This is a long way of saying driving, and has difficulties with participation and mobility with transportation, because he is restricted. And the barrier is the law that inhibits him.
And you can think about the "Forrest Gump," which is the Oscar winning film from this country. Forrest was able to run, but later became a big runner. He ran across America. And he was able to go to the school, but he wasn't allowed, because he had a low I.Q. But at the end he became a millionaire. In the army his I.Q. was even said to be 160.
So that dimension of disability goes from one end to another. I think looking into both ways is an important thing. And how do you then differentiate activity from participation is that you look into the activity, that it is in person level, and it is based on performance of the person. It is an individual attribute. If there is a difficulty, the person is not functioning due to the underlying health condition or body functions.
And basically, the locus of intervention is at the individual level. Whereas, in participation that's at the society level. What they are measuring is the involvement in a life situation or in a circumstance, and the participation is not occurring due to the underlying health condition or activity limitation basically due to social hindrances. And the locus of intervention is at the societal level; that you have to make a societal intervention to change the attitudes, the change the law, to educate people, and such like.
So in a sense, you can take this 2 x 2 table and look into activities and participation in any way. You can take poking(?). The person is poking and is participating. But the person might be poking, but might not be given the right to participate if he is stuttering and such like.
So similarly, it might be urination, it might be driving. So what we have deliberately filled the ICIDHs. That we have some sort of a mirroring system of activities and participation, that we can look into these two dimensions in a cross-sectional way, both from the activities perspective and a participation perspective.
So if a person performs an activity and participates, that's good. If a person does perform the activity, but cannot participate, that's the locus for intervention. You have to remove the barrier. If a person cannot do the activity, and participates, that's likely you might think, but sometimes it happens. It's a good example. We can learn from it, and identify the facilitators.
If a person cannot do the activity, and cannot participate, then it's an issue for awareness and research development and such like. So ICIDH, with this "flow" overlap identifies to the sintangle(?) this floor in front of us, and so identifies the interventions that we can take.
So the field trials -- tell me when you think that I should stop -- but basically we ran through the field trials to refine the conceptual framework. And I think thing that we take pride in is that we created the ICIDH in a multilingual platform. No other classification so far has been created from the alpha state, beta state in several languages. So we borrowed the English language. Sorry for the broken English in the English language, but you are very welcome to make an American adaptation of it for your uses.
We borrowed the English as a server language, rather than the mother or a dictator language. And we addressed some of the basic questions relating to the construct and validity and semantics. We tested some key items, concepts, and alternative options, and tested the feasibility of the classification in different settings, plus some reliability tests in different settings, formats, and versions.
A field test is not the only dictating part with this slide, I want to say. We have inputs from linguistic evaluations, reports from collaborating centers, task forces, disability groups, taxonomic experts, and technical experts.
I will go quickly in view of the shortness of time. We also a very systematic study in 15 countries that is done more in that. Selected different cultures in 15 countries, and have done linguistic analysis, concept mapping, and pile sorting, key informant interviews, and reference case interviews, and focus groups first to identify where there is a unified concept about body function, activities, and participation in different cultures.
And plus, as I mentioned to you, this is a real live comment from the Website. Anybody who can tap on the Website at every level and give a comment on each item or in the totality of the classification we have database that is compiled, and so far we have more than 4,000 entries into this database. So it is open to the public for them to give comment as well.
Plus, how can we use it in real life? One very quick way is using a checklist. What we have done is since this is a long classification, even the short version is some sort of 60 pages with 99 categories at each dimension. We summarize it as one dimension, one page, and giving the salium(?) category a glance, that's body functions, body structures, activity limitations, and participation restrictions, plus environmental factors are summarized. That's one page. I have a copy I can leave for your scrutiny. And we can also -- this is for clinicians and people with disabilities or informants can administer it themselves.
Plus, there is another tool which we developed after cross-culture and psychometric testing. This is something like what you know in this country like SF-36. This is the disability assessment schedule, which covers six domains. That is, basically understanding the world around you; getting around; self-care; getting along with people; life activities such as household activities or work; and participation in society. So it covers the A and P dimensions.
These are the basic six dimensions that we found to be measurable as different rules. And it's that advanced scoring system with difficulty rating as well as frequency ratings.
So I will end here, basically showing a poster of Einstein, who was supposed to have a learning disability in primary school. And he was almost to be expelled from a Swiss school. So thank God he wasn't. But when I mention ICIDH, people are talking about disabled versus able bodied or disabled versus not disabled.
ICIDH, as you might have already understood, it's not a categorical distinction. Nor is it a percentage, that you give 60 percent disabled, 40 percent disabled. Because it is something like saying somebody is 60 percent schizophrenic. It's not the case.
And it's a multidimensional classification. When we say somebody is disabled, we say disabled for what? We only have to identify these six dimensions. When we start focusing on a minority model of people with disabilities, but as a universal coverage for all the populations, therefore it is a good system for covering the national health statistics.
And I think by being neutral and objective, having etiological neutrality, and since this is a social and biological model involving the environment and personal factors, giving operational definitions and multi-level views, it is a positive classification together with the ethical guidelines.
I will stop here and will entertain any questions if you have time.
Thank you.
DR. IEZZONI: Thank you very much for a very quick, but very comprehensive and thoughtful run through of things. We are slightly over time, but because we have the benefit of Dr. Ustun being here, I think we should go a little bit longer.
Andy, do you have any questions for Dr. Ustun?
DR. KRAMER: Well, thank you very much. It was very helpful, and it certainly has been an impressive and ambitious undertaking. My hat goes off to you for the work you have been doing.
I'm curious on a couple of fronts. First of all, do you advocate that this be completed by individuals or health care providers, or either? And do you expect that they would give the same answer? And if they give two different answers, does that matter? Is there a correct score?
DR. USTUN: The answer to the last one is that there is no correct score. I think in the sense that we are allowing this framework to be reported by different persons, either the providers or the consumers. And I think it will be beneficial to know the perspectives of both. And the research so far has shown that there is a discrepancy at times between them, but there is usually a concordance as well.
So currently we are field testing different versions, because the concordance for instance between the self-report and the informant or proxy report is giving a couple values of 0.6, which is at the acceptable level. But at this type of reporting I think if you give a free format, like a checklist and so forth, the concordance is even less.
So the thing is this is a framework. And I think if you want to use it, I would strongly recommend that whatever the data is, that you identify the source of the data; who is reporting. So I think even in ICD we have that type of lay reporting versus provider reporting type of data. So I think that should be an option.
DR. KRAMER: Right. And I guess for comparability then across countries or for national tracking, international tracking, one would want to make that same distinction, or make some requirements or something like that?
DR. USTUN: Yes.
DR. KRAMER: Another question I had is heretofore what we have heard about in terms of their validity testing has mostly what's been conducted in the U.S. And as well, we haven't heard a great deal about reliability testing. Is that available somewhere as to the rigorous testing? The reason I bring it up is in this day and age as you know, everything is being exposed to a scrutiny that never was required many years ago.
I mean as Lisa pointed out yesterday, that the ICD-9 coding system never was exposed to that kind of scrutiny. And in fact, there are still a great many issues with its consistency and reliability across raters and so forth. On the other hand, right now things like the SF-36 have had years of literature supporting them. And that's one of the big issues.
DR. USTUN: I think that is a very positive development, that we are really requiring that type of data from our classification systems, because I think it has to be reliable. And I came to the ICIDH field from the mental health field, and I did the ICD-10 mental health chapter, and was involved in DSM development, with that, defining the quasi categories of mental disorders and requiring the reliability was the only touchstone that we can build a system on. So we work on that reliability, and the approach towards reliability, but it is not the only dimension.
As a response to your question directly, yes, we are testing the classification and the instruments for reliability. That data will be published. Currently, tests are going on. For example, the reason I'm in town is that I'm going to present the reliability figures in the NIMH conference on services research for the WHODAS. And we have done studies in 22 countries, and our reliability results will be presented there.
But it indicates 0.8, 0.9 with the 36 questions. So I think we have pretty good between cultures, between raters, within instruments, reliability, which is only one dimension. We are testing it for validity, whether it predicts health care utilization, health care costs, whether it predicts work performance, and several other types of research about different types of construct validity are underway.
DR. KRAMER: And the timing of those activities as well?
DR. USTUN: The timeline of those activities, I think we have four papers in the pipeline. You know that in a publication it takes time. But one thing is that for example the cross-cultural research, it is some sort of a qualitative research. It is already in the second proof stage. Within two or three months it will be available. So you will have some sort of the alpha and beta-1 testing in that, in different countries.
And by the time this is published, we would be happy to put everything, even though the time is not right for scientific publications, on the Web, so it is there. By the time we have the meeting in Madrid, that data will be made available to collaborating centers for their recommendation.
DR. KRAMER: I could keep going, but I don't think I should.
DR. IEZZONI: Let me just ask one question, and then I'll just warn the audience that I would like to open it up, in case anybody in the audience has any questions for Dr. Ustun. I am a little bit curious about the process, because obviously you've been working on this -- you, the WHO global you, and then you, the personal Dr. Ustun you since 1995, and 1990 respectively.
And yet we heard yesterday from Dr. Andersson that there are a variety of disability groups that he feels are important to comment, who have not yet taken the opportunity, and have not yet commented. We hear some concerns about the structure of the classifications from our Canadian colleagues.
Obviously, you can't please everybody all the time. But we also hear that there is kind of this train rolling out of the station that will end up next June or May, I think you said 2001, with a global approval. How do you plan to kind of reconcile some of the fairly large classification issues that people have raised? Is it important for you to reconcile them at this point? Do you think it's a make or break for countries implementing the ICIDH?
DR. USTUN: Yes, as you rightly said, you cannot please everybody, but we have opened it up for everybody. As I told you, the revision process, at least in the last five years, has been going on, being open to all collaborating centers. We tried to collect systematic data about it.
And I put up this slide with the Tower of Babel so that you can see the impossibility of the mission that we are facing. But the thing is what we would like to create is some sort of an broken English version, if you take the analogy. I have been labeled as the metaphor man in this group in an earlier hearing.
But I think broken English is much better than an unused Latin or Esperanto or Yiddish for that matter. So our philosophy is more of that kind of approach. Plus, I think we would very much like to have not a solid, frozen product, but a process that goes with it. So if you take a software analogy, you can version 1, version 2, version 3. That goes with an update mechanism, so that we have some sort of a standard procedure, and some sort of backward and upward comparability, with different versions and such like.
So that is more or less the blue print of how we think we could deal with differences of opinion. And I think currently if you see these successive versions from alpha to beta-1 to beta-2, and if you compare it with the ICIDH-1, I think the progress in the case that we have more consensus than the original ICIDH-1, and the alpha draft and beta-1 draft, and it's increasing.
So I think that the final version, if you take it that way, a significant proportion of the users will be happy with the common standard. Of course, even with ICD, if you take it, Dr. Iezzoni, Americans have a CM, Canadians are planning another modification, Australians have an AM, the Brits have a nonspoken BM. So it is everybody takes the structure and modified it to their needs, to tailor their systems and so forth. But our proposal is to make the driving force behind these modifications be the driving force for the update and revision of the classification.
DR. IEZZONI: Questions from folks in the audience.
MR. HANDLER: During your presentation, you showed a summary sheet. I guess that's like a checklist to evaluate an individual person. Is that the way that this system is to be used, like a checklist? Or is it one code number to identify a disability? Which is the best way to implement this?
DR. USTUN: I think it is not a classification of one person goes into one box. It provides several boxes that identifies a profile for a person. And for each of these boxes in fact, the analogy is a ruler or a measure that could be standardized. So at least there are three basic dimensions of the classification as body functions, activities, and participation. And each one of them will have five or six salient domains around which you can set these rulers. So in a sense you are going to have a summary profile of an individual, rather than one single code.
MR. HANDLER: What I was leading to was thinking in terms of using this system on administrative records, so there would have to be a certain amount of space on a record to put these codes.
DR. USTUN: I think there should be at least 15-18. It depends on the parsimony that you agree with the different users. But there will be more than one single code, definitely.
DR. RUGGIERI: I enjoyed your presentation very much, and it is certainly great to hear it come from the original source. In this country we have a piece of legislation which is sort of looming over the health care information domain, which is referred to as HIPAA, which stands for the Health Insurance Portability and Accountability Act.
One of the stipulations of that legislation is that health care information migrate towards standards. And I think one of the functions of this committee down the food chain is ultimately to provide information and advice that will support those decisions. But it will incumbent on the secretary of HHS to adopt standards.
And as part of the guidelines for making that decision, it is suggested or I should say powerfully suggested that standard setting organizations be involved. In the United States we have ANSI, American National Standards Institute, and the ISO has TC-215 with its various subcommittees.
My question, will WHO go through the effort to have ICIDH appropriately canvassed as a standard?
DR. USTUN: The answer is at the intention level is yes. We have been together talking with ISO to go through their procedure so that it will cover the ISO codes and procedures for the ICIDH. And the basic quality assurance procedures that we want to take under.
And the basic approach of ISO in terms of terminology and harmonization of informational terminology has been already applied, at least at the basic terms level of ICIDH. But we have not formally registered it with ISO. But with the recent restructuring of our team, we have acquired the terminology branch of WHO, which will undergo through that process, and will have the ISO approval.
But that is a part of a resource thing, and currently in the WHO administration, many of the collaborating centers know that this has also been done in a very cost effective way, to say the least.
DR. IEZZONI: We understand what you mean by that.
DR. WOLFSON: Let me preface my remarks by saying that Bedirhan I have discussed some of these issues for a while now, and we agree on some things, but not on everything. I wanted to pick up on a question that Dr. Andy Kramer raised at the beginning, which had to do with the reliability. Bedirhan was kind enough to share with me the proof of this CAR study, called cross-cultural applicabilities studies.
And you weren't here, but I showed a slide yesterday that contained the following quote from the study, which you were the lead author of. "The classification of items into the three levels, impairment, activities, and participation was not sufficiently clear to all participants; 54.2 percent of the participant classifications did not correspond to the actual classification in the ICIDH-2." That's on page 13.
I guess two questions. How do you perceive that problem? And this second one is given the process of working this through the system, and seeking WHO Assembly approval in May, assuming problems such as this still exist in the version that is associated with the assembly resolution. How would you see the process unfolding afterwards to address concerns like that?
DR. USTUN: Thank you very much for bringing this up, because it allows me to clarify the issues. Of course I wasn't there to shoot myself in foot by publishing this, saying that 52 percent or 54 percent doesn't categorize them rightly if you give them. That would be called pile sorting. It was a gameful exercise where you write the names on the items on cards. You ask the participants to make a card sorting. You see something called urination. Then you put into body functions or activities or participation.
So that's is the type of an exercise that you do. And that shows the shortcoming of ICIDH-1. Unless you give operational definitions by what you mean by that, people will not go for one single term, but we have provided 92 terms which are fixed across cultures, and allows each culture to put 50 terms of their cultural relevance, and to sort that out.
So in a sense that card sorting, i.e., the former use of ICIDH-1 in an international way would have yielded that level of concordance among users. We believe that by providing operational guidelines, we increase it. And in real life reliability studies, if you for example use the checklist for WHODAS, I have already quoted the reliability figures which goes much beyond. In fact, this is a concordance rate as percentage corrected for chance. It's less than 0.2.
So that is the level. We are saying that you can increase the precision error liability by providing operational definitions as to how to measure, what to measure, and how to report your measurements.
DR. WOLFSON: Just a quick point. The items were from ICIDH-1 or from 2-beta-1?
DR. USTUN: ICIDH-2, beta-1. But by definition that also has the ICIDH-1, but without any definition, just label names. So it's this natural taxonomy of people by gut feeling into pile sorting doesn't suffice to make a proper classification. So that's why we are going in the operationalization with that.
The second question is I think we have built a more elaborate system now, good enough than any other estimate method that we can use in a comparable way across cultures if you are ambitious enough to make disability adjusted life expectancy type of measures for population health as summary measures. Currently, there are different ways of doing that, and Michael is an expert in that.
And the IOM last year made a report in this country, and then after we had a meeting in Marrakesh to discuss the different aspects of it from ethical, philosophical, to practical users of summary measures. I think there is a great need to fulfill the need for the paradigm shift that public health should not be measured only by levels of mortality.
Formerly, we were only looking into top end causes of mortality. Heart disease kills. Breast cancer kills. But we learned looking how people live. And summary measures are such that combined mortality with functioning and disability, we merely have to get into this next round of reporting. And if you don't have data, you have some sort of a quick survey and estimation method. But if you install a system like ICIDH that gathers data from the millions of encounters, and you have this mechanism built up in a more reliable way.
So at this stage, that is the ambition. I think it's a process. I think that's the most promising candidate as far as I can see as the framework.
DR. IEZZONI: Andy has begged to have one more question. It will be the last.
DR. KRAMER: I just wanted to ask you, do you think that physicians, as you know them, and physicians in the U.S. who haven't had a background in this type of work, could fill use the ICIDH? What do you think it would take them to be able to do it? How long would it take them to do it? You know how physicians are trained. You know what it would take to bring it around. And we're talking about something here that would be an encounter-type form.
DR. IEZZONI: And let's say not rheumatologists, not geriatricians, but primary care doctors.
DR. USTUN: I think that's a very good command, because I was just going to say that geriatricians actually it, or rheumatologists do. But I would very much like to draw your attention to for instance the group(?) charts that Wonker(?) has developed. There is good enough evidence from Dartmouth Medical Group that group charts in primary care has been used.
I have international evidence from a primary care study that I conducted in my former life that physicians' assessment of case mix relies more on functioning than specific symptoms of mental illness. So in that sense, they see if somebody is unable to perform usual activities or work, and by six questions of this type, physicians assign case mix better than going through a long list of mental symptoms.
So in a sense it is more clinician friendly as an approach, I would imagine, but it will take a time. That will really require some testing, and I think there has been quite good testing in Dartmouth with group charts and other things.
DR. KRAMER: No testing here though? You've never tried to give this to a group of physicians or train a group of physicians?
DR. USTUN: With this one we are testing in Italy. A whole region of physicians of primary care are now using the checklist for example there, and they are taking it on and using it. In a sense it is a matter of a change in ideology. But I'm not saying this because I'm producing the ICIDH, but we are trying to sort of meet the current practice with the expected level of standard. Physicians should not only look into the disease symptoms.
DR. COLBERT-THREATS: Just to follow-up on your comment there. I actually think looking at the scale, and I've had some time to look at it a couple of times, that it would be of use to primary physicians. I practice primary care at Washington University. And if anyone has ever spent any time, I see about 3,000 patients a year. I'm the one who actually fills out all those forms, not the orthopedic doctor, not the rheumatologist, sometimes the gerontologists. Sometime the gerontologist is really acting as a consult.
The current forms that we use to rate disability are ridiculous. They are three pages long, and they don't address issues what you are trying to address, which is function. They say things like if someone can move their arm 20 degree, then they are more disabled or less disabled than someone who can move their arm 5 degrees.
So I think that this is a good assessment, and it is going to be a useful clinical tool with the proper training, and hopefully application that universally when people are rating disability for the state, that we would use a form like this.
DR. IEZZONI: Yes, that's in the context of disability determination for disability pensions, benefits. We are thinking about adding it to the HCFA 1500 form or to the UB-92 Medicare bill.
MS. GREENBERG: I just wanted to make sure I heard you correctly. When it was mentioned about giving a person one code versus multiple codes. And you pointed out that the latter is much more likely, and that you are looking to profile a person. Did you say 15-18 elements?
DR. USTUN: Minimum I would imagine. Look, we are prompting for the WHO, and I don't know your HCFA 1500 form, and I would be happy to see and do a consultation with that. But I spoken with Don Lollar and Jeff Copeland of CDC to see whether the activity limitations might be the first part. In WHO our mantra is saying that if you are going to use ICIDH, all dimensions simultaneously so that we can see the interactions between them.
So I would think that if you go with the way we have developed WHODAS, you can summarize each one of the dimensions such as body functions, activities, and participation in each five or six domains, like cognition, mobility, self-care, interpersonal relations, work, usual activities. So each person could be rated on such dimensions. So that would be the way to go.
If you choose, I would suggest that we can make pilot experiments as to how best that could be. I wouldn't go into grand scale implementation, but I think nobody would do that.
DR. IEZZONI: Not in their right mind. We're not thinking that.
This has been great. Thank you very much for allowing the barrage of questions, and for sharing your thoughts with us.
I, as the chair, have been wondering whether we should take a break now, or just move into the next panel. What do people think? We have a break scheduled for 11:15 a.m., but we're a little bit behind the schedule, and I hate to -- why don't we start then with the next panel, since I don't see a groundswell of people feeling that we should take a break at this immediate time.
Chris McLaughlin, you are the first on our list.
MR. MC LAUGHLIN: I've been sitting here patiently, because I'm hoping that this testimony may respond some of the issues that all of you have been raising.
Good morning. My name is Chris McLaughlin and I'm director of Professional Development Demonstration Projects in the Practice Directorate of the American Psychological Association, which I will refer to from now on as the APA. The APA is the largest scientific and professional organization representing psychology in the United States. With more 159,000 members, APA is also the largest association of psychologists worldwide.
Psychology is a doctoral level health care discipline that embraces scientific research, measurement, and psychometrics, and clinical health services in a wide variety of areas and setting, and with a broad range of populations. Particularly relevant for today's discussion are the areas of health psychology, rehabilitation psychology, neuropsychology, forensic psychology, and clinical psychology.
Our combined scientific and clinical expertise places us in a strong position to evaluate the potential usefulness of the ICIDH-2 system, and to comment on the matters that you are considering today.
My comments are based on APA's experience of participating in the development, review, and field testing of the ICIDH-2 classification system over the past six years. APA has been formally in the ICIDH-2 process as a member of the national task force on mental health and addictive, behavioral, cognitive, and developmental aspects of ICIDH.
Represented by Dr. Stan Verant(?) on the task force, numerous APA members have been involved in the review of both the beta-1 and the beta-2 versions of the system, and have participated in several of the international ICIDH-2 revision meetings. Currently, APA is one of the largest organized participants in the beta-2 field trials. We have over 50 sites in the U.S. and Canada, with multiple providers participating in various field trial protocols.
Our knowledge of the system makes us very supportive of the ICIDH-2. Of primary importance to APA is the fact that the system is etiologically neutral. The concept of etiological neutrality emphasizes that there are variable relationships between health conditions and aspects of disability.
For example, two individuals with similar diagnoses of cardiovascular disease may have greatly different functional consequences from the disease. As a result, the diagnosis alone is an insufficient basis for needs assessment and intervention planning, particularly from a psychological perspective.
I know that the subcommittee has asked presenters to address specific questions, so I will present the remainder of our position on ICIDH-2 in the context of those questions.
Question 1, would using the ICIDH-2 be the appropriate way to capture functional status information in routine administrative health care claims? We believe that the ICIDH-2 would be an appropriate way to capture function status information on health care claims. We also believe that this can be done in a relatively simple and methodologically sound manner.
In preparing for this testimony, my colleagues and I spent a considerable amount time thinking about ways that ICIDH-2 could be used to satisfy the paper requirement for a functional status measure. As a result of our discussions, we developed a simple system that could be expanded over time, but would nonetheless, begin collecting a significant amount of previously unavailable data on functional status.
In order to develop a simple system, we looked carefully at the ICIDH-2 to determine which dimension would capture the most useful clinical information. Given that we were trying to develop a system for use on claim forms for clinical interventions, we focused on the activities dimension as the place to start collecting information.
We chose the activities measures because we decided that the majority of health care and rehabilitation services aimed for alleviating functional impairment in these areas, in addition to treating diagnosable disorders as captured in ICD. We felt that the body function and body structures dimension within ICIDH are more duplicative of diagnosis, and at the participation dimension would be less frequently the target of clinical health services.
Therefore, we believe that a relatively simple way of capturing activities data could add a tremendous amount to our understanding of the functional aspects of health conditions, particularly as they are encountered and treated in the health care system.
As clinicians gain greater understanding of other ICIDH dimensions, and as payers begin to understand the importance of reimbursement for interventions to alleviate the context of impairment, data collected on the activities measures will be an important step in starting to focus the U.S. health system on function improvement, and supporting accountability in these terms.
The activities dimension of ICIDH-2 includes eight chapter headings. We feel that several of these eight measures could be possibly be combined into potentially four or five functional status measures. For example, a list of categories that we have considered include the following: health care, mobility, communication, interpersonal activity, and productivity in either the home, workplace, or the school.
In each of these categories a gatekeeping health care professional could enter two of the ICIDH-2 qualifiers, the first being severity of the limitation, which is the scale that Bedirhan discussed, and the second being the use of assistance. This high level coding would begin the process of collecting functional status data, and would be used to direct the patient to a specialized health professional that would provide a more finely grained classification of the limitations.
Question 2, what the logistical and feasibility barriers? Logistical and feasibility barriers to widespread use of ICIDH-2 vary, depending on the proposed use of the system. In the high level coding scheme that I just described, we feel the barriers to implementation are essentially educational. Gatekeeping professions would need to be trained about the components of the various activity variables, but again, there may one by five to eight of them.
We see this as a minor barrier to implementation that could be alleviated by selective review and study of the ICIDH-2 manual. This type of high level coding system that we have described consists of clearly valid, dimension level criteria, and issues related to reliability and standardization would be alleviate through the use of a simple rating of severity, that first qualifier ranking.
In order to accomplish finer grained coding using ICIDH-2, more work would remain to be done. But this appears to fall outside the mandate of creating a functional status variable under HIPAA. This additional work would require substantial attention to issues of standardization. While the ICIDH-2 is not an assessment instrument in the psychometric sense, the information it yields needs to be both reliable and valid to be useful to health care professionals.
In terms of reliability, two or more persons working independently should be expected to arrive at similar classifications when applying ICIDH-2 to the same person. Furthermore, once classified, any subsequent classification of the same individual at a later time should vary only to the extent the person has changed, and not vary as a result of the classification procedure.
In terms of validity, the data resulting from the classification should reasonably reflect what it purports to reflect. For example, an ICIDH-2 first level qualifier of slight impairment should indicate such, and not become easily confused with moderate or another qualifier.
Given the psychometric expertise that is present in the APA membership, we have approached the WHO with a proposal to develop a standardized assessment manual for use with the ICIDH-2. The aim of this proposed WHO-APA assessment manual will be to provide all health professionals with the necessary tools to accomplish the goal of reasonably reliable and valid classification using the ICIDH-2 system.
This manual will specify a standard approach to the procedure of classifying with ICIDH-2, and provide standard interpretations for the various terms and concepts employed in the system. Wherever possible, terms and concepts will be defined operationally, and in terms of observational events or independently verifiable information.
We intend to develop this assessment manual through an interdisciplinary, collaborative process, and to involve health care decision-makers and policy experts in the development.
Question 3, are there any conceptual barriers to using ICIDH-2 with infants and children? APA does not feel that there are major conceptual barriers to using ICIDH-2 with infants and children. We are aware that the subcommittee has previously heard the testimony of Dr. Rune Simeonsson, the chair of the ICIDH-2 children's task force, who presented examples of the use of codes with children.
We are also aware that he mentioned there are some limitations in the use of the current version with children. We also know that some have also commented that the system does not adequately address developmental issues. Our hope is that many of these limitations will be addressed in the final revision of the system, as a result of current beta-2 field trials. But it may turn out that a special version of ICIDH-2 may need to be developed for children. But we do think that the current version offers a framework that can be used across the life span.
Question 4, what would be the benefits of using the ICIDH-2, and how would the data be used? Health service psychologists recognize that a variety of functional sequelae may result from any given physical disorder. And this often results in functional limitations for the individual. If these behaviors cluster around the recognized signs and symptoms of a psychological or psychiatric disorder, the individual may be diagnosed with the disorder, and receive appropriate treatment, and these services may be appropriately reimbursed.
But if the functional sequelae fall beyond the systems commonly associated with the recognized disorder, appropriate treatment may not be delivered, and health care professionals may not be reimbursed for the care needed to improve functional status.
Because ICIDH-2 identifies the functional effects of health conditions, use of the system will allow health professionals to develop treatment plans that focus on a patient's actual functioning, rather than simply on his or her diagnosis. Use of the symptom would also assist with treatment document, and evaluating the outcomes of treatment.
Question 5, does the ICIDH-2 raise any special concerns about data privacy and confidentiality? A collection of any information related to the functional status of an individual needs to be subject to the privacy and confidentiality requirements similar to those currently required for medical records. Although the electronic collection of health data poses a number of challenges in general, we feel that ICIDH-2 poses no special concerns.
Question 6, do you know of any examples of widespread of ICIDH-2? One example of the current use of the ICIDH-2 system by psychologists is an intake system that is under development by the Division of Rehabilitation Psychology in the Department of Physical Medicine and Rehabilitation at the Ohio State University School of Medicine. The Division of Rehabilitation Psychology has its own practice at Ohio State. The practice plan is developing this system to codify patient limitations present at intake.
Using selected chapter headings in the ICIDH-2 bodily functions and activities dimensions, much like the proposed system I described earlier, intake providers assign ICIDH-2 qualifiers to the various measures. These qualifiers indicate the severity of impairment and the use of assistance in these areas, and are then used to develop the goals of treatment, and an individualized treatment plan to meet those goals.
In closing, I would like to address your final question, what would you recommend about the routine use of ICIDH-2 on administrative health care claims? As I mentioned earlier, APA recommends that the ICIDH-2 activities chapter codes, or a subset of chapter codes begin to be collected on administrative health care claims forms. These data could be potentially be reduced to approximately ten number fields, depending on the number of categories considered.
APA applauds the NCVHS for recognizing functional health status and the value that it will provide to clinical assessment, effective treatment planning, and outcomes measurement. APA would be happy to assist or work with the subcommittee in any way to further refine its thinking on functional status measurements, just as we intend to continue working with WHO to develop a more reliable ICIDH system.
Thank you.
DR. IEZZONI: Thank you, Mr. McLaughlin, for your comments. I think in the interest of time what I'd like to do is move through the four speakers, and then we'll have questions for everybody at the end.
So Dr. Threats.
DR. THREATS: I have taken the basic six questions, and I have turned them around in order, and combined a few, but they are all addressed. First, I want to talk about from ASHA's(?) point of view, the benefits of using the ICIDH, and some demonstration of its actual benefits.
In the manual they go through four main applications or uses of this ICIDH-2. I'm going to talk about those in relationship to communication disorders. One, as a statistical tool. This means you get better epidemiological data about communication and swallowing disorders. And there is attention given to these silent disorders. Those people who are signing deaf persons are able to fully advocate for themselves.
Many of our patients are not able to fully advocate for themselves. They do not have a disorder that they can always draw attention to themselves, so in many ways they are not given as much attention. So we think that this could influence of in terms of grant money allocation and health resources, so that we know how many people do have these difficulties with communication.
You can count, for example, for a given insurance company, how many wheelchairs for example, they have paid for in a given year. You can do things to grossly get at some of these other things. But with communication disorders, they often are not those more measurable, easily measurable types of ways to get at how many people have this disorder.
As a research tool needed to compare outcomes of real life functioning, because it provides a common language. This could spur increased research on these real life functioning aspects, and especially improve research on the environment, since we do not cure our patients. We teach our patients to adapt, and that adaptation by definition is dependent upon the environment.
But research on the environment is a lot harder thing than research on laryngeal vowels or something like that. So of course there is a tendency to want to research things that are a little more clear cut, but this kind of classification system would encourage people to look at some of these bigger pictures, as some already have.
Also, research and development of assessment tools and procedures to evaluate the dimensions of impairment, activity, and participation. And probably the most important relationship would be the relationship between these dimensions.
As a clinical tool it is needed to spur development of clinical tools to evaluate impairment, activity, and participation needed to improve outcome measures across the rehabilitation process and quality improvement. Currently, depending on what aspect of rehabilitation, you are going to be evaluated by a different scale with different language.
It is very difficult to follow somebody from the initial time of a stroke for example, through acute care, skilled nursing facility, rehabilitation hospital, outpatient care. Common language is not used, and is difficult to track, as well as difficult clinically to compare programs in terms of how one program is doing as opposed to another. And also would influence thinking processes and clinicians to be more patient-centered and more interdisciplinary.
What I'm going to show you now are some transparencies from a program at the University of Chicago Hospital at Schwab Rehabilitation Center that I have been given permission to show. And what they did is they have adopted over the last year, ICIDH-2 as the model that their rehabilitation is going to operate around. So they changed over to the goals of trying to make it more streamlined, accessible, meet regulatory standards, clinical practice patterns, to see what they could do using this.
I should say from the outset that they are not using the actual codes. They are using it as a structure, not practicing giving the actual codes at this point, but that is what they are working towards. So their process and their rehabilitation, and again, this includes PT, LT, and speech, interdisciplinary task force, clearly defined project goals. Everyone is involved in the project, in many drafts, and many pilots.
What they have produced was instead of OT, PT, and speech having separate reports, they have produced a single report. And by looking at it in terms of activities and participation, you can do that. So for example, you might have as the goal for the patient, to be able to attend church services again. That's what they want to do. Now all therapies are gearing their individual treatment toward reaching that common goal.
So by looking at it globally, you get everybody focusing on the same thing, and each person is still doing their own individual therapy, but they are doing it towards a common goal. So they have an actual one evaluation form. There are not three separate forms. So that evaluation, patient history, functional status summary using ICIDH-2 terminology and terms, rehab potential, treatment planning, and goals are all done. All therapists sign on the bottom of the same sheet. Progress discharge and team conference is also done.
So in terms of implementation, well, there is a substantial learning curve. It took a long time. It took a lot of meetings to teach everybody this.
Now this is a very important transparency. A lot of talk has been about, well, isn't this just an extra thing that people are going to have to do. Well, if you think differently, it's not an extra thing. If you think differently, in fact sometimes you can reduce. So they went from 19 forms to 6; 50 pages of reports to 18 pages of reports. So again, sometimes you shouldn't always think of something new if it's a radical restructuring, as just an add-on, and that would somehow cripple a system.
Outcomes, single form, patient-centered, team, relevant information. Some research on percentage of staff. They were very satisfied the documentation provides meaningful information to other team members, and you can see the substantial increase. So there was increased communication between the therapies as they knew more of what each were doing.
More patient-centered goals, patient ownership. By doing the goals this way, you can't make an activity goal for somebody obviously without asking them what activity they would like to get back to. What can't they do now that they used to be able to do before? Now as individual therapists -- it's not good therapy -- you could develop a goal for somebody separate from what they were thinking. You can't do this type of goal separate from what the person misses, what the person wants in life.
Thus, that patient ownership is important, because by them being a part of it, they are not really a passive person. You are not treating them as just somebody to be done to. But they are part of the process. By being part of the process, I also believe that the patients and families have been more involved in therapy, are more compliant.
A lot of times we talk about non-compliance, and that is because we are telling somebody something that they don't really want to do. They are too polite to tell us, but that's what is really going on. This way you get them keyed in, in the beginning, and everybody knows where therapy is going.
So they have adopted this model; better consistency in clinical practice, across levels of care, better communications across levels of care. So acute, inpatient, outpatient. The University of Chicago system obviously is a large system that has different levels. So we are talking about better communication between those different levels.
So their conclusions after using this were that the investment was team education, new staff orientation, and logistics. The payoffs were quality patient care, case-by-case clinical problem solving, patient advocacy, clinical education, and an implementation of a philosophy of care, a way of thinking about things they found to be useful.
There are many other examples of people in speech pathology using the ICIDH model. This is one tied to stuttering in terms of participation. He's got a whole series of scales for each level. So how much the stuttering affects you overall. How much the reaction to your stuttering affects you overall. Other people's reaction. This is looking at that participation, that societal level that was spoken of earlier.
Also it looked at the activity level. Talk during a job interview. Answering the phone at work. Making phone calls at work. Talking with fellow employees. So those actual activities of work that involve some kind of communications. And that's for all the other areas also.
Then lastly, this can be used for social policy development, which is very much what we are talking about here. We need to demonstrate economic benefit, and to establish a moral right, and that the relationship between the two are not in conflict, because the best uses of economic resources is actually the most moral thing.
If you were to have a family, and squander the family's resources so that basic things were needed, that would be considered a bad thing. The same thing with the government. You want to spend your money in a way that gets the most effect from it.
Now that we have technology to keep people alive, we do have a moral obligation to see to the quality of their life now that we've saved them. And that's where modern medicine has increased the life span lately.
And also I believe that when the core staff is understanding of all our code of ethics, which is a charge to advocate for our patients, that the environment is the problem, then we have an obligation to fight to change it, including such environmental things as social attitudes. So it's already in all of our code of ethics that are to advocate for the patients.
If you look at it from this model, we really want to help people with communication disorders, then we need to do something about policy for communication disorders. We need to do something about people's attitudes when they talk to somebody whose speech is slurred, so they don't automatically assume that have a reduce capacity or intelligence because of that.
There are logistical and feasibility barriers, data, privacy, and confidentiality, retraining, as everybody is going to say, adjustment of philosophy. And as I said yesterday, concern for who will be assigning codes. We want to make sure that the right person that is capable of assigning the code is given it.
With infants and children the scale is current, not fundamentally developmental in nature, but could be used with further study by a task force that is currently working on it.
Then lastly, would you recommend the routine use of ICIDH to administrative health care claims in the United States? ASHA as an organization has not officially voted on it, because we haven't seen the final form of it. We fully support its philosophy and goals.
We are impressed with the change from beta-1 to beta-2 in terms of improved areas concerning communication and swallowing. Our input was taken and made substantial changes, so we trust that the changes that we are recommending, not all, but they will taken. So we are prepared to support its use for routine use on health care claims.
A little more about cost. That was alluded to already before, in the sense that in this case they were able to reduce paperwork. But I would also say what's the cost of not switching to this system?
My job I had in the early eighties, I worked for a nursing home, a large rehabilitation nation company in a nursing home. When I wrote my first report, my supervisor came in and said, well, these reports -- and I was wondering what I had done so wrong. She said, well, they are a little too clear. We have found -- this is the company speaking -- we have found that if Medicare doesn't really understand the reports, they are less likely to reject it.
And then she went on to say, we also know that all they want to look at is that chart. As long as you start off with this percentage, that the next month it's at an increased percentage, and the following month it's at a higher percentage, that's all they are going to care about anyway. So the idea is to confuse them in the report, make them just look at the chart, and you're in the money. This is what we are being told.
Now that kind of philosophy makes you do such things as have people find a set of 10 pictures, and have somebody work on naming them. They will probably, if they have a failure, be near 0 percent at the very beginning. If you keep working on those same 10 pictures, they will probably be 20 percent by the end of the month. If you keep working on it, they will probably go higher and higher in percentage.
Well, this is a system that is wasteful. This is about wasting resources. Wasting resources is wrong by having a system -- that's why time period to reimburse has an affect on behavior. We all know this. You get more of what you pay for; you get more of. So a system that had you look at actual improvement, and had to report on that, would certainly be preferable to the cost of a system that can let people get away with that type of thing.
One more thing about the report. So she told me that the reports needed a lot more jargon. So I was very angry at her. So I went back and I wrote a horrific report. It was complex sentences, passives where they should be actives. I did, because I wanted her to say, now I think you've gone too far. I was trying to make a point.
Well, what happened actually was she called me into her office again. She told me that these were great reports. And not only were these great reports, she was sending these reports to the national office, and these were going to be used as an example to teach other therapists to write reports for Medicare. So this is the guilt that I have to carry to this day. I contributed to something that I was trying to destroy by extreme.
So there is a cost right for what we are doing. And that's a bad cost. Medical care dollars are too scarce to waste on a bad process. We are replacing what we are doing, not just adding to it. So if somebody is doing a series of tasks they weren't doing, and I say now I want you to fill out this form, and they say, well, I'm already doing those things. I don't have time for this form, the recommendation is well, look at what you are doing. Maybe something you're doing isn't very efficient. Let's replace it. Let's not add to it. Let's strive to replace it.
That of course leads to a question about cost effectiveness --
DR. IEZZONI: Dr. Threats, we're going to have to hurry a little bit. I apologize. You've been great. I really apologize for that.
DR. THREATS: I just want to show you, people have been talking about reliability. In the trials that I ran, we did basic questions, ten researchers, ten clinicians, ten persons with hearing impairment filled out a very long form about basic questions on feasibility. We did 20 cases of interrater reliability, 20 cases of test/retest.
This is the form they filled out. So each person is this interrater. Now only are you rating them on the codes, and whether they agree on the qualifiers, but also on the confidence of the rating, and how confident were you. So I wanted people to see that this was indeed being done. And then we did even some more questions about the use of this case and all. So these are the actual live patients.
So I just want to end on the fact that this information will be available to you. And you will be able to look at that. The system can be learned. I think that you have to challenge people to improve, and we can do that. It's not going to be a perfect system. And we know that, because classifying people into numbers is never going to be a perfect system. And we accept that, and we're willing to accept a scale, even if it doesn't have every single thing we want on it.
Thank you.
DR. IEZZONI: Thank you. That's great. We have really benefitted from this. These examples are terrific.
Dr. Stark is going to first.
DR. STARK: I'm Susy Stark. I'm from Washington University in St. Louis. My position here today is as a representative the American Occupational Therapy Association. I'm also an educator in the OT school at Washington University. I'm a researcher. My research interests are all social participation of people with disabilities as impacted by environmental barriers. And I also have a clinical practice, so I struggle with the measurement issues every day.
Today I would like to respond to the committee's questions. First, I plan to review how the ICIDH is used by OT practitioners, review the model, and then finally, identify the strengths and limitations of implementing the model.
I would like to first show you a slide. I think it's really important to know that OT has used the ICIDH model pretty regularly. Last May, I believe, we published a practice guide for OTs in which the ICIDH was a major component of understanding the relationships between person factors, environment factors, and outcome factors. It is important to note that all of the emerging scientific models of occupational therapy recognize the interaction between person, environment, and the outcome of performance or social participation.
We address this in our uniform terminology, which is the language that occupational therapy practitioners use to communicate with each other. We are hoping that the ICIDH works as a language kind of neutralizer, to make communication between disciplines easier and more effective.
It is important to note that the essentials of all OT education also require an understanding of this interaction between person and environment, and the resulting occupational performance. I think it's also important to note that AOTA has given an official position and has participated in the ICIDH development and beta-2 trials. I think that reflects the commitment that we have to this model and our belief that its development is important.
I think there are more general applications of the ICIDH that exist in OT that I wanted to make you aware of. The ICIDH is emerging in textbooks. A couple of textbooks that were recently published last year, one on client-centered practice primary law, one entitled, Enabling Function and Well Being, which is one of the major textbooks for OT educators was published by Christianson and Baum(?).
I think we find it cited in the literature pretty often. It is also emerging as a model for new program development. Winny Dunn(?) has just recently been reviewed by her advisory council to begin a doctoral program at the University of Kansas Medical Center, and will have an OT program basically on the ICIDH model, or the notion of participation.
Our own emerging doctoral program and the program in OT is considering the use of the ICIDH model as the model for our doctoral program in disability studies. I think you will find that there are many emerging and interesting practice models in OT that use this concept. We are moving into the community as a general practice. OTs tend to be less medically-based, less hospital-based, and you will find that this model is guiding our practice.
As educators I think it's also important. Interestingly, we were talking about educating physicians. I taught what classically was known as the pathophysiology course for OTs. And I have turned it from a kind of basic diagnostic model to an ICIDH model that explores the impact of impairments on activities, and finally on participation, which I think is very important. It's been very successful, and gives our students a great understanding of the impact that disability has on performance.
I think that this model has great value to our profession. The classification scheme addressed the outcome measure of individual performance across disciplines. This hasn't been done before in our medical system. It is also a model which explicitly recognizes the contribution of the environment of the performance of people who have disabilities. And it is the classification scheme, as I mentioned that we hope will eliminate language barriers between professions.
As the pressure for accountability increases, and the evidence of clinical effectiveness becomes a requirement for reimbursement, this model may become a clear measure to determine if the appropriate clinical outcomes have been achieved. This model would help payers recognize the performance gains in therapy. No longer would range of motion be the only measurable outcome that we have.
It may assist in guiding assessment. There is a direct and logical connection between what is assessed by therapists, and what is expected as an outcome by consumers and their families.
Deterministic models of disability that only recognize the contribution of the person to their performance have been criticized in the literature, and this model considered the interaction effect of the environment with the characteristics of the person as predicting performance.
Clearly, this model more closely reflects the reality faced by people with disabilities across cultures and social systems, and assists professionals in describing and justifying the treatment strategy that they build to support their client performance.
That being said, there are some current limitations for the model for our profession. We have been participating in the beta testing, so we understand that the model isn't ready for full blown use at this point. We found that there are some problems with the categories.
What we did is we had a consensus conference within our group to answer some of the basic questions about the ICIDH-2. We had special interest group leaders that reflected participation across different practice areas including mental health, folks that work with children. They received Web-based training, and we had a teleconference, and responded to the basic questions.
You probably have been introduced to those, but briefly we wanted to explore which model we thought would be most useful visually, and the questions that related to categories and operational definitions. Essentially, we found that we preferred a transactional model, because it was compatible with our scientific model.
We thought that the ICIDH-2 was open to great interpretation at this point, because of the operational definitions. And we hope that categories would soon be better defined. We thought that there needed to be better distinction between activity and participation, and activity and body function; some of the comments that have been echoed by other speakers.
But I think that these are really findings that can be expected in a new model. We have to remember that ICD is on its tenth version, and we understand and we support that this is a developing model, but we think it's worth exploration.
We think that the strength of this model is that there is explicit recognition of the environment to performance. We think that use of participation is an outcome, a meaning individual outcome. It is critical. We think it's an appropriate way to capture functional status information.
Certainly we are collecting a lot of body function level and some activity understanding. But I think the understanding of participation environmental contributions to disability will be not only valuable, but certainly critical in understanding how we can affect health of people with disabilities.
We think this is going to be a great tool to track performance consistently across the continuum of care, as we move from acute care to more community-based care as our population grays, and as technology keeps people alive longer and longer, I think this is going to be critical.
I found in my clinical practice, which is mostly community-based, FIM scores become sort of meaningless. When I met my favorite client, Ray, who had been living in his bed for five years as person who had post-polio syndrome, it occurred to me that he had a pretty low FIM score. But the problem wasn't the services that I could offer. The problem was the fact that he had an ambulance come and take him to a psychiatric appointment once a month, because he couldn't get out of his bed or down his steps.
The problem was he didn't have the appropriate equipment or environmental supports to help him do the things that he needed to be able to do. And after three years of intervention wasn't funded by any insurance company or third party payer, we finally got in place all the things that Ray needed in order to participate. His goals were to be able to go the place that he heard about, but had never seen, where you actually drink a cup coffee and read books at the same time.
When I told him about Barnes & Noble, he did actually get out of the house. We had a community agency build a ramp. We had a lot of donated equipment. But unfortunately, Ray passed away two months after his ramp was installed. So the man who waited for seven years to read a book and drink a cup of coffee, wasn't able to participate in those activities long enough. So I think this measure is going to actually capture and help us understand the things that need to go in place to help someone participate.
I would suggest the most important aspects to include would be as reflecting Paul's comments yesterday, the participation and environmental levels. Those are the things that are missing in our medical system, that we need to understand.
I do think there are some liabilities of using this model that have been mentioned. Our members didn't find that it was appropriate for children or people who rely on caregivers. I think it gets at Nora Wells' concern about reflecting and understanding that family interdependence that needs to happen for participation for people that have disabilities.
We need to remember that it's not a classification. I have run into folks who are using it as an evaluation tool, and I think that's a danger and a risk we need to be very clear about. We don't have reliability and validity information yet. It is forthcoming. I think that's important to remember. I do think it's a very broad tool. And I think a team approach would probably be the best way to address these issues.
Again, operational definitions within major categories need to be addressed. Privacy is major concern. Ray Seltser's idea yesterday was pretty intriguing and sounded very interesting, the idea of having an annual evaluation. My concern as I thought about that was what if someone is stuck with a bad rating given a rater who doesn't understand fully that person's abilities? For example, what if someone is rated as not having money skills, or not having good decision-making skills? Will this information follow them? Is any piece of medical information in this society today, what if your employer gets his hands on it, and doesn't give you a job based on somebody's rater?
We also need to recognize that people are static, and the environment is static. Thus, participation will change over time, as abilities change, and as environment barriers -- or supports get put in place, participation will change. And I think we are going to need to have a system to recognize that.
AOTA is prepared to include the model in practice guidelines. We are working to use it to support our understanding of disability. We are in favor of working with PT and speech and other disciplines to develop this model more fully, and to support rehabilitation and the outcome measurement goals.
I do think that specialized instruments may need to be developed to serve various disciplines. We are currently working on those at Washington University. We do believe this is the best available model as it recognizes again, participation and environment. And we do believe that it will be very exciting to see what occurs as a result of the beta testing, and the changes that are made in order to reflect the needs that our members feel need to happen.
Thank you.
DR. IEZZONI: Thank you, Dr. Stark. Dr. Guccione.
DR. GUCCIONE: Thank you. I'm here on behalf of the American Physical Therapy Association, the major professional organization of physical therapists, physical therapists' assistants, and students. I actually asked to go last, because we were last informed of this event, and actually now in the interest of time I will skip all the comments that are in concert with my colleagues, and concentrate primarily on the negative or the disagreements which you might find within the profession.
The American Physical Therapy Association does not have a stand on the ICIDH-2, nor is it likely that it will ever take one, the reason being we have already well adopted and integrated into practice, research, and education, down to the level of accreditation, the NAGI(?) model, the Institute of Medicine model. And that is primarily based upon the inadequacy of the ICIDH in the early nineties.
The ICIDH-2 might have offered a better option at the time some of these changes were taking place, but at the time that we began development of The Guide to Physical Therapist Practice, published in 1997, which is 400-odd page outline of body of knowledge in physical therapy, the best available model was in fact the NAGI, IOM, NCMRR variations on the same theme.
Now in concert with my colleagues, in terms of an overall structure of four different baskets, or in this case now three different baskets in which to sort different kinds of clinical phenomenon, any system is probably better than no system at all.
And I think that it would be one of my concerns that the research that Dr. Threats reported, the assumption is made that the introduction of the ICIDH set of baskets or sorting terminology and sharing the language is somehow better than others. When in fact I think the comparison group as described, is no set of common terms at all. I think we must be particularly careful in that regard.
One, I have heard a lot this morning about the reliability issues relative the ICIDH-2, but one needs to go actually to question the validity in some cases. And we would have concerns about the face validity of the model, particularly in the area of participation.
What we note, and this has been commented in the only time were asked as an organization to comment officially on the ICIDH-2, and we did so directly to Geneva about a year and a half ago, although individuals and groups of individuals who are physical therapists may be participating in the beta demonstration projects, there has been no official participation of physical therapists in the development of this language.
Our concerns about participation are at the face validity level in how it mixes the environment, which actually has an impact on activities. In using the old draft, I could point out to you, but maybe in the interest of time we will just talk offline, where there is lack of specificity about the environmental impact on the activity.
The reason why the American Physical Therapy Association did gravity towards the NAGI/IOM model was because of the notion of disability. Now the term "disability" may be an infelicitous word for good reason, and one which we may not want to use in the future, although it is still used by many, many groups, even those who would self-describe as disabled.
Inherent in the NAGI/IOM model is the notion of role, and role disability really was the essence of the disability concept within that model. The embedded concept was pretty much unexplicated until Disability in America was published, with the notion of role set. And that is a notion which has been lost in the ICIDH-2. One moves immediately from activities, tasks, functions, that one cannot do, to lack of participation at a societal level.
And I was intrigued that Dr. Ustun this morning kept responding that problems of participation need to be responded to at a societal level. That ignores the level at which problems of participation, formerly described as disability, need to be addressed by myself and my colleagues in occupational therapy or speech therapy.
So we still have a problem with the model at the level of face validity. The concept of participation is too deep, and doesn't adequately distinguish between those restrictions to participation which might be handled at the individual level, which according to the model then become activities. But the roles are not described until you get up into the participation basket.
And then the participation barriers, which obviously must be addressed at the societal level, and which many of our organizations, as well as individual therapists participate in changing societal restrictions on the rights of people with differences.
So we do have a validity problem in terms of adaptation, besides the fact that the horse is well out of the barn in terms of defining everything up to and including the level of our house of delegates. All of our positions and policies not only are written in NAGI/IOM language, but are mandated to be written in that language. So it would be a huge political internal change of our educational structures, research structures.
We just recently published a clinical research agenda which also adopted elements using the terminology derived from the framework of the NAGI model, which did not commit us to the specific vocabulary. Now at the level of vocabulary the ICIDH-2 is very exciting. It is always nice to have somebody else figure out all of the different ways of coding something. And in terms of providing a crosswalk to charts, to the development of computerized documentation systems, a lot of good work is probably here.
And certainly I know from the informatics world, the idea of having a unique identifier for every term that might appear in a medical chart, it would seem to me that the ICIDH-2 would potentially be a great resource for such crosswalking.
However, at the level at which we begin to think about reimbursement, particularly in light of the face validity problem, we often are able to reduce what was in one model known as disability by individual solutions. It is not clear how we would begin to document, which is being projected as having a societal solution.
It is very distressing to me, having worked pretty closely with colleagues from AOTA on ASHA on such instruments as the MDS, the MDS-PAC, and the OASIS, that these other instruments which are in currently use, and in many ways have had a rather difficult introduction into the clinic, that the other systems which are currently out there mandated by the insurance world, both either private insurance, but mostly particularly government insurance, it is not clear at all to me how these will interface with the systems that are currently out there. But that may be a question that needs to be addressed.
In terms of barriers to implementation, I can tell you that we are currently -- in fact, I just left a group this group working on the second edition of our Guide to Physical Therapist Practice. Because of the ICIDH-2 and its apparent appeal to other professions, we will contain somewhere in the neighborhood of two to three paragraphs about how to crosswalk between the systems if in fact you are in a health care system, or facility, or an environment which chooses to use that model.
However, we are deeply embedded in the realities of reimbursement. And that is the realities of the HCFA 1500, the realities of insurance forms. We have created our own template for documentation that has already been published in draft form, and will be published again. It is model neutral. It would be quite possible to be a NAGI-ite, or an ICIDH-2er, and use our approach to examination and documentation of intervention.
But the fact of the matter is that much of this process may be irrelevant if you do not suddenly hook up with the insurance community and with your own governmental community that is developing instruments. Yesterday morning I participated with a representative from AOTA as well as others, several physicians and nurses representing other groups on the validation study of the MDS-PAC, which will become a widespread instrument for the documentation across the post-acute, chronic life of a patient or former patient.
This isn't in there. So where is it going to go? Now could you crosswalk it? Yes, absolutely it could be crosswalked, and there might need to be major changes in terminology. But that kind of effort has not been mounted. And with the mandate for use of the MDS, for those of you who are not familiar with that, unless you fill out the minutes and fill them out properly, you don't get paid.
And the accuracy with which you fill them out not only determine whether you will get paid, they determine how much you will get paid. And not surprisingly, you don't get to provide services that you don't get paid for.
So the impact of introducing a language as an intellectual enterprise for research purposes, versus the clinical rubber hits the road of filling out a form, and then being qualified into a certain category, receiving a certain quantification, and then being reimbursed accordingly, whether that be as an individual or as facility is quite real.
And that is a barrier to implementation of any system, because any system which does not facilitate the process will meet with widespread objections. While the professionals might not be able to inhibit the widespread introduction of the terminology of the ICIDH-2, I can assure you that the health care industry can nearly put to a halt anything with which it has not been included.
And that seems to be a political aspect of this process, the representation of those people who do actually create these systems of quantification and payment, that may not have been adequately considered for the introduction of ICIDH-2 or any system into this particular American health care system.
Finally, just one brief word. One should not presume that in the interest of communication across disciplines, the disciplines will not continue to develop their own specific languages. I left up my colleague from OTs breakdown of some very distinct OT terminology and concepts, which having been fortunate over 22 years of clinical practice, I think I have had some extraordinary OT colleagues, as well as some extraordinary speech colleagues, but I am less familiar with their theoretical structures than OT.
How an OT thinks makes a whole lot of sense to me in terms of what they do. And in terms of what a classification scheme has to do for a professional who is collecting, sorting, evaluating data, and then implementing an intervention that is designed to achieve an outcome. I could not think like an OT, because I don't provide the interventions that an OT provides.
Similarly, PT and RS have some similarities, but some distinct differences. We don't have a theory of occupation. We don't identify purposeful activity as our major therapeutic modality. Those two facts alone have forced physical therapists to develop their own way of thinking about things.
The real barrier to getting the ICIDH grammar, not its vocabulary, but its syntax and its thought processes into the minds of physical therapists will be dependent upon the way in which it improves the classification abilities of a physical therapist to do the job that a physical therapist is hired to do, or similarly as an occupational therapist is hired to do.
At the basket level, the broad framework, and let's not get too discrete about what may be right or not exactly right about your individual concepts of yes, there are bodily functions, activities. And then there is participation. And I would break that participation; that can be individually affected, versus participation that must be societally affected. Those are helpful frameworks.
And the degree to which the ICIDH-2 provides a vocabulary that can be used within any of those frameworks is fine. But the way in which it enhances a physical therapist's ability to classify your phenomenon, your clinical data in a way that will make me come up with the most cost effective, efficacious solution, there is a large gap there. And I think that that must be considered as well.
We may use your words, but whether we will have ever gravitated towards the ICIDH-2 as a way of exposing what we do to ourselves to make us more self-reflective practitioners, that still remains to be seen. And that is the ultimate barrier at to whether it is truly adopted in clinical practice not, as a practice tool, as opposed to a language for quantification and documentation.
It's all in there.
DR. IEZZONI: Thank you for your clear and forceful presentation. It was very useful and very thought provoking.
I'm looking at my watch. What I would like to do, because there is a logical kind of pace to these kind of presentations, it's kind of break and then bring people back. I would like to maybe have some questions until 11:30 a.m., and take a break from 11:30 a.m. until 11:45 a.m. And then if there any remaining questions, have some time for those, but then start the next panel at 12:00 p.m., and end by 1:00 p.m., which is the end time for the meeting.
So are there any specific clarifying questions, because maybe we can deal with those right now.
DR. PLACEK: For Chris McLaughlin, this is an American Psychological Association project of WHO. Is there such a thing as an international psychological association also? And if there is, are they going to be involved? The other thing you said was that the tool you were going to develop, the fours tools I guess from eight chapters are going to be useful for a variety of disciplines. I wondered if you have planned to work with other professions in developing these tools?
MR. MC LAUGHLIN: The APA does have international members. There are psychological associations in other countries. There is no one, singular international association of psychologists, but we do have quite a few international members.
So in developing an assessment manual, it would have to be an in iterative process, and these are some of the things that we are still in discussions with WHO about, and also are determining for ourselves. Initially, we had thought it might be for the North American market. But that still has not essentially been determined. If it were to be for an international market, we would include international providers, and that entail a much more involved, much more complicated process for developing a manual. That is still to be determined.
The manual would be just one single manual. The goal would be to make this useful for all professionals. So as a result of that, we would include other disciplines in the development of the manual. What I was describing was really sort of two things earlier. Initially, we think that there is a way to simplify ICIDH for use on an administrative claim form. That would be a distillation of the system to the most clinically relevant areas that could be sort of the first step at getting the system integrated in the U.S. health care system.
That's the high level system that I proposed. The assessment manual is what is going to be required for the finer grained use of the system, because those validity and reliability issues need to be addressed. And there needs to be more of a common set of definitions to those.
DR. IEZZONI: Andy, you had a question.
DR. KRAMER: I would just like to address it to the entire panel, except in your case, because it comes from the work you just raised. And that is for the issue of reimbursement, which I think was very well raised here, and has always been a concern of mine in this regard, everybody is going to be faced with the need to provide this information in a way that it could ultimately be used for reimbursement purposes.
Even if it didn't start out being used for reimbursement purposes, everybody talks about that being one of the uses. And if it is going to show up on a claims types form, it certainly is ultimately going to be one of the uses.
Do any of the rest of you, occupational therapy, speech therapy, psychology association, do you have concerns about how it might dovetail with other issues like other forms or strategies that are used for reimbursement at this point?
DR. STARK: OTs work in the area of activity, and we are reimbursed in the area of activity right now. So I am concerned that the structures dovetail. I am concerned that there is not an overwhelming amount of paperwork; that we add on layers. I am interested in looking at how we can remove some things, and be able to identify two other of the very important things I basically feel need to be involved in reimbursement.
And that is how can we get reimbursed for environmental modification? Because we are not now. And would this structure in some way allow payment systems to support people getting in and out of their homes? And the other area is we can then write goals, and have treatment activities that support participation. And some people feel that's kind of a vague category right now.
That's the area that I think we are moving as a profession to addressing as OTs. We have this activity domain under our belts, but I think that there is always a push to understand so what does it mean in terms of outcomes and quality of life and health for the individual? And that's why I think our profession is interested in pushing the envelope to get those things quantified.
DR. THREATS: Whenever you are talking about reimbursement and insurance companies, they can always take the good and turn it toward the evil. We wouldn't like to see a code for swallowing for example, say well, that's $50 for that code. That's something that people have tried, attach a dollar amount to a particular thing. So that would be a possibility. That's our job to battle, however. That would be with any system where they try to limit you. And it's up to you to have the research, how much time does it take to do this or that, and how much would that cost be?
Blue Cross is looking at it, and they are willing to work with us. The have stated that. It has been shown to some insurance companies. And of course their idea is they really do want to say for a given code, this is how much money. Then they want to have that as a system.
MR. MC LAUGHLIN: We are clearly very interested in reimbursement issues. Part of what we think is so useful about the system is that it is etiologically neutral, because at least in clinical settings, psychologists often find themselves providing services to make clients more functionally -- to raise their functional status, but yet these services do not necessarily congregate around an identifiable diagnosis.
So it's been an issue that has been of concern to psychologists for a long time, is trying to figure out some different way of classifying functional status, and encouraging people to reimburse for services based on that classification. We think that ICIDH may provide a step in moving in that direction.
We are also concerned that the health care payer community is not particularly connected to ICIDH. In this assessment manual that I have discussed, it would be a major, major investment on the part of APA to develop. APA has a very major publishing house that would be willing to do such a thing if we can show that we can at least break even or make some sort of profit on developing this thing.
That's one of the issues that we have to determine before going into it. We are pursuing and considering developing some stakeholder meetings to try to see just where insurance companies, HCFA, some of the other providers are on this issue, and see if they are interested in further development of the system. We think that in the long run a reimbursement based on functional status may be more cost effective than reimbursement based simply on a diagnosis. So we think there is a great opportunity here. It's a long-term opportunity, and it's going to take a change in the way people think about it.
DR. IEZZONI: Okay, thanks. Quickly, and then we'll take a break.
DR. GUCCIONE: If I can just add, because maybe physical therapy is in the most advantageous position, because obviously a great deal of what we deal with is at the impairment activity level. Obviously, all three of us as therapies, the three therapy groups, rehabilitation therapy groups are clearly organized around functional evaluation and functional status oriented diagnoses.
But I would still hold out, and I can say this as an association person, formerly a clinician who probably deals as much with insurers and governmental, voluntary credentialing, regulatory agencies, perhaps more than my colleagues on a daily basis, I would not be really enthusiastic about going forward with the definition of participation as it currently is defined and identified.
I totally appreciate the occupational therapy concern that at long last contextual factors and environmental modifications may be coming to the fore. But I would be very concerned that they are in the wrong place, because the participation definitions I do not think are going to mesh easily or well with current modes of thought in the insurance, private or governmental. And that that needs to be sorted out, because many of your contextual, architectural changes, environmental task modifications are at the level of activities, not as defined here, at the level of participation.
And that's merely a definitional what's in which basket problem, but it's a scary one. And I'm glad, I think, that I don't have to go to an insurer to explain the wonderful notion of participation as defined here, and what it is going to do to the health care system, and why you should pay for it. That's just a practical association hack response.
DR. IEZZONI: Okay, quickly.
MR. MC LAUGHLIN: I would just like to respond to that. We were concerned about that too, and that's our first cut at trying to figure out a way to make this functional, we were focusing on activities as the more similarly clinically related activities to what is currently being reimbursed.
DR. IEZZONI: I wish we had more time, because I think that there is a lot more to say. I think what we will do is take a 15 minute break, and come back and start with the next panel. We have run over.
[Brief recess.]
DR. RUGGIERI: My position is supported by a grant that is shared by two institutions. The section of medical information resources, the health sciences research section at Mayo is where my mess is. That's where my desk is. But I also have support from the Division of Health Computer Sciences in the Department of Laboratory Medicine and Pathology at the University of Minnesota.
I can't help but note that this meeting is a tremendous coincidence of Minnesota. Yesterday I met Judy, and it turns out she lives in the town right near mine. And here we are at the Hubert H. Humphrey Building. If any of you didn't know, Gloriajean's tropical landscape, that was taken in Minnesota. I guarantee you my assertions won't be as reckless as that.
But anyway, I'm happy and grateful to have the opportunity to offer you mind and my colleagues' perspectives on WHO ICIDH-2, particularly in the areas that comprise the agenda today. I'm a newcomer to the ICIDH-2 discussions, and I'm excited to join the forum and learn more, and hopefully make a contribution.
I would like to thank Paul Placek for helping me get up to speed. He was very responsive when I first began to show interest on this.
My general area of interest is medical concept representation, particularly as it relates to the computerized medical record. The section of which I'm a part at Mayo is involved in the development and evaluation of standardized, controlled medical terminologies, classifications, and nomenclatures, and the provision of terminology services to support medical concept representation in routine clinical practice, ultimately for storage in the automated medical record.
My own special interest, I'm a rheumatologist, and I have 20 years of practice experience, my main interest is the modeling and representation of functional status. I think it would be useful to just as an aside, make a couple of points that I think we're all probably aware of, but would be useful to remember, particularly in discussions about ICIDH with other colleagues in our field.
I am a fan of functional status. I know how important it is. And I think everybody else here realizes that as well. I think we are all joined at the hip on that issue. I think criticisms or praises of ICIDH have to be taken separately from that fundamental standing. Just because ICIDH may not serve a particular use case within the functional status domain, or you particularly don't accept ICIDH as serving a particular use case within a functional status domain does not mean that you do not recognize the importance of functional status concepts.
And also just because ICIDH may not be the appropriate classification for a proposed functional status data field in a HCFA 1500 form, does not necessarily mean that the classification might have useful applications elsewhere. So with that, let me proceed.
The agenda issues as were communicated to from Dr. Queen's e-mail had to do with a data field for functional status data within meta data registries serving administrative health care purposes. That's a fancy term for a health claims form. And also the second issue is ICIDH-2 adequate for meeting the representation needs of that particular field?
Fundamental to addressing these issues is a clear understanding of the requirements of the use cases, the scenarios, the processes, the decision activities that this data field would support. Administrative health care data naturally serves administrative health care purposes. While there exists much health care data out there, much of it is not realistically accessible, and we often by necessity must rely on data resources which are practically available, such as health care claims data.
Nevertheless, the most convenient source or vehicle of data is not necessarily the most appropriate resting place for it. It is important to ask are there administrative activities or use cases which would use this data field? If so, what are their information requirements? Or is the administrative health care meta data registry serving as a convenient vehicle for the data to be used in a yet to be defined process, or to serve an incompletely elaborated use?
I would imagine that the creation of a new data field, regardless of the extent of the definition of its use would impose a not insignificant burden on those who would collect it, store it, as well as those would have to record and submit it. Will those burdens be worth it, at least for the known use cases that we have for this type of data today? It is possible to expend much energy and resources in creating a data field and filling it, and realize it ultimately did not support the goals we had initially intended for it.
If a functional status data field were found to be justified and implemented within administrative meta data registries, the structure and content of the data field should reflect the level of detail, or the level of data granularity of the specified use cases. Granularity requirements of the data uses must be understood up front, and accommodated accordingly.
Decision on granularity of data use requirements bears directly on the suitability of a particular off-the-shelf data representation scheme, classification, or terminology used to fill the data field. Will the data uses be at the level of individual patient, or for populations, or for both? Will the data uses involve administrative decisions, resource allocation, knowledge discovery, diagnostic or therapeutic-related medical decisions, advocacy?
Finer data granularity requirements may be necessary for data uses around the individuals and populations, and may vary for each according to the specific use. For example, if the change of 10 percent in functional status is used as a criteria to decide on whether a therapeutic effect has occurred, or an intervention is indicated for an individual patient, and the level of detail of the data representation must be at this level of incremental change, if the data representation scheme can only represent at a coarser or lumpier granularity, it may not be appropriate for the purposes to be served by the functional status data fields.
Just to summarize the points I have made so far, and to emphasize, ideal circumstances would have us possess a clear understanding as much as possible of the use case scenarios and data requirements that would supported by this data field. In particular, the representation scheme selected for the data field should be compatible with the level of detail or granularity required by the use case processes.
If we were to go through the hurdles and the obstacles to have such a data field established and filled on an administrative health care form, then we would want the effort to pay off by knowing that the data structure and representation fulfilled all the needs we had identified for the functional status data fields.
In particular, it's important to know that the granularity or level of detail is not overly fine for some uses, and not overly coarse for others. Just because a classification exists and is coded, does not mean it is capable to serve the data representation needs of a particular use case. With standard terminology classifications, it is more than just a case of if you build it, they will come.
A representational scheme, whether it is a classification, a terminology, or a vocabulary must support the actors in the functional status arena, whether they be patients reporting on their needs, a health professional formulating a treatment or therapy decision, an epidemiologist tracking the functional status changes of certain populations, or an advocacy leader testifying to Congress.
I would like to switch my focus now to ICIDH in particular, and I want to emphasize a few points about ICIDH, reflecting my personal feelings about it. I believe that the emergence of ICIDH is an important event for several reasons. Its emergence offers an opportunity for achieving conceptual clarity within the functional status domain, a landscape that has been not all too clear in the past.
ICIDH's most powerful feature, I believe, is its embodiment of a tacit -- I am emphasizing the word "tacit" -- yet convergent conceptual model of human function developed through consensus among broad based domain expertise. I believe it is a very significant landmark in an effort to conceptually represent knowledge and concepts of functioning.
ICIDH describes itself in many ways. It describes itself as a multi-purpose classification, a common language for describing functional states, a description of situations with regard to human functioning and disability, a unified and standard language and framework for the description of human functioning and disability.
And I think this is an important clarification. A classification describes things or states around common features or attributes, and as a result, can only represent the members of the classification at the level of those aggregating features. A language on the other hand, serves to represent all the concepts within a domain, with the capability to represent complex compositions and states of those concepts.
Its compositional capability or what is referred to as post-coordination of expressions, confers representational flexibility, extensibility, and representational power to the language. For instance, one can have atomic terms representing the concepts penicillin, prescription, and allergy, and the post-coordinated concepts of penicillin prescription and penicillin allergy.
Or in a more functional status oriented example, terms for the functional status atomic concept of the act of walking and another term for the concept assistive device can be used to represent the compositional concept of walking with assistive device.
These post-coordinated expressions allow representation at different levels of detail or granularity, depending on the representational or expressive need specified by the use case. It is very difficult for a classification, any classification of functioning and disability that partitions and separates conceptual entities along aggregations rules and common attributes to be a classification, and at the same time a language.
Is ICIDH-2 attempting to be both a classification and a language? Are we expecting it to be a language? Much of the concern expressed by various representatives from the different use case domains that we heard about yesterday I think were expressing this reaction, is that ICIDH is possibly being promulgated as a language. And the first reaction is, is this language going to support the needs that I am particularly serving in a functional status domain?
ICIDH-2 and its documentation states that a classification should be clear about what it classifies, its universe, its scope, its unit of classification, it organization, and how these elements are structured in terms of their relationship to each other. I think we have to hold ICIDH to its own bar.
I believe it is fair to say that ICIDH-2 as a classification, seeks in some way, in some level of granularity to provide representation and describe and communicate information about functioning in the context of health-related conditions, but only at its predefined level. This is certainly a noble effort of considerable need and merit, however, as a classification ICIDH-2 will likely face constraints in its representational capability and breadth, particularly if we expect it to be a language.
These constraints need to be carefully considered before selecting ICIDH to serve representation needs. It is important to critically assess ICIDH to these representative capabilities and limitations for uses we have in mind for it.
There are several representational issues of concern that exist that might preclude its blanket acceptance as the representational scheme of choice for a functional status data in an administrative health care form. ICIDH-2 has yet to express in a formal, explicit manner the conceptual model of functional status on which it is based.
While based on a shared, agreed, and I think in many ways elegant mental model of the functional status domain conceived by many domain experts working in collaboration, this tacit model has not been subjected the rigor of a formal modeling effort, using an explicit and exact modeling notion.
Examples of such a modeling notion language is UML, standing for the Unified Modeling Language notion. I mention this, because UML is the modeling language of choice of Health Level-7, which is the health information standard setting body.
The benefits of such a modeling effort include the elimination of hidden concept ambiguities, inconsistencies, promotion of widespread understanding and acceptance, and providing the opportunity to build interfaces and mappings to existing concepts within functional status information structures and data, as well as other models that exist about functional status.
Examples of such other content of functional status data would be SF-36, and the HAC, often referred to as the health assessment questionnaire. Such a model would be useful in guiding iterative improvements in modification of ICIDH.
Many ambiguities seem to exist in ICIDH-2, among its classes and relationships. Three people in my department, which included a nurse, health services research scientist, a computer programmer data modeler, and a physician carefully reviewed ICIDH-2 and its documentation, and all could not agree on its complete conceptual intent.
This slide lists some of the ambiguities that have arisen from joint efforts in our group to model ICIDH-2 in Unified Modeling Language notion. I won't go through them, but I will be happy to discuss them later.
We believe that definition of a formal and explicit data model for ICIDH-2 can resolve potential ambiguities and inconsistencies in conceptual definitions. The creation of such a model should precede formal adoption of ICIDH-2 to represent a functional status data field in administrative meta data registries.
I'm likely naive to many of the use cases planned or anticipated for the functional status field by NCVHS, however, there are many uses that I can conceive as a clinician that seem to demand a rovering granularity that ICIDH-2 have difficulty consistently matching. For some uses, ICIDH seems too coarse; for others it seems too fine.
ICIDH does not provide a hierarchy of global functional status, nor does it address composition of one from the functional status dimensions. Global functional status representations may convey information of clinical value in some clinical domains. Certainly, I can speak for the field of rheumatology, and global functional status assessments do have power.
ICIDH-2 cannot represent an activity in the context of a participation, which might be desirable and meaningful clinically. The coding structure cannot permit a promotion of an activity into participation, which evolving understanding of the functional domain may require. For other use perspectives, ICIDH-2 can overload a data field's function by providing too many values at too fine a level of detail. Which dimension would be selected for a presumably single data field in a health care claims data form?
How would one decide on a domain within a dimension, and at what sublevels? The situation might be akin to confronting a salad bar with multiple appetizing choices, but with only a small plate available to fill. Different uses might call for different data views, containing different ICIDH-2 elements. Would the data field begin to creep and expand to multiple fields?
ICIDH-2's uniform qualifier may be too coarse and arbitrary. Are mild, moderate, severe, and complete and their assigned percent deficits adequately expressive enough and appropriate for all uses?
Among the criteria that should be met prior to adoption or selection of ICIDH-2 to represent a functional status data field is insurance of the stability and permanence of the classification over time. Are the developers of the classification committed to proper maintenance and versioning of the classification effort? Are there guarantees that there will not be deletions, reassignment, or modification of leading codes over time? This has occurred with ICD-9.
There should be a standard protocol that will guarantee concept permanence by means of formal retirement and remapping of codes. ICIDH-2 should also have mechanisms in place for incorporation of new functional status concepts easily. The code hierarchy currently limits expansion in certain sublevels to only 10 concept classes. Finally, WHO unfortunately has carried on the ICD tradition of including the unspecified concept classes, which can foster ambiguity and confusion.
ICIDH-2 defines it scope to be human functioning in the context of health conditions. I believe its attempt to classify environment factors such as social, political, economic, and civil entities could be regarded as an extension beyond its stated scope. It is not clear what the individual health-related uses of such a classification would be.
The ICIDH-2 environmental axis classification invites etiological conclusions about environmental entities and their impact on individual function which might viewed as controversial. This may lead to wholesale rejection of ICIDH-2 in its entirety by some, despite some of its more useful within scope features. Perhaps only selective parts of ICIDH-2 should be considered for representation of a functional status data field.
Recommendations. We would hope the following recommendations regarding ICIDH-2 might be constructive. The concepts, class definitions, and their relationships intended by ICIDH-2 should be formally and explicitly expressed in a widely accepted, standardized modeling notion language, and I have suggested the Unified Modeling Language.
Such an explicit and rule-based model would allow broad agreement and understanding of classification across disciplines and uses. A model will serve as a blueprint, which can enable ICIDH to relate to other representations of functional status present in the health care record. These would include the SF-36, the HAC, another other validated and tested functional status information gathering instruments.
Information use case requirements to be supported by ICIDH-2 should be elaborated for various entities with interests in the functional status domain, with particular attention to granularity needs. Information use case needs, which a functional status data field is intended to support, should be delineated across entities with interest in the functional status domain, with particular attention to granularity need.
The acceptability of ICIDH-2 should be based on how well it enables the roles within the identified use cases. It will through satisfaction of the use case requirement, and empowering actors in their roles on the health care stage that ICIDH-2 will catch the conscience of its users.
There is a temptation always to use a classification simply because it is there, and is coded. If the classification does not support the intended uses, then another representational scheme should be considered, or only useful in pertinent parts of the classification selected for use. A representation scheme should not define the data uses, rather the data uses should define the representation scheme selected.
Scientific evaluation and testing should be performed and reviewed to determine how well ICIDH-2 provides content coverage across functional status related domains before it is adopted to represent a broad base health care claims data field. Solicitation of comments and study by the different medical and social science domains with interest in the functional status should be sought.
Infrastructure, tools, resources, mechanisms, process organization should be devoted to maintenance of ICIDH-2 to assure its permanence and responsiveness to change in evolving functional status domain need. If ICIDH is used to represent health care information in the form of functional status data over time, then we must be assured of concept of and class permanence over time.
Let me try to summarize some of the things I have said. If I have to officially come on one side of the fence or the other, I would say that I would not recommend incorporation of ICIDH-2 into a functional status data field on health care claims forms. This does not take away from other valuable potential uses ICIDH-2 might have.
I think ICIDH-2 holds a promising conceptual core around which standardized representation of the functional status information domain can crystalize and evolve. ICIDH-2's suitability for representation of functional status will depend on designated uses, and will hinge on granularity issues. Every effort should be made to elucidate those uses and their data needs. Studies should be done to evaluate how well ICIDH-2 serves informational needs in various clinical settings, and especially domains holding particular interest in functional status.
U.S. collaborators involved in ICIDH-2 development should help to focus efforts in needed areas for iterative improvement of ICIDH. These would include formal model elaboration, as I had mentioned, content coverage evaluation in various health-related settings, convergence with other information standard development efforts. I sort of alluded to this in my question to Dr. Ustun this morning.
There is a convergence right now of health care information standards, and it's sort of a coming out party, because of the introduction of the whole concept of automating medical records. There are a lot of well dressed players at this party. And ICIDH needs to make sure that it's dressed in proper attire.
In order to do that, it needs to be canvassed through the appropriate health information standard setting bodies. There are health care information standard bodies at the international and national level. Some of these would include HL-7, ISO TC-215, which stands for International Standards Organization, Technical Committee 215, which deals specifically with international health information standards. And lastly, ANSI, American National Standards Institute, the Health Information Standards Board.
And I mentioned the concerns this morning about HIPAA. I am not an expert on HIPAA, but sometimes I wonder if all of our discussions are going to be moot if ICIDH-2 has not been officially recognized as a standard by some body.
Finally, we should be sure if we use the classification, that the appropriate infrastructure is in place for its proper maintenance.
DR. IEZZONI: Alex, can you wrap up quickly?
DR. RUGGIERI: I just want to mention that there are guiding principles for promoting patient medical information standards that were communicating through Dr. Lumpkin to Sec. Shalala on behalf of NCVHS. All of my recommendations are congruent with these, and I have listed: meet the data needs of the health community; be supported by some sort of accreditation body; have timely developmental testing, implementation, and updating; be precise and unambiguous; and incorporate flexibility to more easily adapt to changes.
I'll close with this quote. Terminology is important. It was recognized as such as far back as 1839, and it is an important issue to settle, however, we have to make sure we understand the uses that we are intending for the terminology before we make a commitment.
Thank you.
DR. IEZZONI: Thank you. That was a very clear and helpful presentation, especially with the standards things. I was looking at Marjorie, because she knows what all these numbers mean, and some of us are not as clued into it.
Are there any quick questions -- very quick questions -- on clarification of the presentation? Okay, then you stick with us, and Dr. Andresen is our final speaking. Could you actually introduce yourself briefly? You weren't here when we first kind of went around the room.
DR. ANDRESEN: It's Elena Andresen. I have one of those names that causes lots of problems. It's Norwegian. I actually know something about Minnesota, even though I'm from the Midwest right at the moment, I might fit in culturally to some of those around the room.
I am happy to be here; my luggage is not. None of my handouts are here. And so even though I have my slides, I have been doing this for long enough to figure that if I can lose something, I will. I'm going to go ahead and use a little bit of my prepared presentation.
I'm an epidemiologist. I don't represent anybody today. I don't represent epidemiology. I don't represent an institution. I don't represent a group. I don't represent therapists, medical people, any of that sort of thing. My job is an epidemiologist, and if you haven't heard from one before in these proceedings, I will try to make myself more population-level discussion, rather than some of the subpopulation-specific discussions we have heard so far.
And in that vein, although some of my slides were going to talk a little bit about how to select a measure of function, and not specifically just on ICIDH-2, the reason I'm here this week I think was because I have to do a study section at AHRQ at the end of the week, and I couldn't come the time before when things were a little bit more general. But I do have some feeling about ICIDH-2, having been involved in some of the earlier alpha testing of it. So I can address those as well.
I think since you have a handout of my slides, none of the supporting documents that go along with it, what I'm going to try to do is talk a little bit about some of the major -- what I usually talk about and what I do research about is how do you select the best measure for the best purpose. Since I'm an epidemiologist, the questions that are usually addressed to me are how do you use the best measure in?
And then it's at the population level, and the specific study, in an outcomes research application. I'm an unusual epidemiologist in that a lot of my training has been in health services research. So usually I'm with a bunch of economists and sociologists, and I'm the only population epidemiologist kind of person there. But I bridge both worlds very comfortably, and I like them a lot.
So when people ask me that question, some of my funded research has had to do with head-to-head comparisons of different ways to measure effects at the other end. And for a number of years, first more in older adults, and now mostly in populations with disability, the question has come to be about how do you measure, and the word has usually been health-related quality of life?
And we use that term very loosely. It has evolved over the years. I'm old enough and been around long enough that it used to be called health status, and before that functional. What I mean by that is people have talked about for example the SF-26 today as having measures of function. That always surprises me a little bit, because it does a scale about physical role, mental health role. And it has a scale on physical health. But as a measure of function, it is not particularly good at the level that most people want a functional assessment measure.
So I think there are a lot of different ways to get at the kind of information one might want on an administrative record. And I want to start by saying it depends on what you want to use it for. At the population level I would think that we wouldn't be always using a measure of function just as a rehab outcome, or just in a rheumatology clinic, or just in OB GYN. We could have the list go on and on.
And because the use, therefore, is for a lot of different kinds of uses, it becomes extremely challenging to pick what is the best way to measure it. I like the ICIDH in its fullest form as it goes along. It would be just absolutely wonderful, regardless of what kind of use that you would. You should be able to find enough detail there somewhere to apply it to the use that you are looking for. But then we get into questions of practicality.
So I want to go through a few of the issues that I think probably need to be there. I think what I won't do is show you what some of the findings that I have seen in terms of how well some of these measures of functional status work. It was nice to hear this morning that there has been some testing of the ICIDH on some of these key measures. I like to see that it's gone past the developers before I accept a good liability statistic. I'm sure that we would all like to have that happen over time with ICIDH.
Unfortunately, it sounds like you going to try to make a decision before all of that has occurred. And that is going to mean that --
DR. IEZZONI: No, our decision might mean further study.
DR. ANDRESEN: Oh, wow, okay then, I'm in good company.
So these are the kinds of people I have been working with for the past few years to try to evolve some sense of what is the best measurement tool. Those include folks in academics, people in public health, public health departments, and people at the federal level.
I'm going to go through a selection of criteria and make some comments about them. And if I have time, I will just show you one use of a functional assessment kind of measurement tool, and some of the pluses and minus of selecting any one for a use.
The first one that I think is important is to talk about comparative norms and standard values. What I mean by that is whatever you choose, whatever goes into this administrative record, whatever is deemed to be the most useful, you want to be able to use it to compare different groups, to look at the population in general. And you need to have some sort of standard values that everybody can understand.
And so if you move past something like plain vanilla activities of daily living, or elements that are on the minimum data set for nursing home use, that what you need to know is are you going to be able to look at a particular health care setting, and find out whether or not that population looks like the population at large from which it came? Or does it look like all the populations that go to these kind of medical encounters?
And it is going to take a long time to establish normative data. So if you change to something else for which there is not some sort of national level normative data, then there is going to be some time lag. If it isn't the same measurement tool, if you are not using something for example from the behavior risk factor surveillance system, or the national health interview survey, or other sort of nationally normed kinds of data, then you need to the best -- the word that was used earlier by someone is crosswalk between whatever is in your administrative record, and whatever is used by other kinds of systems.
We have a lot of different kinds of definition and measurement tools for things like function and disability and health status and so forth. What that usually does is provide a fairly large cloud for understanding differences between populations and groups, and across the country at large, and temporal trends. All of those kind of nice things we like to know in order to make plans for the future, or to figure out if we've got groups who are underserved or overserved, or some of those kinds of things.
So if a selection is made to put something in an administrative that may work pretty well in let's say in crosswalking with the minimum data set, then there will be some problems with it. If a decision is made to put something in it that looks exactly like elements about function that are on the national health interview survey, then you'll be able to do a really good job of taking administrative information, and looking at the population at large on a regular basis.
I'm not saying that's the right decision. But that's one of the abilities you have in picking a data element that already has a history, and sort of normative data that go behind it.
People have been talking a little bit about -- your topic was granularity. And I think of that as being one of the aspects of having the right measurement models. And what I mean by that is that something like ICIDH, or the function assessment of activities of daily living may have enough variability for some populations. For example, nursing homes and people who get home health care, people who are in acute rehab, and so forth and so on.
But the population at large for a lot of ADL-type items you will have 1 or 2 percent of people who answer the question affirm it as a yes. And that might be very interesting information. But if you are using something that measures basically from some level of impairment that is usually part of a fairly regular medical encounter, at least acute medical encounter to the most severe, then almost everybody who goes to the doctor on a regular basis, who has a head cold or this, that, or the other thing, is probably going to be a no on a lot of those elements.
That's okay, but that does mean that you are imposing something on an administrative record for which the majority of people might be nos on it. You have a functional assessment tool that does not describe the differences between say me and my husband. I see a rheumatologist and he doesn't, but on an average medical encounter you probably wouldn't pick up anything on any of the kinds of scales we are looking at to define me.
So a selection that has to do with a tool that measures some level of impairment, and there would be a line up to sort of the maximum impairment, will probably leave out a lot of people. And even when we talk about very broad tools like the SF-36, this is a tool that has a lot of problem with its measurement model. If you have some scales on it that look at for example, mental health role, for whom 60-70 percent of the population gets scores of 0. And they probably different one from another, then there is a problem.
So it's not just an area of functional assessment that might be useful in medical encounters for people who have acute or chronic problems. You might distinguish between them and not then not somebody coming in for an OB GYN exam or something else. It really occurs in tools that are a lot more general as well.
Burdens. When I talk about burdens, I am usually talking about imposing burdens on subjects and administrative level. And because this was a discussion about administrative records, I decided to keep most of my comments having to do with that. But you can ask people themselves about their functional ability. And we've had lots of practice doing it. I usually collect data using a computer program, so that I can tell you exactly how long it takes on average, for a particular question when you are interviewing people.
So something like the 14 questions that are used or could be used on the behavioral risk factors surveillance systems, the BRFSS, which is a state-based annual survey done in all of our states. I don't know what it's like to go through medical training. I do know what it's like to get people with medical training to use any kind of data collection tool, and it's extremely challenging. Extremely challenging, even with some training.
What has often happened in any kind of imposition of data collection in any kind of clinical setting in health services research that I have been involved with is that somebody like a nurse or nurse practitioner, and sometimes the clerical person on the unit is getting a lot of that information. What that means is there is somebody who has to be trained on an ongoing basis. You have to change everybody over to using a new form. And then you have to retool people on an ongoing basis.
One of the affects that we haven't talked very much about or I haven't heard today has to do with quality assurance. Whatever tool that you are picking up that is going to be used, you are going to make it cheap, simple, understandable, and then someplace, somebody is going to have to do quality assurance on an ongoing basis to make sure that the information is good.
In an academic medical consortium that is now defunct, the used the SF-36 for a huge number of patients. They all agreed to do it at some academic medical system. When it came right down to it, and I was working with a student working with a thesis, only 70 percent of those data were complete enough to be used. Well, the SF-36 takes a little while to administer, 7 or 8 minutes on average. But if you only had 70 or 80 percent of the records with that information in it, it's not quite as useful as an administrative tool as one would like.
There has to be ongoing quality assurance to make sure that things are in fact universal. Missing data is just an extremely problem regardless. Everybody's best intention could go out the window in getting a form filled out at the moment when you are dealing with somebody who has come in with something more complex. And that will get done first.
I can tell you at my rheumatologist and the people who know the most about me are the medical students, interns, and the residents who spend usually 40 minutes sitting with me doing long interviews before the rheumatologist comes in. And so it's possible that in those academic medical settings, you could add all kinds of information and get lots of detail.
But it is extremely hard to get people to collect this kind of information on a regular basis. Especially, although we might think of it as cheap, simple, and understandable, they don't see it as having anything to do with helping them with medical care.
Responsiveness. We have talked a little bit about that. It's one of those things that doesn't happen very often. Even the SF-36, which is usually held as the gold standard, has a problem being responsive. Nobody is talking about whether or not if you would ever use this, to find out whether or not over time if you do an intervention, you will see a change in this. Or whether or not it's more useful as a population tool to talk about medical encounters.
Responsiveness has its own science. It's very complex. There are scientists who could come talk to you about it. And usually it's not as good as people think that it should be.
Reliability we have talked a little bit. Are they are repeatable by individuals over time, among raters, by proxy respondents, that sort of thing. I had the opportunity to test the new census 2000 questions on disability, and you know I guess I wish they were a little bit better than they were in terms of family members and people with some health care training responding similarly to the person with the disability themselves.
Regardless of what the choice is, I think that over time it is going to take a lot of work to make sure that it's really well known whether or not you can get the same answer from whoever fills it out, not just when it was developed in an academic center, with lots of data collectors available, but in the real world.
Validity, you want functional assessment to have been the truth. Does it do the same as other tools. Is it convergent with others? I can tell you from comparing different kinds of functional measures that even when they sound very similar, you get very different answers. And that's a worry. If you have a question on the census, it is supposed to get at does somebody need help with personal care, and it doesn't measure the same thing as a personal care variable on the behavioral risk factor surveillance system, it's like is one of them better than the other? Is it wrong?
Well, the main problem we have is that they don't seem to get at the same kind of information. Whatever you choose, and if it doesn't get at the same kind of information as the other, you can't quantify how different it is. It's going to be something of a problem.
Discriminate ability. Nearly every functional assessment in disability and health-related quality of life tool I have ever looked at in the literature does just one bang up of a job. A great job. If people actually are different, usually these things work fairly well. So that's usually what you hear people talking about. Can you tell the difference somebody who has cognitive impairment and someone who doesn't? Can you tell the difference between somebody who has upper body level impairment and someone who doesn't? Somebody who is in a wheelchair, versus somebody who is not?
Discriminate ability usually is great. And you see lots and lots of information about people looking at discriminate abilities. The rest of these kinds of functions are really usually not quite as good.
DR. IEZZONI: Elena, we are running a little shy on time, and I notice you have quite a few examples.
DR. ANDRESEN: I'm just going to leave you with the slides. I'm going to show you one example. I don't have any in the article, so it really doesn't make any sense for me to be going over this. I just want to use one example of a question that is used at the population level.
In many states there is a question on the behavioral risk factor surveillance system that says are you limited in any way in any activities, because of any impairment or health problem? And in Missouri we have tracked this over time. It came down quite a bit, as it did in a lot of places in the country in the mid-nineties. We're seeing a little bit of a trend back up again.
Well, now that's great. That gives us some information about Missouri. We can also compare ourselves to other states that have these data, and at the national level. What it can't tell us is why this might be occurring.
So getting back to the ICIDH, when you have something that gives you a piece of information about a functional ability, but it doesn't give you any information about the why, let's say it's the social environment, or some other sort of thing that is changing as well. Then it's not quite as useful, except as a very simple descriptive kind of a tool.
The other example is this exact same question that used at the national level in a study of minority and majority women looking at physical activity, but also had information about physical impairment based on questions from the behavioral risk factor surveillance system. And this also was very useful. It's nice to know and be able to look at data that suggests that there are differences in majority and minority women, about whether or not the need personal care assistance, and have severe work disabilities.
These are also questions that can be derived from the behavioral risk factor surveillance system. But again, even though there is this information, even when looking at -- in my case I tried to explain these differences by any kind of variables I could -- none of the kinds of information about participation in environment were really there, other than these kinds of variables. So something like the ICIDH, that has the ability to at least compose limits in different areas, would help a lot, I think, at the population level, and even in administrative data records.
We talked a lot about errors and measurement issues and so forth. I just want to add that it would be lovely to see these kind of data in administrative records. See places like -- since I'm reviewing grants for AHRQ at the end of the week -- watch those dissertation proposals come in five years from now with some of consistent kind of data element that would allow health services research doctoral students to do a better job of talking about medical encounters, and need, and unmet need, and all the rest of it based on something that was better than what is currently in those kinds of national data sets that they have available to them.
But I don't know if you have had a chance to talk to AHRQ. I'm not aware of any plan on their part of incorporating this kind of information in the special kinds of data sets that they use for more research purposes, that they make available. So that would be great, if you could get the research community and evaluation groups kind of on board with this. I didn't see anybody's name from AHRQ in the meetings that have come before this.
I think I'll stop there and not worry about the examples that I have from data.
DR. IEZZONI: Thank you. Is your suitcase going to show up at some point?
DR. ANDRESEN: That's a great question, because I would certainly love a change of clothing. Unfortunately, I'm traveling with my husband and he had the Woolite in his baggage. I have no idea. It's in Cleveland, as far as I know. I'm not in Cleveland.
DR. IEZZONI: Well, this panel and the prior panel actually brought up the whole issue about an assessment tool versus a classification. I think that we are getting a little bit mixed up in how we are thinking about this. We are probably not alone in this.
I think Gerry and Paul, you clearly say that the ICIDH is a classification system, not an assessment tool. It's not like the SF-36 or a scale, or am I putting words into your mouth?
DR. ANDRESEN: I have been told that there will be research applications, it just has to come down the road.
DR. PLACEK: There are assessment tools now for ICIDH-2 beta-2.
DR. IEZZONI: There are, but what WHO has promulgating and what will be voted on in May 2001 is the classification system that can be used internationally. It's not an assessment tool.
DR. PLACEK: No, there are tools developed from it.
DR. IEZZONI: There are tools that can be developed from it, for example, the MDS and OASIS could possibly be crosswalked with it. And so I think that what we need to understand is to what extent it would be hard to crosswalk OASIS and MDS PAC and MDS with it, and what would be the utility of doing that. It may or may not be terribly useful.
DR. ANDRESEN: I can tell you that crosswalking among different kind of measurement tools sounds better than it works in practice.
DR. IEZZONI: It's not a measurement tool. It's a classification system.
DR. ANDRESEN: I'm using an inappropriate word then.
DR. IEZZONI: I'm still struggling with what the distinction is too.
DR. ANDRESEN: You could mandate that ICIDH replace MDS for the elements that have to do with classifying people's functional levels, and that would work. But can you use the current elements on the minimum data set that classify people in terms of their function, and crosswalk directly with the ICIDH? I have used those data, and I can tell you people like Vince Moore would say no way. I don't know. It would be lots of work.
DR. IEZZONI: Andy, do you want to weigh in on this?
DR. KRAMER: Well, I'm also having difficulty with this whole issue of what the ICIDH actually is. And if it's not what it is, what the sort of halo of it is. Because when people talk about it, they talk about it in the context of it sort of connoting a whole structure of care as well, which is again, along the lines of the MDS was intended to do. It was intended to be sort of an assessment. It also triggered RAFS(?). It also triggered a whole approach to care.
So I understand this sort of ambiguity. It's quite different from sort of an ICD-9 code on some level that says here's a diagnosis. This tells you there is nothing wrapped up in this diagnosis that tells you how you look at an individual case.
On the other hand, it may just be our lack of incorporation of sort of function and activity limitation in care practice that makes it appear to be such a lengthy assessment. And could in fact -- could it be filled out based on information available?
DR. IEZZONI: But I think Alex has a good point, that there is a standards process for looking at medical classification systems, and ICIDH has not been subjected to that scrutiny yet.
Can you introduce yourself and make your comment?
MS. KENNEDY: I'm Cille Kennedy. I'm chair of the international mental health task force that has been working on this since its inception. You have had wonderful testimony I know from many members of my task force yesterday and today, and I'm very pleased that we were well represented.
And because I know you have had a lot of rich and thoughtful material, I would like to refresh your memories to some of things that Bedirhan Ustun said this morning. You were asking about ways of either documenting or assessing disabilities. And I would like to remind you that he showed you the ICIDH-2 checklist, which is a very brief way of just documenting the disabilities, and it has the ICIDH codes in it.
DR. IEZZONI: I don't think we want to assess.
MS. KENNEDY: This is documenting. And then for research purposes, and just to make the clarification, for research purposes there is the World Health Organization Disability Assessment Schedule. One of the things we have done in the field trial is to use the ICIDH-2, the checklist as a preliminary way when we are interviewing live cases, individuals or summarizing out of medical records, just documenting it before putting it on another form.
DR. KRAMER: But I guess the question is how you get to the boxes, the checks that you are documenting.
DR. IEZZONI: We're not really getting into that whole assessment issue. We're kind of treating this the way that ICD-9-CM functions.
DR. KRAMER: Where you write down a code.
DR. IEZZONI: Where you write down a code that classifies what you, as a clinician, in your own implicit judgment have decided ought to be represented there in terms of the patient's functional status.
MS. LUX: Just very quickly, I think that the point has already been made that the assessment essentially are already ongoing by all of the disciplines represented and others. We all have our things that you do, if you will, to evaluate the language function and the whole nine yards. This is simply a structure then in which those existing assessment results are placed to give a more comprehensive picture of what's going on with the client situation.
DR. IEZZONI: The code to classify that. So across disciplines, based on your own disciplinary evaluation.
MS. LUX: And/or the patient self-report.
DR. IEZZONI: Well, I think we are going to have to, as a committee, really struggle with clarifying our own understanding of this. We will have complete transcripts of these meetings, and we'll go back over the exact words that various people have used, because I think the people testifying to us have actually parsed their words very carefully from their own disciplinary perspectives, to try to convey what they are interested in. But we are going to have to be sure that we can be clear about what our use is.
I did promise, and their people whose jobs depend on me being on time, that we would end in about 5-10 minutes. Are there questions for the speakers?
MR. HANDLER: I have one question for Dr. Ruggieri. A number of times you mentioned modeling, whether this is an appropriate model to use. Well, ICD-9 is not a model. It's a way of categorizing causes of death or procedures or disabilities. If that's not a model, why do you consider this a model or a series of models?
DR. RUGGIERI: It's true, ICD-9 is not a model. ICIDH, based on the documentation, there is a declaration. There is a concept functioning. And there is a declaration that functioning is comprised of dimensions. So you begin to talk about entities. You begin to talk about relationships. And on the basis of that tacit model, and I think it is a very elegant model. To be honest with you, what it needs is formal, rigorous expression in a syntactically rigorous modeling language, which will ferret out ambiguities and inconsistencies in it.
Because there are some ambiguities and inconsistencies in that can lead to its I guess in the world of epidemiology it might be in this classification a type of bias. But anyway, I think ICIDH is based on a model, and I think the diagrams that are posted on the ICIDH site indicates that there is a mental model. It's kind of like a chick trying to break out of an egg shell. But like I said, I think it needs formal, rigorous expression.
DR. IEZZONI: Can I ask you, you know something about the U.N. process and procedures that the rest of us may not. Do you think that ICIDH could be put through this egg beater type of process that you seem to be describing, and come out the other end in any way that looks like it looks now, that people could recognize it?
DR. RUGGIERI: Yes.
DR. IEZZONI: You think it could?
DR. RUGGIERI: I do, but like I mentioned, in any modeling effort, there is a way you do modeling. You bring in someone who is a data modeler, and you bring in domain experts. And you put them in a room, and you shake. That's how the model was elaborated. That's what we have been trying to do. And we have come to fisticuffs over what ICIDH is trying to exactly say.
And the reason we want to be exact is because we want ICIDH to be incorporated into the ultimate computerized medical record. In order for that to occur, there has to be exact definitions, because you can't hand the ICIDH to a programmer right now. They will hand it right back to you and tell you there are inconsistencies.
DR. IEZZONI: That's really fascinating.
DR. ANDRESEN: Can I ask a question? When you are talking about what you are looking for is a classification that can be applied to administrative records, and I think there is a language disjoint here that probably I'm responsible for. Classification in epidemiology is just a method to get to something else. So that a classification system that would be applied to administrative records would be useful for some kinds of purposes at the other end.
One of those might be evaluation, and one of them might be assessment. There might be all kinds of uses for which these data were used. So the selection of the classification system should be in hand with how it were going to be used, not just as a descriptive tool on its own, or have I missed something?
DR. IEZZONI: I'm using classification as ICD-9-CM uses the term. A classification of in the ICD-9-CM, it says diseases, although it is not all diseases. There are symptoms and signs and other things in there. But it's a way of kind of representing with numbers, concepts, quantities.
DR. RUGGIERI: Could I just mention one brief thing? Regarding a model, if anyone would like to look at a UML model, the CDC U.S. Public Health Service has I know made a document that expresses a public health information model using the Unified Modeling Language notation. And I can't tell you where that document is available, because I stole it from somebody.
MS. GREENBERG: I think it may be on the Website, but also --
DR. IEZZONI: Which Website?
MS. GREENBERG: On the CDC Website. I can find out. I have some hard copies of it myself. But actually Dan Friedman, who is a member of the National Committee, a former member of this subcommittee has suggested that the NCVHS have a presentation on that public health data.
DR. IEZZONI: Okay, so we are ignorant of it.
MS. GREENBERG: People who have worked on it with CDC, domain experts I guess, also have been very involved in the HL-7 modeling process.
DR. RUGGIERI: Just one brief thing I want to tell you. Regarding a UML model for ICIDH, and a functional status informational model, we are working on that right now, and hope to have that.
MS. GREENBERG: By "we" are you referring to HL-7?
DR. RUGGIERI: "We" are referring to Medical Information Resources at Mayo. The people I work for.
DR. KRAMER: A quick question for you Dr. Ruggieri. What is Mayo then going to do about classifying functional status? It sounds like you are working on this ICIDH model. But if that doesn't resolve, do you have another solution?
DR. RUGGIERI: Well, Mayo's philosophy is, from my perception, is that our goal is to represent what the provider, whether that be an occupational therapist, physical therapist, respiratory therapist, physician. We want to represent the expression in terms that are meaningful to them in a computerized format. And if ICIDH serves some uses in the clinical arena, then we want to provide that classification in that scheme, available in an automated form.
That is pretty much our approach to it. We of course record functional status. We use all the health assessment instruments that are available to everybody else. Some of them have been in-house designed health assessment instruments. We have no formal classification system. We will provide ICIDH to our users as they so desire. I hope that answers that your question.
DR. IEZZONI: Then the question is going to be who desires this. I know that there are people who do. Probably your OTs might, the PTs might not.
DR. RUGGIERI: To be honest, I don't know about the institution-wide awareness of it yet.
DR. IEZZONI: Yes, that's a problem I think systemwide.
Thank you, it's been very provocative. And it's good for us to get presentations that push us on paradigms and language.
Yes, Paul?
DR. PLACEK: Do you have time for a one minute commercial?
DR. IEZZONI: One minute, okay.
DR. PLACEK: This room is filled with Americans, except for Michael Wolfson. I want to point out a meeting coming up in two months where the Canadians and the Americans get together, and the U.S. citizens, whatever you want to say. The North American Collaborating Center meeting on ICIDH revision will be held here in Washington, September 14-15.
That's pretty much an equal mix of about 20 U.S. researchers, 20 Canadian researchers, and usually a half a dozen other countries represented. So if you want to broaden your discussion to include more Canadians and other countries, and take another look at the ICIDH, you are all welcome September 14-15, Radisson-Barcello(?), Washington, D.C.
DR. IEZZONI: Paul, I'm going to be in Dallas then. Andy, I don't know whether you might be able to come. We should whether maybe one of the subcommittee members could go to that.
MS. GREENBERG: It would be excellent if one of our subcommittee members could be there. Just to follow-up on what Paul said, this meeting will be our really opportunity to try to bring all of our thinking and research and beta test results, et cetera, together for the Madrid meeting, the international meeting, which is seen as kind of input to WHO on the kind of revised draft that they would do before the Madrid, and then the Madrid meeting. So we certainly would benefit from any guidance that the committee might give us.
DR. IEZZONI: It sounds like it's going to be a great meeting. Andy is going to check on his calendar.
We can check with Paul and see whether anybody else can do it.
MS. GREENBERG: The problem is that the full committee is meeting on the 19th and 20th too.
DR. IEZZONI: Oh, thanks for reminding us.
I want to thank everybody for the time that you spent. You have really stuck with us through the two days. I think actually it's been a wonderful two days. Again, thanks Susan, Queen, and Paul, and Gerry for putting it together. And thanks to the speakers for coming and sharing your ideas with us. Safe journeys home.
[Whereupon, the meeting was adjourned at 1:00 p.m.]