Conference room E-275C, Lowrise
John F. Kennedy
Federal Building
100 Cambridge Street
Boston, Massachusetts
A P P E A R A N C E S:
Lisa I. Iezzoni, M.D., M.S., Chair
Hortensia Amaro, Ph.D.
Richard
K. Harding, M.D.
Vincent Mor, Ph.D.
George H. Van Amburg,
M.P.H., M.S.
M. Elizabeth Ward
Paul Newacheck
Carolyn M. Rimes, Key Staff
Olivia Carter-Pokras, Ph.D.
Dale
C. Hitchock
Ronald W. Manderscheid, Ph.D.
Dr. Jack Fowler
Dr. Phillip Caper
Brian Burwell
Barbara
Fain
Wendy Mariner
Nora Wells
Mr. Shekar
Mr. Robbins
Mr Fowler
Mr Caper
Mr.
Burwell
Ms. Fain
Ms. Mariner
Ms. Wells
DR. IEZZONI: My name is Lisa Iezzoni, I'm the chair on the Subcommittee. We're here in Boston, happily my hometown and the hometown for a number of us. I'd also like to thank staff to the National Committee for all of their hard work. Carolyn Rimes deserves a special thank you as does Jason Goldwater and Lynnette Araki. And so with that, I'd like to go around, have the committee members introduce themselves quickly and then everybody around the table just so we can save time. And we'll hear from the audience. Elizabeth, do you want to start.
MS. WARD: I'm Elizabeth Ward, member of the committee and I'm from the Washington State Department of Health.
MR. VAN AMBURG: George Van Amburg, I'm a member of the committee from the Michigan Public Health Institute.
DR. NEWACHECK: I'm Paul Newacheck, I'm a member of the committee from the University of California.
MS. MOIEN: I'm Mary Moien under contract to synthesize the findings here.
MS. RIMES: Carolyn Rimes, I work in Hancock.
AUDIENCE MEMBER: I can't hear you.
MS. RIMES: I'm Carolyn Rimes, I just work here at Hancock.
MS. COLTIN: I'm Kathy Coltin, a member of the committee and I'm with Harvard Pilgrim Health Care in Boston.
DR. IEZZONI: Dan.
MR. FRIEDMAN: I'm Dan Friedman, also a member of the committee with the Massachusetts Department of Public Health.
MS. GREENBERG: I'm Marjorie Greenberg, National Center for House Statistics, Executive Secretary of the committee.
DR. IEZZONI: Dale, do you want to just introduce yourself very quickly.
MR. HITCHCOCK: I'm Dale Hitchcock, I'm with the Department of Health and Human Services. I'm with the data policy group and I'm a Massachusetts native.
DR. IEZZONI: Oh, great. Oh, good. Okay. Can we just go around the audience.
MS. WALK: Racheal Walk, Health Care Finance Administration -- (inaudible).
DR. IEZZONI: And Lynnette, yourself.
MS. ARAKI: Lynnette Araki, National Center for Health Statistics.
MR. GOLDBERG: I'm Jason Goldberg - (inaudible).
MS. GASOWSKI: Gretchen Gasowski, Lisboa Associates, Inc.
DR. IEZZONI: Okay. Great. Sam, do you want to give us some brief comments and then I think we'll hear from Tony briefly as well.
MR. SHEKAR: Thank you, Lisa. And, again, welcome to Boston. Tony and I are partners in crime, along with Judith Curlin, the regional director in working on issues of interest to public health and health care delivery in this office and on behalf of this region. First of all, I'd like to welcome all of you to Boston today for this important meeting. As a former HVHV meetings in the past, as the regional health administrator for regional one, I'd also like to thank Carolyn Rimes, and the HVHV staff, Betsy Rosenfeld, the deputy regional health administrator, and Laurie Golden, our student intern for ensuring smooth arrangements for today and tomorrow's meeting. We applaud the idea of having regional meetings. We look forward to more regional meetings with my colleagues in the future and we want to compliment you on agreeing to do so. This allows a variety of experts to present, who otherwise may not be able to in Washington, and any movement towards increasing the interaction of the federal government with those fellow citizens in the states and communities exist to be applauded. The subjectunder discussion today is an important one, particularly in New England. Medicaid Managed Care is a growing phenomenon particularly in this region. Over 55 percent of the Medicaid beneficial population in the six New England states is now in one form or another under managed care arrangements, over 500,000 Medicaid beneficiaries in Massachusetts alone. And the growth, that growth raises all sorts of issues which arise from the significant change in Medicaid responsibility from being a claims payer to being a valued focus purchaser of care. Among the many issues to be explored today and here and elsewhere is the issue ofpublic accountability for services and data. It stands out particularly in this region where states and community have had a very strong and abiding interest in health care issues. Our office has long been concerned about the growing intersection of managed care and health care delivery evolution with traditional public health functions. And we see a strong rationale for mutually collaborative activities between those two growing entities, between these two major entities to improve the community health. To give a little more indepth discussion on the public health managed care, we have with us as well representing the regional office Dr. Tony Robbins who's the editor of the public health reports and former director of that vaccine office, et cetera, et cetera.
MR. ROBBINS: Thank you very much, members of the National Committee, for coming. I'm sorry that though we had this meeting planned long in advance, the IRS seems to have set the national date for filing taxes on April 15th and you all had to put up with the crowds downstairs. The -- I guess the reason I'm here is I've been around New England for a long time and I thought I would tell you, if you weren't already aware, a lot of interesting data work began in this part of the country. Jack Fowler is here and he goes back to the beginning when Allen Giddleson and Jack Wenberg set up the data system in Vermont that led to the first National -- first time in the country having all hospital discharges from a given region, given state in one data base and tied into census data. So for first time we're outside of managed care plans and organized populations. One started to learn about what kind of effect on health a medical care system was having. Connecticut contributed DRGs; in Maine there was really the first and best collaboration between the medical society and statewide data; Rhode Island had the search -- had Search, which was another statewide data base. And so New England has been on these problems for a long time. I'll leave it to Dan to make -- to talk about what's been going on and what has been going on in the past in Massachusetts. It was partly because of that kind of interest and because the New England states tend to have strong state healthdepartments, minimal city and county health departments that there seemed to be a logical focus for the -- this region to take a serious look at whether organized medical care, and you can read managed care out of that, could somehow contribute to a better handle on the public health and on the health of the people. And so, a lot of thinking has gone on in this area, and we in the regional office have just convened earlier this year for the first time the six state health officers and all of the major managed care plans to see if collaboration is possible, really with a goal of public health as much as anything else. It's not intended only to improve medical care, but to assist state health departments to do their job better. The one other thing I will say, and I will bring up later for you, is a piece that we published last year by -- in public health reports, which I edit. There are a few copies of the latest issue around. But last year, Allison Pollic and Dorothy Rice published a piece in the journal which I think is -- you may have all seen it by now, but it's a fairly strong argument for doing what you can do to protect the public uses of medical care data. Because, as they point out, we're the only country in the world where we still try tofigure out what the medical care system is doing by surveys, and other countries have gone on to using their data generated out of the medical care system. So welcome and thank you all for coming.
DR. IEZZONI: Thank you very much Tony. Good introduction. We are interested in data and we're interested in what data can tell us about the impact of Medicaid Managed Care. And so the first panel, I know some of you have been very involved in both looking at data and actually generating data and so we're anxious to hear from you. We're going to follow the order that Carolyn has established. And what I'd like each of you to do is just introduce yourself briefly as you begin your comments and then I think we have about 15 minutes scheduled per person. What I'll do is I'll ask the committee for specific questions about content or clarification after each speaker. And then after the first panel, we'll pause for a while and see if there are any more global questions from the subcommittee. Okay. So, Jack, could you introduce yourself. Okay.
MR. FOWLER: Actually, I have some overheads. I don't know if I need a mike.
DR. IEZZONI: You do.
MR. FOWLER: Yes.
DR. IEZZONI: Maybe somebody -- Lynnette, do you think you could do the overhead. That would be great. (Discussion held off the record)
MR. FOWLER: We can do it that way, I can sit and she can do overheads and that will work.
MS. RIMES: Or we can slide a microphone over. There's all kinds of options here.
MR. FOWLER: Well, the low-tech one seems to be easiest, often that's the best.MS. RIMES: Okay.
DR. IEZZONI: Okay.
MR. FOWLER: My name is Jack Fowler and I work at the Center for Survey Research at U. Mass. Boston. And when Carolyn called, I tried to think about what I had to say which might be helpful, which 15 minutes would be the most useful. And I guess there are two things that we've been doing that seem to relate most to what is of interest to you. One is that we have been working with the Department of Medical Assistance which is in charge of the Medicaid program in delivering care to people in Mass. Health. And doing -- helping them with collecting data from enrollees about their experience with managed care and not so managed care, and as an accountability process. And that's something we're -- we have an ongoing thing that we're doing. But in particular, I'm also part of something -- I guess you can do the next one -- called a CAHPS project, which is a national effort to try to get feedback from patients about their experience of care. And one of the groups that we're working with is the division of medical assistance in Massachusetts to try to learn more about how to do surveys of Mass. Health beneficiaries better. And it seems to me that that convergence of interest may be the topic that would be of most interest to this group. So that's what I thought I'd talk about. And we do have a -- you can do the next thing.
One of the things that -- the goal of the CAHPS instrument is to attempt to get an instrument or actually a set of instruments that can be used across different populations so that -- and the vision. I mean, it seems to me to be quite important that the notion that you can get actually comparable data with the experiences of Medicaid people with privately insured folks; that you compare managed folks with not so managed folks; that you can comparewith the people who have significant health experience with a cross-section of population and be able to make statements about whether their experiences are the same or different seems quite critical and a very exciting potential that the interest and accountability has generated. So this is the goal of sort of working on the instrumentation for that kind of thing is what the CAHPS project is about. You might want to do the next one. One of the things we worried about a lot is what's the best place to get information of various sorts. That surveying people is only one way to collect information about people. And there's some things that people are uniquely able to tell you that you can't get anywhere else. And there are other things that probably you wish you could get from people that enrollees are probably not very good sources for.
For example, the three general areas that we found where a lot of testing that people are good at is they can tell you about Access, problems with getting appointments, getting referrals, getting care off hours, on hours waiting in waiting rooms, things of that sort clearly -- patients under better shape than anybody else to have that. No one keeps track of that stuff and they can tell you what their experiences are. They can tell you about their interactions with providers and physicians and nurses and office staff and other things like that. And they can tell you about the interactions with plans about things like claims or payments or hassles, et cetera, et cetera. And those are things that people are very good at, and it's reasonable to get from people.
On the other hand, when we started to make our list of what would folks like to know, it turns out they're getting information about costs is very poor from people, they don't even know what they pay themselves. Even things like premiums turn out to be hard to get because they get deducted and things, much less what is paid for their care when some of it's paid directly, et cetera. They're not very good about saying what's covered and if what's covered is good or not. They run into something they need that's not covered sometimes. But that's not a very good topic of information for people.
When they ask them to rate their doctor and whether they're good at doing things like diagnosing and treating, they really don't have a clue. And, actually, it turns out when they rate doctors, how much time they spend with them is probably more important than -- that drives the answer to how qualified they are. And outcomes are really tough, not because people can't tell you how they're doing but you have to -- but whether they can attribute how things turn out to their treatment experience is really hard. And unless you sample in very specific ways, you don't know where people started so it's hard to reach conclusions about where they ended up. So those are things that are not easy to get from patients except under very specialized circumstances. Another one of the really exciting things that we've been doing and I think one of the perhaps most important, I'm a methodologist so that's the kinds of things I think about a lot; is that CAHPS has generated a lot of systematic interest in methods issues about how to collect data and how to solve these problems. And particularly relevant back to your topic, which is Medicaid and DMA, we'll be working this spring on, I think a set of very innovative tests to figure out better how to solve some problems of data collection that distinctively are problematic for collecting good data about Medicaid population. One of the key problems of most people who surveyed Medicaid populations have is response. Folks like mail surveys because they're cheap and anyone with a typewriter and postage meter can do one of those surveys. And the unit costs are pretty low, if you think of 25 or 35 percent response rate is okay. But a lot of people worry about that sort of thing. And indeed, surveys of all health plans suffer if they want to stick with mail.
So one of the things we've been working on is the problem of, one, how to maximize mail returns, but also what the additional benefit is of switching to other modes such as adding telephone follow-ups and maybe even inperson follow-ups to enhance response rate and get it out. And, in fact, we're doing an experiment in Massachusetts where we are doing exactly the design I just mentioned; we'll start with mail, then we're going to do telephone, and then we're going to go in person and try to get a description of what you miss, what are the characteristics and what are the reasons why you didn't get them at different accepting low response rates, and what the benefits would be of trying to collect data from all or a sample of nonrespondents and some alternative mode.
That leads to the second issue which is effective mode on the data collectionon the data of estimates you get. Obviously, if you're going to use more than one mode in order to collect data, you need measures that behave the same way so that you don't change the data by collecting data in a different mode. Again, we've been doing a lot of testing to try to develop questions that are immune to mode and one of the things that CAHPS has been prettysuccessful about that in our test, at thispoint, we have gotten almost all the measures that are in the core of the instruments are -- seem to not be effected by motive data collection, at least phone versus self-administration which is a big advance, it's a different way of thinking about problems.
A third problem which is distinctively important for Medicaid populations so it effects everybody is the problem of dealing with languages. And languages are an issue of two sorts; one is just the inherent problem of getting -- of saying the same words -- getting the same concepts and words translated into two different languages and being -- actually measuring the same thing. And that was a consistent standard for CAHPS. And I'll just give you some -- for those of you who worry about survey research at all, one of the standards for the questions in the core was that they translate as well as we could manage. And so, two examples, it turns out that some of the questions about sort of how you're doing psychologically include things like downhearted. And blue, for example, and blue doesn't work very well as you go across different languages. And so that's a question -- example of a question. If that's a standard of yours, you might want to figure out some other way to ask it. And one of the things, for example, in ratings, it turns out fair, doesn't work very well. If you happen to be translating and one of the CAHPS decisions was go to numbers of 0 to 10 rather than excellent, good, fair, poor and one of this criteria for that is, we could do a better job across languages with numbers for our ratings than adjectives. It's a hard problem.
The other thing we are worried about and we're doing a test again in Massachusetts is how best to collect data. About 15 percent of people don't speak English. But 15 percent of the Mass. Health population is identified as Spanish. And there are two ways,at least that I know of, to collect data from Hispanics; one which we tried in an earlier survey with DMA was a Canadian-style instrument which read in English from the front and if you flipped it over, you would get Spanish the other way. And it worked very well if the standard is to get a good response from Hispanic speaking people, the rate -- it looked like it was right around the rate that you would have expected given other information, which was quite encouraging. What we didn't know is whether it has a suppressive effect on English speaking people to get a two-language instrument in this culture where that's not a very common thing to do.
We're trying the Canadian style again and we're going to send a postcard in Spanish in the other, please send me -- I prefer to answer in Spanish, would you send me a Spanish questionnaire. Our guess is we're almost sure that the dual language approach will work best with Hispanic population and we'll have documentation that either does or does notsuppress the response for nonSpanish speaking. So that's an interesting kind of -- (inaudible).
Collecting data about adolescence and this is really important not only for the Medicaid population but for everybody. In our testing and focus groups, et cetera, everyone's really clear that adolescents are clear that their parents don't have an include what goes on in their doctors offices or what they have to say. The adolescent don't know anything about their health plans. And we've got a survey that asks about health plan experiences and physicians. And when we talk to the parents, they're pretty sure they don't know what's going on either. So we're fairly confident that collecting data from adolescents, that some aspects of their experiences is essential. Including questionnaires saying, Please ask your child to fill out one part and have the parents fill out the other part, we're not sure how confidential that is from the kids' point of view if they get all sent back in the same envelope. It may lead to all kinds of data collections but we're doing one study where we're -- (inaudible) -- what they're anecdotalfeedback was that either they do or don't answer questions differently. So that's one we don't have down. But, again, we're doing experiments in that area which I think can be helpful if all kinds of settings. So the last thing I’ll mention, and I think my 15 minutes is probably up -- is there one more or did we use them all up?
MS. ARAKI: That was it.
MR. FOWLER: I can just tell you the other points of this is that the CAHPS visionwas that there was a core of questions that might apply to all patients; people with special needs; people without different kinds of managed care. But then it spun off some additional supplemental questions; if you're oversampling some groups or ask comparable questions than those that are asked for adults for themselves and children, there's another set of supplemental questions that are asked for people who are identified as having chronic conditions or special needs of some sort because there are issues that become relevant for those populations that don't show up as relevant for a cross-section of adults.
For example, emergency care or even specialists care is not that common within a six-month or one-year period for cross-sections of patients. And so asking in much detail about that is not effective use of survey time, but it's critical if you're going to try to find out with people with particular conditions or more significant conditions are handled.
Another variation on the family of CAHPS instruments is behavioral health issues; people who need counseling either for mental illness or for personal problems, et cetera. Again, for cross-sections, those aren't relevant because they don't happen often enough. But to have related kinds of instruments that can be in a forum that you can add and coordinate with the core is another goal. And, again, we're doing testing or at least we have some draft plans todo testing again in Massachusetts. We are testing the ones for adults with disabilities and also for children with special needs. And we'll be getting our first experience with those and be able to see how discriminating they are in terms of experiences. And we're discussing getting the behavioral health component tested here as well. So, that's what I have to say.
DR. IEZZONI: Great, Jack. That was really interesting. Are there any points on which people want clarification?
DR. NEWACHECK: Jack, I wonder, are you translating the CAHPS instrument in any other languages, in Spanish, for example, other languages that are important like in states like California.
MR. FOWLER: Actually, we clearly have accounts that has been a bit of a moving target and Spanish is the largest second language, it's the only -- that's the only one for which we have an official translation at this point. But clearly, once it gets more frozen, I think we'll try other languages. But, at the moment, no.
DR. IEZZONI: Dan.
MR. FRIEDMAN: Do you have any response -- (inaudible) -- in course of loans?
MR. FOWLER: We don't, and that's one of the things we're really looking forward to.
In a privately insured managed care setting in Colorado, we have those data about who doesn't respond to the mail, who we can pick up on the phone. But the Massachusetts test will be the first one where we'll actually have a Medicaid one where we can separate and contrast what the differential nonresponses by mail, by phone. What I'm really excited about is actually goingout in person and find out who the people are. We can't even list at all by any strategies and see how they're different. So I think it's going to be a very important effort for I think all kinds of purposes.
DR. IEZZONI: Jack, stick around because we'll have more global questions for you after the next speakers.
MR. FOWLER: Okay. Great.
DR. IEZZONI: Phil, do you want to introduce yourself and go for it.
MR. CAPER: Thank you, Professor. I'm Dr. Phillip Caper, Chairman and CEO of the Codman Research Group. With me today are Man on Spitzer Ruben, Vice President for Applications Research; and Jo Ann Lutz, Director of our State Health Care business. I think you already met them, most of you. They will provide the subcommittee with specific examples of the -- the general examples I'll describe which underline our approach of beneficiaries moving into managed care.
CRG was founded in 1984 to serve as a vehicle for developing scientifically rigorous population-based tools for medical management tools. Today CRG is recognized as a leader in the provision of ongoing development in support of such tools in industrial strength that is capable of dealing with the increasingly large and complex data among our clients. Put to you simply, we're in the business of providing tools and services to assist our clients in converting raw data, their raw data into statistical information. That is, information is then usedto measure, monitor and manage medical care delivered across a broad range of sites to define populations often those enrolled in managed care programs.
Despite our name, despite the fact that a great many important findings have come out of our work, we do not view ourselves primarily as a research organization. We do not use data primarily to answer a specific research question. Instead, our products and services organization, display and permit manipulation of data. They are used as monitoring, surveillance and management tools capable of many applications. This epidemiologic surveillance approach often leads to questions we and our clients would not even have thought to ask. Our clients include Blue Cross and Blue Shield plans, commercial health plans, provider sponsor organization, Medicare fiscal intermediaries interested in the detection offraud, waste and abuse and state Medicaid programs. Our Medicaid clients at the present time include the states of Tennessee, Arkansas, Mississippi, Kentucky and Texas. The raw data we use most often is client claims and enrollment data. We help clients organize andclean their raw data and provide the tools and/or services to convert that data into statistical profiles of various types. Data quality is always a concern, but we have learned that the perfect need not be the enemy of the good. We often used datas that was never intended for analytical purposes but was theonly data available. Tolerance thresholds for imperfect data vary with analytical objectives,and the process of analysis itself can be both a powerful tool and a powerful incentive for the improvement of source data.
Unlike patient-based clinical studies, population-based statistical analysis is a relatively forgiving methodology. Our analytic efforts provide status, access to care,resource use and costs for defined subsets of the Medicaid population. The profiles we produce invariably show large variations in the utilization rates and costs in medical care between geographic regions and other subgroups of the total population such as race and socioeconomic status. We help clients do assessments of the intensity and distribution of morbidity among their members, including the blind and disabled and medically needed. We have found great variability even among Medicaid populations that look at first blush to be very similar, such as the TANF, formerly AFDCbeneficiaries. We will provide specific examples of these differences later in thepresentation.
Once statistical information is produced, the ability to drill down progressively even all the way to individuals to the power of our methods. Their utility for the surveillance of medical care depends upon provide practice patterns, resource supply and organization or patient characteristics that systematically drive the large variations in and hospital admission rates and ambulatory care that we observe. Therefore, we are strong advocates of the collection and reporting of highly detailed data, assuming appropriate use of existing safeguard to protect patient confidentiality. We have an example of a minimum data set for the committee's use.
A population-based analytic from all sites and types of service, including butnot limited to hospital inpatient and outpatient, professional, laboratory, x-ray,pharmaceutical, home health, behavioral health, and dental care services. This comprehensive view of medical care is vital to Medicaid programs for three reasons. First, only a population-focused model can detect underutilization by monitoring rates of events. This permits measurement of what is nothappening as well as what does happen second, this population oriented view is more representative of care as it is experienced by program recipients than conventionalprovider-center views and, therefore, is a better measure for evaluating program performance. Finally, the health risk assessments necessary to identify and manage financial risk can only be applied through comprehensive population-based measures. We routinely incorporate clinical grouping and morbidity assessment systems into our work, as long as their methods are scientificallyrigorous and described in peer review articles.
The credibility of the methods used in any analysis of medical care is critical to its acceptance, especially by physicians. We currently use Adjusted Clinical Groups or ACGs and Diagnostic Cross Groups, DCGs, both morbidity assessment tools; Episade Treatment Groups and RVUs, a physician work output measuring tool. In addition, we have contributed importantly to the development of the additional methodologic approaches, such as the concept of ambulatory care, which has proven particularly useful of the indirect assessment for the adequacy of and access to primary care by Medicaid beneficiaries.
In our view, these grouping and morbidity assessing methods are in theirinfancy, but are evolving rapidly and will continue to improve. As that happens, we must incorporate those methods into our tool kit. Ms. Lutz will now provide some examples of the work we're currently doing --
DR. IEZZONI: This is going to have to be quick, Ms. Lutz.
MS. LUTZ: Will be.
DR. IEZZONI: We need to have you miked, microphone.
MS. LUTZ: I'm sorry. Okay. (Discussion held off the record)
DR. IEZZONI: Great.
MS. LUTZ: I will make this very quick. And what I've done is put up fourexamples to show you what I believe is representative of the experience of the states as they transition from service to managed care and, likewise, for the managed care organizations that have entered into agreement to serve this state Medicaid population.
Again, we're stressing the idea that it has to be an eligible essentric model or population-based model as opposed to a patient-centered analysis. Using the eligible composition of the eligible group for the managed care plans, what we've done is create -- taken a systematic approach with our clients to identify, given the encounter data records that are received, what percent of the expected service level are these managed careorganizations supplying. And it does take an empirical approach to quantify that givenfeedback, which is very important to these managed care organizations, you often in that dialogue under -- are able to identify what are the factors that drive this lower than expected service rate, and not unusual that you find that there are potentially some access barriers that do get corrected under service and most typically under reporting. And with that, we see some correction in how the typical managed care provider then supplies those services.
I think it's important to note that the state organizations have to have an ongoingdialogue with these health plans to monitor the quality of that care delivery. The point aboutthat is that it needs to, we believe, take an empirical approach doing that a core set of the data elements must be specified in the contractual requirements to these managed care organizations, and there must be sanctions in place that the state can invoke when the levelof care does not appear to meet the state health plan requirement or guideline.
As we go forward, the science of understanding the health needs of our populations has really come to a point where we can use it to manage the delivery of care. Using an example of our Kentucky client, a state that is in progress of developing a partnership program where there are eight dedicated health delivery networks to serve the Medicaid population. We're using data today and some of these morbidity assessment tools, you can identify or quantify the absolute health status and morbidity of the population that is to be served by that health care network. In this case, we find that there is a decidedly more complex eligible population that's be being served by one of these partnership networks, 18 percent greater complexity in the eligible population.
The experiences that they are currently spending about $450 less than the statewide average and we can project, given their health status, if the system was meeting all of their measurable health service needs, that we expect their spending to be even higher. This is a very, I think, rich example of the kind of position the state can be in today to quantify the health status needs of a population and importantly, appropriately fund that health care network to ensure that the eligibles receive the needed health services.
To reiterate, certainly health status is found to date explain about 40 percentof the variation in resources and claim and encounter data is going to be vital of the ongoing measures of these populations, and we believe that should be coupled with health survey instruments and together be able to better assess when the care level's meeting the state guideline for appropriateness.
Additionally measuring access as we change dramatically, the financial incentives the provider community, it will be in the forefront of the state program to ensure thatthere's access to care, ambulatory care sensitive. Admission rates continue to be a very rich example or rich instrument for measuring the continuity and availability of primary ambulatory care. We see today that there are still a great amount of variation in those ambulatory care sensitive admission rates. With the states that we're currently working with demonstrate a range of 18 admissions per thousand to a high of 45.2.
Interestingly, as each state has some degree of primary care case management,that there may be some correlation with the improving the continuity of care through managed care and it's measured impact on reducing those potentially avoidable hospital admissions. The states we're working with are interested since they use a common data model in sharing information on how the variation in that particular access measure compares from state to state. And one of our states that we're working with is looking at why their admission rate is 45 when other states under that range of 20, and estimating that there's potential for about 18 million dollars in savings in facility costs alone in that state if they can reduce. And they are working toward reducing that admission rate by looking at the continuity of their ambulatory care.
Finally, and I think it really underscores, that while states have a learningcurve that they're working on along with their providers to manage encounter data using an empirical data model, this actually represents the state of Tennessee, that across their MCOs have really gotten comfortable that they are receiving 100 percent of their encounter data taking now some very important quality measures. This is an example of well child visits, what ratio of continually enrolled eligibilities ages three to six receive a well child visit in a 12-month period of time, they have used a HEDIS 30 measure. I think that where they demonstrate the importance of using that information, this information and a subset of 12 HEDIS measures are now publicly available to make available to the eligible population an informed choice about whether or not the provider that they've selected is meeting the quality of care guidelines.
This information, along with information on infant mortality is now publicly available. These reports are on their web site. And I think it really demonstrates thatthere is a lot of pain, but a great deal of gain, but not relinquishing your quest as a state to bring your encounter data up to the level of integrity of our traditional fee-for-service claim data.
DR. IEZZONI: Thank you.
MS. LUTZ: You're welcome.
DR. IEZZONI: Great. Are there any points of confusion that you want clarificationof? We'll have time for more broad questions after Brian has spoken. Anything that you want to ask either Phil or Ms. Lutz? Okay. Great. Brian.
MR. BURWELL: Thank you, Lisa.
DR. IEZZONI: And could you just introduce yourself. (Pause)
MR. BURWELL: My name is Brian Burwell, I am a Vice President at the MEDSTAT Group in Cambridge, Massachusetts here. Lisa and I go back a long long ways.
DR. IEZZONI: I was just telling Carolyn, 19 years, it kind of came out.
MR. BURWELL: Amidst that is a health care information technology company with corporate headquarters in Ann Arbor, Michigan. We do lots of things, but our core skills of development are the integrated health care information technology company. Like the Codman Research Group we also develop and design proprietary analytical tools for looking at the data or decision support products, and we provide consulting around our health care data bases and our software products. We do serve all segments of the health care market; that includes large employers, managed care organizations, insurers, federal government,state government, state Medicaid programs and providers. Our Medicaid clients currentlyinclude states of Massachusetts, Georgia, Montana, Indiana, Georgia, Nebraska, Wyoming, Mississippi and most recently, California. Many of our health care data bases integrate both traditional fee-for-service complaints and encounter records from managed care plans, both in our commercial products and our Medicaid products.
We also provide encounter data technical assistant services to the market. Twoexamples of that; we are currently HCFAs contract to provide encountered data technicals systems to 11 15 waiver states. Our HCFA officer is Racheal Block who is right over to my left. And we were most recently selected to be HCFA's contractor to do data validation for Medicare risk plans submitting data and the Medicaid choices demonstration.
I think the reason we are invited here is that last year we were awarded acontract by the Massachusetts Division of Medical Assistance, the state Medicaid program to build an encountered data base for their Mass. Health program. This data base includesthree major components, it includes the Mass. Health HMO program; also their primary care clinician plan, which still the majority of Medicaid managed care population is enrolled, aswell as their behavioral health program. Their behavioral health program is a carve out for those enrollees who are in the PCC program and those services are managed with a separatevendor in Massachusetts, and our data base will include those data.
Also, on the HMO side, not all Medicaid services are capitated, some are still carved out, the data base will include those services that are provided on a fee-for-servicebasis. Some HMOs do not have, for example, prescription drugs in their contracts, those are still provided fee-for-service and we will get hose fee-for-service claims and integratethem. The claims and encounter records will also be linked with the Medicaid eligibilityfile so we'll have complete eligibility data for people in the data base.
There will be -- we have both -- DMA is both a customer and HCFA is a customer,and part of the contract involves transmitting a minimum data set to the health care financing administration as part of the terms and conditions of Massachusetts 1115 waiver which was just recently approved. We will also be using the this administrative data set togenerate quality indicators for Mass. Health enrollees. And DMA is very interested in the possibility of risk adjusting their capitation payments to their HMO plans. And this data base will be used as an information source for developing those adjusted rates, depending upon where DMA chooses doing it.
The reason we were selected as the contract for DMA is primarily with ourexperience with the Group Insurance Commission, which is the state body that administers health benefits for state employees in the state of Massachusetts. We've had that contract since1993. The Group Insurance Commission contracts with many of the same HMOs that participate the Medicaid Managed Care program, therefore, we have experience in collecting and building data from encounter records from those HMOs. There's not total overlap between GIC's managed care plan and DMAs, but they're a significant overlap, names that are very familiar toeverybody in this room.
I want to talk a little bit about our approach to building data bases that involve a significant amount of encounter records. We take a very proactive approach in working, both with our customers and with the plans in ensuring the quality of the encounter information that we receive and process. We meet with each plan to learn about their information system, capabilities. There have been a lot of transitions in the informationsystem within plans. There have been a lot of mergers and so forth. There are a lot of issues that need to be talked about with plans in regard to their ability to generate encounter information. We develop very detailed data specifications for each plan.
I have a copy of the data specifications we've sent to the Massachusetts plans, that includes all the data elements that we are requesting from, the data field definitions; record layouts; what the media requirements are for transmitting the data; standard coding procedures; and the data quality and standards that we expect plans to adhereto. We then receive test data submissions from each plans, process those data and feedback test data quality reports to each plan, and talk about their ability to comply with the data specifications relative to the standards. Then we proceed with full data submissions and we provide ongoing consultation with each plan.
Each plan submits -- we work with each plan to develop an action plan that addresses their data quality issues and that is submitted -- will be submitted to DMA.
We are currently in the process of receiving test data from each plan. And we have a formal ongoing data quality improvement process that we maintain as we continue to with the plans. Once the full data base is up and going, the data base is refreshedperiodically, but we have an ongoing system in regard to working with plans around that and current data. The process that I believe that we have in place in Massachusetts is a very good one. We work with a number of state Medicaid programs around encounter data issues.
And I just want to talk about some of the common problems that we encounter -- in working with encounter data and state Medicaid programs. Number one, as I've kind of alluded to before, there are system deficiencies in Managed care plans. They're inability to capture encounter data within plans; there are system problems; there are problems generating information. Many states do not have good procedures for processing the encounter data once they receive it. They receive the encounter data and they really don't know whatto do with it. And they can't tell whether it's any good or not. A number of states have triedto modify their MIS systems to be able to process and encounter data but in many cases, that is not done correctly. And MIS systems are not designed in large part to encounter records. A lot of the encounter records will be -- (inaudible) -- which are primarily system for paying fee-for-service claims.
So states have to make a decision to modify in order to deal with current recordsor to -- (inaudible). Many times there are poor data specifications from states to plans aboutwhat the expectations are. Some states say, Well, just send us what you have. It's not the right way to go about it. Communication is key. There must be a key point of contact at each state that's responsible for collecting encounter data from the plans. Even though we're a contractor, we deal with fairly technical issues with each of the plans. But there are obviously bigger issues that we need our state person to get involved in terms of plans meeting deadlines; getting they're data's in on time; negotiations around timeliness, et cetera. One thing that we -- one area we think needs a lot of work is standard processes for validating encounter data that is submitted. We get involved in a lot of encounter data validation projects as a contractor. Medical record review is obviously the gold standard, but it's very costly, takes a lot of time, it's extremely burdensome. It's not really necessary in all cases. There are other ways to validate data. There should be more standard procedures for identifying those points in the data generation and transmittalprocess where encounter data quality problems are more likely to occur, and the interventionsshould be appropriately targeted to where the problems are.
Medical record review is a good idea in some cases, but it should be used in conjunction with other data validation methods. Sometimes states have very unreasonableexpectations of plans and what plans can generate in record to encounter data. There is no formal data quality improvement process. It's kind of an ongoing formal set of procedures to look at data, evaluate it and implement procedures in order to improve its quality. And I -- (Headphone problem) -- how they intend to use the data. And one of the most common complaints you hear from plans is, Why should I give the purchaser all these data when I know it's not going to be used; or, you know, We have no idea why it's going to be used, and the data gets sent and they just sit there. So a very important part of ensuring compliance with data requests on plans is to talk to the plans about how the data will be used.
A few concluding observations. We feel very strongly that person level encounterdata is absolutely essential for effective management of state Medicaid programs. We do not believe that performance measures or aggregate data are sufficient for effective program management. We do think, it's not easy, but it is possible to develop encounter -- Medicaid encounter data systems that have complete and reliable data in them. As I said before, data simply requesting data for data's sake does not work. Purchasers, those who request the data must be very clear before they make the request how they intend to use the data.
Another important aspect is show plans; provide value to the plans from the datacollection process, feedback information about the value of the information that they're submitting to the plans, have them be part of the process. If they see the value of generating encounter data, compliance will go up. A continuous data quality and improvement process on an ongoing basis is absolutely essential. And that there are many opportunities these days for linking Medicaid encounter records with other personal level data such as vital statistics data, maybe even Medicare data for dual eligibles, et cetera, that will provide for further value of these data systems. Thank you.
DR. IEZZONI: Brian, thank you for an exceptionally clear and helpful talk. Are there any quick questions for Brian before we open up the floor? Yes, George.
MR. VAN AMBURG: Brian, you've sent out a proposed data set for Medicaid, encounter data set to the plans. What is the reaction to collecting the same data for other than Medicaid patients, plans that have commercial and other patients, would they collect the same data set are we willing to do that, you think.
MR. BURWELL: The data specifications that we have sent to the Medicaid Managed Care plans are very similar to the data specifications that we send to the same plans for the Group Insurance Commission. I mean, obviously Medicaid and the Group Insurance Commission are two of the largest purchasers in the state of Massachusetts. Now because the population, the group insurance population, the Medicaid population are obviously very different, some of the data elements differ, and we're requesting some data elements that are more specific to the Medicaid population. But we try -- we have tried to minimize, tried to make the data generation process a similar process in order so they can use the same programs to generate the information.
DR. IEZZONI: Is it possible for us to get a copy of what you've sent the providers or is that confidential?
MR. BURWELL: To the managed care plans?
DR. IEZZONI: To managed care plans.
MR. BURWELL: That's up to the state of Massachusetts. I don't think that they would have a problem. They'll be here tomorrow. Our project officer for DMA is Tony Asciutto who's on the agenda for tomorrow.
DR. IEZZONI: Okay. Carolyn, you'll check into that?
MS. RIMES: Yes.
DR. IEZZONI: Okay. Great. I'd like to open up the floor for questions to Brian, Phil and Jack. Members of the committee George.
MR. VAN AMBURG: This is for Phil. You showed the slide that showed the variation of 18 to 45 in hospitalizations for sensitive conditions. What is a variation in those plans as far as encounter data versus fee-for-service data? I assumed there was a -- MR. CAPER: Let me turn that over to Manon or perhaps Jo Ann to answer.
DR. IEZZONI: Whoever answers has to come to the mike. Thank you.
MS. LUTZ: While Manon puts up the slide, I will note there was plan A through 8; A had a rate of 18, that was 100 percent encounter data because it is the state of Tennessee. The other four states have varying degrees of managed care, none of those other four states are currently capitated. In other words, the primary care x-rays management models are submitting fee-for-service data. Tennessee's rate of 18.5 has been validated by using the Department of Vital Statistics data and they use the birth records -- they bring in the birthrecords which have allowed them to do the infant mortality study, and they found that they had about 99.5 percent of the birth in their encounter inpatient data base. That and other measures have led them to feel very confident that that rate of 18, while it's lower and is coming purely out of encounter data is validated in terms of having what one would consider to be as close to 100 percent capture. Does that answer your question?
In other words, I assume that your concern was that -- and you notice that there are increasing levels of managed care with the lower rates. I think that it's -- this is,again, I think fairly public data. State B is the star program in the state of Texas, which is100 percent PCCM, and their rate is low. Kentucky is state C, they have the KENPAC program which is a high degree of primary care case management, about 60 percent of their total Medicaid population. What we find -- and I think it is a very important point about cross-state comparisons in Medicaid, Mississippi, for example, and Arkansas, state D is Arkansas and Mississippi. I think that the eligibility characteristics of our Medicaid programs must be taken into account when we look at comparisons, the socioeconomics of the Medicaid population.
At the Codman Research Group, we certainly found that morbidity and health status are directly correlating with the levels of poverty. And I believe that there is a fair amount of variation in terms of the economic, socioeconomics of these Medicaid populationsand, in part, that's driving the Medicaid's population access behaviors.
And, likewise, I think that it's important to note that the more rural the constituency of a Medicaid program is we also find higher ambulatory rates. But what we've done is we've broken down these rates by age, sex, cohorts for the state to begin to cut into what's driving this variation. And the literature continues to support the fact that Medicaid populations or insurance is a factor in access. In the most recent issue of Health Affairs is supporting that we're seeing this level of variation.
DR. IEZZONI: Thank you. George, following up on that?
MR. VAN AMBURG: I have a follow-up on that. If you run a similar analysis for referral sensitive surgeries and what would the data show?
DR. IEZZONI: Can you --
MS. RUBEN: There is a lot less variation in the referral sensitive surgeries. In general, the variation in overall surgery rates is probably 40 to 50 percent what it is inmedical admissions. The ambulatory care sensitive component in the medical admissions is the most variable component. So when you compare Medicaid utilization with privately insured commercial populations in the same age groups, it is the ambulatory care sensitive admissions which are most different from a commercially insured population. So you wouldn't see anything more than maybe a 30 percent variation in the surgical rate. We also track the so-called marker conditions.
MR. VAN AMBURG: Okay.
MS. RUBEN: There's no difference in the hip fracture rate among these states, which is another way also of checking the integrity of the data that we're receiving.
MR. VAN AMBURG: Okay. Did you use all the ACSs or a subset?
MS. RUBEN: All of them. And in the -- there's an attachment to Phil's testimony that lists the ACS conditions that we aggregated for this. But the majority of them are what we call the rapid onset, which are largely pediatric and the chronic conditions.
MR. VAN AMBURG: Okay. Thank you.
DR. IEZZONI: Hortensia, has just joined us. Could you introduce yourself and you have a question.
DR. AMARO: Hortensia Amaro, Professor at the Boston University School of Public Health. We've been very interested in the collection of race and ethnicity data. And I was wondering whether you're collecting that as part of the work you're doing and whether you'redoing any analysis to look at issues of access and other measures of quality of care by raceand ethnicity?
MR. BURWELL: The answer is, yes, in some states. States have kind of different policies in regard to whether they actually correct race and ethnicity data during the eligibility process. Some states believe that that's kind of confidential information and do not capture that data at all, so it is not in their systems. Other states believe it is part of the Medicaid program administration. So, for example, the Medicaid -- the system that we are developing for the state of California definitely captures race and ethnicity data. You know, one of their major goals in terms of program analysis is to look at Medicaid utilization rates by those variables.
DR. AMARO: In the Massachusetts contract, is that included? Does the Massachusetts collect --
MR. BURWELL: We -- is race on the eligibility files? We have not yet linked inwith the eligibility files. I -- you should ask that question -- I don't know. I don't know the answer to that. It's a good question to ask the Massachusetts people. Maybe Dan knows.
DR. IEZZONI: Paul, and then Dan.
DR. NEWACHECK: I have a question for Jack about the administration of the CAHP Medicaid populations. The three modes that you're going to try in terms of administration of mail first, mail plus telephone, mail plus telephone plus personal interviewing --
MR. FOWLER: Right.
DR. NEWACHECK: -- do you have any response for these three modes? And, also, do you have expectations or at least sort of a target that you're looking for in terms of what would be an acceptable response rate? And I realize this is sort of speculation because you're going to be doing experiments.
MR. FOWLER: We do have experience in other states in Medicaid populations and there's a range. I believe New Jersey had somewhere around 20 percent, which a lot of people would consider as pretty low. Another experiment in -- there was at least another one -- it might have been California that ran -- did it? I think they were in the low 40s by mail alone. And we actually collected some data from Massachusetts by mail alone earlier which was sort of modeled by the NCQA instrument, and that was 38 percent.
So our guess is in Massachusetts will be in the 40 plus or minus a little bit bymail alone. I would expect we'll pick that up in 15 to 20 percent points by telephone. Andthen I don't know how far the rest will go because no one, to my knowledge, has tried to go out and look for the balances of population in person. I'm certainly hoping that we'll atleast get another 13 percent and get it well up into the 70s, but I'm not sure that that's what will happen.
In terms of your minimum standards, actually, one of the things that has happened in an attempt to getting a consistent instrument that's used in lots of settings, NCQA which is in the CAHPS process, have converged, I believe is the word with quite a bit of effort. And one of the things we've tried to converge on is minimum standards for response rates. There wasquite a bit of haggling about that, but the number that was chosen as a minimum was 60. And that -- and so NCQA presumably, and we'll say the data that don't achieve that are not -- have not met minimum standards. CAHPS is not in the minimum standard business, they're in the recommending business. But 60 percent has been our target as -- again, as a minimum. And a lot of us are hoping that we would like data even higher than that. So that's the ballparkin which we're hoping we're operating.
DR. NEWACHECK: Can I ask a follow-up question?
DR. IEZZONI: Absolutely.
DR. NEWACHECK: That's very helpful. By the way I've heard that NCQA has moved 50 percent.
MR. FOWLER: I have lots of data, but I think for 1999, it's going to change.
DR. NEWACHECK: Do you have any sense about the relative costs of those different modes, like going from mail to mail plus telephone to mail plus telephone and personal?
MR. FOWLER: I do. The inperson part is almost impossible to generalize about because it depends so much on the degree graphic spread, and what you've got by way of staff. But certainly, using a dual mode methodology, which is the one that is sort of that people are working with is being possible. The preferred mode which would be mail with telephone efforts to interview nonrespondents, and that probably gets you to 60 percent with some population. So everyone's not sure you'll get it with the Medicaid populations that way. Those extra telephone interviews probably cost twice as much per unit as the mail responses. That's kind of vague, I know, but -- but that's about as good as I could get.
DR. NEWACHECK: Thank you.
DR. IEZZONI: Dan, did you have a question?
MR. FRIEDMAN: Yes. Brian, you mentioned opportunities for linkages between the Medicaid encounter data and other data sets, and you mentioned birth files. Are there files that you've done linkages with or that you regard as useful -- potentially useful to link?
MR. BURWELL: I am, in candor, not familiar with all of the data linkages that have occurred in our Medicaid systems. We have linked with vital statistics records I know inat least one state. I believe there may be initiative to link with other disease registriesin other states. And obviously the opportunity to link with third-party liability, the opportunity to link with Medicare claims and encounter records for those people that are enrolled in both programs I think has a lot of potential, particularly as dual eligibles seem to be the next population targeted for managed care.
DR. IEZZONI: Ms. Lutz, you wanted to follow?
MS. LUTZ: If I could just make a comment. In the state of Kentucky, we've just established their data base to improve the vital statistics, the Department of Mental Health. So their bringing in more than just Medicaid as a pair and there is -- they're all pair dischargerecords to that the Medicaid will be in view as the same data model as a subset of the statewideresidents. Dan.
MR. BURWELL: Which also sparks, we also in Massachusetts, part of the data base will include nonMedicaid eligible served by the Department of Mental Health.
DR. IEZZONI: Dan, you looked as if you had a follow-up.
MR. FRIEDMAN: Do you happen to know, is that done in Tennessee on a blinded individual record basis or individual clients?
MS. LUTZ: In Tennessee currently they have only brought in the information from the Department of Vital Statistics, the births. You're talking about in terms of the eligible identification?
MR. FRIEDMAN: Yes.
MS. LUTZ: I know that they tied the mother's birth record at the ID level, which was one of the real significant as tests to see whether or not the encounter records compared to be complete and that was done at the ID level.
MS. RUBEN: Don't they also bring in birth weight?
MS. LUTZ: Right. Correct.
MS. RUBEN: They also bring in new birth rate on the side with the same linkage.
MS. LUTZ: Right.
DR. IEZZONI: George.
MR. VAN AMBURG: I have a follow-up question for Jack on the surveys. We heard out in Arizona when we were there that the plans was doing survey, the state was doing surveys some independent people were doing surveys of the clients and the plans were getting surveyed out. And it's pretty well known that response rates for surveys are going down, whether they're mail surveys or telephone surveys, for all types of surveys. Can you comment on whatyou're finding on that and perhaps what might be some solutions.
MR. FOWLER: That's been a concern about oversurveying populations has been around for a long time. And I've never known quite what to do because it's hard to get comparable data collection efforts and comparable demand levels of interest, et cetera. I have to say that in our experience with the CAHPS surveys, with members of plans, that refusals are pretty rare when we get on the phones. For example, so if we can reach them, one of the things, plan records are not that great so it's not uncommon to loose 10 or 15 percent because you can't get a decent telephone number or somehow reach people. But it's been our experience, and we've done quite a few, we did Washington State employees in 20 different plans, some Colorado people, et cetera. Mail is only a fair way to enlist cooperations, it's not a great way to do it.
And I think the plan experiences for surveying people have been in the -- with CAHPS instruments have been 40 to 55 percent, I think is the range that they've gotten with mailalone, with multiple mail, but not going beyond. But I would say, if you add telephone,refusals are pretty rare. And we've certainly gotten experiences going up to in the 60s and 70s with telephone and privately insured people. So I don't know quite what to say.
Medicaid populations have been more resistent. They're tougher for all reasons. There are more challenges in terms of the data collection modes which may not feel as comfortable to a lot of the folks on Medicaid to use, and their motivation may be different. So, that's not a very good answer in one sense because it doesn't say, is it harder or not. I think surveys of plan members are quite feasible, particularly when you sample people who have a condition, which occasionally people want to know the experiences of people with particular kinds of situations, or have more chronic conditions. Their level of response is distinctively good. And I think the issue that, you know, if you're really doing something that is relevant to your well-being, you're really giving feedback on a plan that you belong to, public, et cetera, feedback. I think people are discriminating on the request being made on their time and stuff. Other people might not be as sanguine. But I think it will do quite good if it's done well.
DR. AMARO: I had a follow-up question on the race and ethnicity. So it seems like you vary by state and experience. I was wondering if you could elaborate a little bit more on how -- for the states that do collect it, how you're using it and -- how you're using it in your analysis and your information back to the health plans.
MR. BURWELL: The business of developing Medicaid decision support systems is a relatively new business. And a number of our data systems are still in the process ofinvestments, as in California. I think one important point that needs to be made is that wedeveloped decision support tools for end users within our customer base, so our -- the nature of our business is to develop the data base, provide analytical tools and then support our customers in their analysis of the data. In many cases, that does end up doing the analytical work ourselves. But it's both a process where we provide the data, we provide the tools, and end users within state Medicaidprograms are able to look at the data themselves with our tools.
DR. AMARO: Okay. Then let me ask you this; have any of your users conducted analysis on race and ethnicity, you know, as a way of looking at issues of equity and access across populations? I guess we're trying to get at, you know, are these variables being used and not just collected but used in any way across the board to look at issues of access and --
MR. BURWELL: I believe there are. I can't provide you with specifics. I'd be glad to follow-up with you and -- about specific analyses that are done on race. I certainly believe in the state of Mississippi we have done and the customer has done a number of analysesin regard to access and looking at race as variable.
DR. AMARO: That would be helpful if we could get examples of how people used it.
MS. RUBEN: In the five states that are shown above, all of them have done analyses based on race, ethnicity and income qualification. There are a little bit of variation among them, but all of them at a minimum are covering that as well, rural, urban.
MS. LUTZ: And I would just like to add a point that in a fully capitated managed care circumstance like Tennessee, each of the 12 MCOs are also segmented by the dominant ethnic category as well as age category. It's been important to answer to the advocacy groups. And the key advocacy groups that have certainly shown themselves in that state are women's health care, some of the ethnic groups and the add advocacy groups for the poor. And we've used the median income for the eligibles in a zip code as well as race to demonstrate that there is sufficient access to care.
And I think that the other point to make, especially in a state like Texas where youhave high, and California would be that way, too, where ethnicity such as the Hispanic population, you have to be able to advertize a part of that data. Whether or not it's provider discretion in terms of some of the service rates or whether it's really driven by the ethnic composition. I would say that diabetes, to understand your diabetic population, ethnicity is a very important factor as well as hepatolic colecysectomy [phonetic] rates, you've got to be able to look at what population of that rate is Hispanic origin.
Likewise, managing asthma in our Medicaid programs. And the urban world differences and the median income of the zip code really is an indicator, zip code of the kind of environments these kids are living in and can help target programs. I really believe that managing with these population-based methods today offers a significant public health opportunity to the Medicaid programs that have moved away from becoming simply fiscal programs and really -- and I think for the states it's been a real, you know, sort of a wake-up call in the past few years. If they move into managed care, they can become health care delivery programs. And I think that these kinds of handles in the data sets that we've produced of encounter data become really an instrument to do what the Medicaid programs have tried to do a for a number of years.
DR. IEZZONI: Kathy.
MS. COLTIN: Just a follow-up to that. Do you know whether in Tennessee or any of the other states that are using the racial and ethnic data, whether they are sharing those data elements back with the health plans?
MS. LUTZ: Absolutely. Absolutely. The health plans receive their reports coming out of the systems understanding how they are.
MS. COLTIN: I don't mean the reports, I mean the raw data on their population. Inother words, as part of the enrollment data that comes to the health plan --
MS. LUTZ: Right.
MS. COLTIN: -- from the state.
MS. LUTZ: Right.
MS. COLTIN: Do the health plans receive information about the race and ethnicity of the enrollees?
MS. LUTZ: That's a question I can't answer.
MS. COLTIN: Because it often happens within the health plans.
MS. LUTZ: Right. And I can't answer that.
MR. CAPER: Of course the health plans -- it's not so much the job of the stateas it is with the health plans in some ways. The health plans should be --
MS. COLTIN: It really -- (inaudible).
MS. LUTZ: That's correct. And I can't answer.
MR. BURWELL: In regard to Massachusetts, I think that would be a good question to ask Massachusetts. I think that the plans have access to the eligibility system in order to basically track who's in their plan and who's not in their plan. And whether they have full access to all the data elements on the eligibility system, or only a subset of those, I don't know.
DR. IEZZONI: I do have a couple of questions. I've been very quiet here letting my colleagues go, but I wanted to check because I noticed we only have one of our panelists for the next section in at her seat, Wendy Mariner. Are there any people in the audience who are on our next panel? Okay. One person. Okay. Why don't -- both people. Okay. Great. Can I just ask a final question then, a quick question. And this can be addressed by all of you.
I'm interested in people with disabilities who are among the most needy and high utilization of the Medicaid beneficiaries. Jack, especially for you, how will the CAHPS instruments be able to get information from people who may have cognitive impairments, other impairments that may get difficult for them to respond effectively to these kinds of surveys.
How will the CAHPS process make sure that these people are not forgotten by the state as their survey and their Medicaid beneficiaries.
And then for Brian and Phil, the kind of information that you get on claims, ICD9CM codes are notoriously bad at identifying disabilities. There is a V code for wheelchair use, but there isn't that much more that gives you a sense of what patients actually looklike. Have you explicitly looked at Medicaid data for people who are eligible for Medicaid due to disability? And what kind of scenarios can you fabricate basically using the information? Jack.
MR. FOWLER: Actually, I'm going to answer your first question and answer one question that you didn't ask that I forgot to mention. In terms of people who can't speak for themselves or write, that's a -- we do not, at the moment, have a special strategy for that any more than we have for people that are elderly. And the proxy approach is all the researchers know how to do, and that is used in a moderate amount. But we don't have a special -- that's a problem we need to address in a more systematic way and I think it's a good one. And the one I thought you were going to ask as part of our testing, what wesaid, can we identify -- (inaudible) -- that we could test our questions on to sort of see how well they seem to differentiate and capture what's going on. It terms out, of course, you can get illegible for Mass. Health in a variety of ways. And if you become eligible by virtue of your income category they don't classify you as being disabled. So it turns out there's a set of people who may have disabilities that are not so labeled in the administrative files because they're eligible virtue of something else.
One of the little pieces we're doing, along with others, is at least we'regoing to be able to estimate the rate at which the cross-section of AFDC-type eligible peoplehave disabilities and figure out what that rate is, and to the extent you need to concern about capturing those experiences as well. But finding how to sample people is a problem, if monitoring how they're being managed is one of your concerns.
DR. IEZZONI: Okay. Quickly Phil and Brian --
MR. CAPER: I'm going to ask Manon who's had a lot of experience with.
MS. RUBEN: Increasingly we're able to get through the Medicaid program access to -- I guess it's the Social Security file that contains the information on what specifically qualified them for the disability. That is, in the standard Medicaid eligibility file, you have an eligible -- you have an aid category and it's under disabled. But now in at least two of these states, we have -- they have facilitated access to the file that actually established what was the nature of their disability and the level of impairment that caused them to be qualified for that aid category and that enhances it.
The other thing that we do though -- you're absolutely right about the ICD9s. And we haven't really perfected this, but we're making progress, is that we couple the aid category, the ICD9 codes and the specific service or procedure categories that indicate the assistance that they are receiving. Sometimes it's as far as ventilator support or something like that. In other cases, it is DME or physical assistance through therapists, et cetera.
MR. BURWELL: For the federal government for ASPI, I have done, in fact, Ijust completed a study of children with severe disabilities in which we linked SSI disabilitydata with Medicaid claims data. However, on an ongoing basis, I'm somewhat skeptical of suchlinkages due to confidentiality concerns within the Social Security Administration.
DR. IEZZONI: I was just going to raise that point. Am I interrupting you, Brian? Do you have something more to say? Because I am concerned about what the Codman folks just said. Because, were you getting that information from Social Security under a program specifically for Social Security?
MS. RUBEN: No. No.
DR. IEZZONI: Well, why -- I mean, I'm concerned about them giving a proprietary organization --
MS. RUBEN: And there's no clinical detail in what we have received. It's simplyinstead of saying; A category, disabled, it has about 15 categories that say whether the disability is mental or whatever, and that's the extent of it.
DR. IEZZONI: Sorry, but --
MS. RUBEN: It doesn't have any other clinical information. And I understood -- and I don't know if it's the SSI file. I don't know where it comes from. It came through theMedicaid program with the eligibility data.
DR. IEZZONI: Brian, do you have anything to follow up on that?
MR. BURWELL: It would most likely come from the Medicaid program because they do receive files from the Social Security Administration in terms of the eligibility process in terms of enrollment from the SSI folks. But in regard to building information systems that involve Social Security information, I do think that's a separate issue. And my -- this is a federal question, I believe that those -- the Social Security Administration has tightened its policies in regard to sharing administrative data sets.
DR. IEZZONI: We don't actually have time to talk much more about this. We havesomebody from the Attorney General's office on the next panel, and this might be an interestingsegment. Brian, just quickly.
MR. BURWELL: Just to go back to your original question. In terms of opportunities to getting more robust information on Medicaid recipients with disability, I do think that weare entering a phase in which state Medicaid programs very much would like to be able to do that. And I know a number of states have considered the possibility, for example, a link with MDS data which is collected routinely on all nursing home Medicaid, nursing homerecipients. And I know a number of states have considered using the same instruments for their home base care waiver program so that they have a similar set of information not only for their nursing home population but for the home care operation.
DR. IEZZONI: Phil.
MR. CAPER: Lisa, could I just make a comment?
DR. IEZZONI: Yes, quickly.
MR. CAPER: As you know, there's federal information pending that deals with that. And this is a debate that's going to go on forever. But, you know, it's -- theiravailability of data to its sort. I know that we can demonstrate it to be used beneficially. It can also be used for discriminatory purposes. And, in fact, you know for many yearsin this country there have been data bases which have been used by insurance companies and othersfor discriminating. That's what insurance companies do, that's their business.
I think the reason it's now coming out is that we are using it for other purposeswhich highlights that issue. But I think to think of these issues in civil rights terms, in isolation is a mistake. I think that there are other considerations, and I've urged my friendson the hill and others to consider those issues as well. And I think we have to be careful about the way we use data. But just because you can use a hammer either to build a house or to bludgeon somebody to death, doesn't mean you ought to ou tlaw hammers. That's --
DR. IEZZONI: Okay. Phil I, thank you. I think that our committee actually wasasked by the health insurance portability and accountability act to advise the secretary onprivacy and confidentiality issues. So this is something that we've certainly discussed backwards and forwards. I just kind of laughed when I heard, you know, that there was information about disabilities from individuals coming from Social Security. But we are movingon to our next panel. The panelists from this morning are welcome to stay or you're welcome tonot. But we really thank you for a very informative morning.
MR. CAPER: Thanks for the opportunity.
DR. IEZZONI: Yes. Thank you. (Discussion held off the record)
DR. IEZZONI: We're not taking a formal break because we actually are a little late. So can the committee members get back, to the extent that you can, around the table so we can hear from Barbara Fain from the Attorney General's office. So we are actually going tobe starting again now. So if we could have quiet. Barbara Fain for the Attorney General's office. Could you just introduce yourself, say what role you play there, then we'd like to hear your comments.
MS. FAIN: My names Barbara's Fain, I'm in the Regulated Industries Division of the Public Detection Bureau. But certainly what we do in the Regulated Industries Division is we do both utilities and the insurance areas. My particular area is health insurance and areas asspecific to managed care, long-term care.
DR. IEZZONI: Could we have your comments. We're having each person introduce themselves at the beginning of their comments.
MS. FAIN: Let me just say when I was invited to participate, I did mention that the Attorney General's office has a very limited role in Medicaid Managed Care and particularlyin data collections. So I'm hoping that my comments will have some relevance to the work that you're doing. And please feel free to interrupt me. And if there is an area that I'm gosling over that you do want to hear something more about, I'd be happy to -- I would be happyto address that. And, likewise, if I'm digressing into areas that have absolutely no -- I don't want to waste your time.
So, basically let me give you a quick overview of what the role of the AttorneyGeneral is in Massachusetts, because not everyone is aware of all of the work that we do and what our mandate is. Basically Scott Harshbarger is the Attorney General, he is thechief law enforcement officer in Massachusetts. And basically the Attorney Generals role is to represent the Commonwealth in all of the legal matters in which the Commonwealth is party or has an interest. And also the Attorney General has a broad responsibility to protect the interests of Massachusetts consumers. So these are very vast diverse responsibilities.
There are over 200 Assistant Attorney Generals in my office and we're divided into six bureaus, lots of different divisions. We do many different things from criminal prosecutions to representing the state when the state is a defendant in civil actions. We havea family and community crimes bureau that deals with issues of children, elderly, disabled,domestic violence and oversees a victim compensation in an assistant division. We have a public charities division which regulates all the public charities in the state. We enforce the state Consumer Protection Act which -- and as part of that, operate a consumer protection hot line.
And as I mentioned, the regulated industries division represents, it represents rate payers in -- before regulatory bodies in regulatory proceedings, auto and healthinsurance proceedings. And the goal in the rate proceedings is to keep rates affordable for consumers. We also, in the regulated division, enforce state laws that protect consumers from unfair and perceptive insurance practices. My work, again, deals exclusively with health insurance.
So I know you're not interested in utilities, and I'm glad that you're not because I probably couldn't answer too many questions about that. The role of the Attorney General in health insurance, in particular, I think is somewhat unique in this country. For the pastseven years, health insurance issues of coverage issues have been a very high priority issue for the Attorney General. And that's not true in most -- in most states. Most states, Attorney Generals do not get involved in the, health insurance area. Although we do get calls from different offices around the country and some of them are beginning to. But it's an area that ispart of the Attorney Generals mandate to protect consumers' interest.
Obviously health care and health insurance issues are very important and probablysome of the most important issues that are facing consumers these days. So the role of the Attorney General in health care, what have we been doing; we have been a key player in securing passage of several pieces of legislation, state legislation that are designed to decrease access to health care and health insurance for undermedically cared organizations. We scrutinize to poor profit institutions to protect the charitable proceeds and the community stakes in these resources.
There's a specific -- we are required to do that under state law. We have established community benefits guidelines for both hospitals and HMOs. And if anyone's interested, I have copies of the hospital guidelines, the HMO guidelines and also the reports, this year's reports which have been issued recently to reporting on the efforts of hospitals and HMOs and providing community benefits.
We also, with the former Weld and Cellucci administration, have launched a bipartisan task force for long-term care financing, and the goal of that task force is toexamine ways in which state government can encourage the development of better options for the private financing of long-term care. And the objective is to try to help elders avoidimpoverishment if they need long-term care services, and also to relieve the strain on the state Medicaid budget. We've worked very closely with the division of medical assistance, our Medicaid division on that.
We, in the regulated industries division, operate an insurance hot line for consumers, and we've acted in an unofficial ombudsman role to resolve formal complaints that we receive from consumers about their health insurance. And we -- statistically, we get probably about 450 formal written complaints a year regarding health insurance that includes -- it includes managed care, it includes indemnity, it includes pretty much anything. We also receive about 300 calls a month regarding health insurance. We don't classify these as complaints, we consider them inquiries. Some of them result in formal written complaints. But it's -- we have a staff of paralegals that are there to answer questions and gets very involved in mediating these complaints, and we -- ultimately we, if necessary, we could litigate some of these. We never -- I don't -- in the past several years, we have never litigated -- ended up litigating any of these. We've always been able to resolve the complaints favorably just through mediation.
Because I was coming here, I asked the person who runs the hot line to take a look at the Medicaid complaints just to see if we had received any, and there were about ten of them.
And we don't classify them. And this might not be totally accurate, but ten complaints Medicaid beneficiaries over the past two years. So they tend not to come to us. We don't know whether they go to the Division of Medical Assistance or whether they don't go to any stateagency when they are having difficulties obtaining health care. These complaints are -- and the work we do through the hot line, I think, have provided us with some insider prospective on the issues that are facing consumers in health insurance and managed care specifically.
And in what we've concluded is that these concerns fall into a few differentcategories; one, we would call rationing. A lot of the gist of a lot of the complaints wereceive from consumers, although they're wide ranging, there's a common thread, is this fearthat managed care organizations are somehow motivated by profit to ration their care. Andthere's a sense that there's something they should be entitled to that they're not receivingbecause of this profit motive and that they are, therefore, being denied medically necessary services. And I think what this really reflects is a lack of consumer confidence, lack of trust. Many fears are probably unfounded, some of them are probably real. But that's what's going on. And it's bad for consumers and it's also, I think, a lack of consumer confidence makes it difficult for government officials to deal with the real issues in a constructive way.
Another second concern or issue would fall into the category of discontinuity ofcare, or lack of choice, or lack of control over who provides your medical services. Manyconsumers voice fears over losing access to certain providers, particularly those with whomthey have long-standing relationships. They feel that they're at the mercy of health planswho alter their networks. And employers, if they are receiving their coverage through healthinsurance plans, switched health plans year after year to get better rates. And thisconcern is obviously particularly strong in the case of those with chronic or serious medical conditions. And many of these people assume that they have all of this power and control under a fee-for-service medicine, which may or may not have been true, but that is the feelingthat's out there among consumers.
I think, in general, the third and final sort of big concern is just there's achange perceived and real changes in the health care landscape. A lot of consumers are just frankly confused, they don't know how to navigate the system. If they have a problem, they don't know where to start to get it addressed in a logical way. And so I would say, probably most of the complaints we receive, it's really simply a matter of either directing people into the right -- into the system asconnecting them with the right people or in some cases, making the connections ourselves and navigating it for them. And I think, as I said, we don't receive many complaints from Medicaid beneficiaries, but I think these concerns are probably similar to the concerns that Medicaid beneficiaries would have; that they are moved from a fee-for-service into managed care plans.
In response to the complaints that we've been receiving regarding managed care or analysis, we did a more formal study of the issue a couple of years ago and last year came out with a report on the Attorney Generals report, managed care in Massachusetts. And I did bring copies, and I see that you have them. And basically, it's -- the report attempts to get beyond sort of the anecdotal information that we have been receiving about managed care.
And it makes a series of, I think, detailed recommendations for state regulation of managed care organizations. And the approach that we took, and we feel pretty strongly about is that the appropriate state regulation, this area not be sort of body part by body part micro-management that you have when you have individual mandated benefits, pieces of legislation, but more regulatory oversight to make sure that there are standards that are followed and really to preserve, I think, what in Massachusetts it's a pretty -- compared to many other states anyway, a pretty reasonable responsible managed care industry that comes up from sort of a different tradition.
Most the large ones are still not for profit, they've been around for a long time. There's a sort of like a culture of doing health maintenance and preventative care, which I think was the original motivation more than just strictly cost containment, and they operate with very wide networks of providers. Unlike in many states where a lot of people who are in managed care plans are limited to very small networks and providers. So, the idea is to try to by -- through state legislation, through oversight, create a level playing field that will preserve what's good about managed care in Massachusetts and also be something there that can address when more problems do arise.
And one thing I would mention is that most -- many of you are probably aware of ARISA and ARISA preemption, any state regulations that we would do would only apply to people in plans that are not ARISA plans, which is basically probably estimated to be about half of Massachusetts insureds would be, or in plans that would be subject to state regulation; the other half, the state would be preempted from regulating. So I can -- I don't know whether it's worthwhile for me to go through some of the recommendations specifically or I'd be happy just to --
DR. IEZZONI: Now why don't we move on, and I think we'll probably have some questions for you after we hear from the next speakers. Thank you very much. Wendy Mariner.
MS. MARINER: Hi. I guess this will do. Hi, I'm Wendy Mariner, I'm Professor of Health Law, Boston University, School of public Health and I'm very pleased to be here. Although I really wasn't sure why, at least why I'm here, I certainly know why I'm pleased. But, I am delighted. And it occurs to me that there may be something I might add. There's been a lot of very productive discussion on methodology. And I think we are privileged tobe in a state where a lot of the research and work on methodology and data collection is at the cutting edge. But the methodology problems can be so complex and so difficult to implement that they sometimes overshadow the reason the data's being collected in the first place. So I think that I may be here, as Barbara has begun to remind you of what you already know and thatis, why this information is being collected. And I would like to raise just a couple ofpoints that may assist in thinking about this.
There are sort of several uses of the information; beneficiaries in obviously need information about health plans and care in order to choose a plan now if they are going into managed care. And in that they act like consumers, but they also need information to use a health plan. And in that, they act like patients. Medicaid needs information to choose health plans with which to help to contract as appropriate for its population, so they act like buyers and like employers in many respects in the marketplace. And Medicaid also needs information to evaluate the plans to determine whether or not it -- as a contractor and as, in essence, representative and protector of the people. It's carrying out its statutory duties to the beneficiaries because, of course, even when providing and the provision of delivery of care is contracted out, unlike many employers, you can't delegate ultimate legal responsibility for what happens to the beneficiary. You remain responsible for ultimately for the provision of the care and the carrying out of that duty, which I think allows Medicaid and Medicare, in general, to the extent that these are coordinated, so much the better, to essentially set a standard for not merely the collection of information but the use of information to improve the health and welfare of the population.
And as it has been done with financial standards and first with introduction of DRGs and the various kinds of things can become essentially a gold standard for health plans that can spill over to the rest of the population thereby making everyone's lives more consistent, more informed and obviously data collection easier because it can be coordinated and collected. I think that helps, obviously with ARISA, my favorite topic. So, people are wondering why I taught that many, many years ago, but now they know you don't have to worry about ARISA in Medicaid because you can set standards for an entire population without having to worry about ARISA preemption limitations.
Let me just make a couple of points about the uses of information and the reasons for collecting it. And what I really want to focus on is that there are differences in information collected for consumers and information collected for patients. In the several years, we have seen an enormous amount of focus on consumers; consumer information, consumer protection, we have consumer bills of rights being written up right and left. And we have all shifted our terminology from patients to members, enrollees, and now to consumers, so all the patients are consumers. And I worry sometimes that that will sometimes overwhelm the notion that we have patients to deal with as well. And I think there are different kinds of information that are needed for consumers than there are for patients. And in Medicaid you have an opportunity to focus on both because there are clear and useful elements of both in the patient population.
Consumers are theoretically people who make purchasing decisions that are informed in an equal bargaining relationship with another entity and, therefore, are presumed to be able to use knowledge to make choices. That can work with many Medicaid beneficiaries who are given sufficient information about health plans to make decisions on the basis of acceptable levels of quality of care and is the product, in a sense, reliable. Is the service warranty any good. When they go back with problems; can they get service, can they get responses, for example, in a grievance mechanism. So there's information disclosure and choice to provide information for consumers.
But as the law has learned through hard experience, consumer choice is not always sufficient and there has to be concern kinds of mechanisms put in place where the ideal of micro-economic theory where purchases by consumers is not working properly so we have consumer protection laws. And those are designed to ensure that consumers are adequately informed that certain kinds of inappropriate contractual provisions are not included as against public policy.
There's regulation on marketing which is something I haven't heard described yet this morning. The kinds of merging information that can be given to Medicare in marketing their advertising practices and certain requirements for the product itself. Many of these kinds of consumer protections are already in place with Medicaid by virtue of defining the benefit package by defining certain limitations on marketing so they are, in fact, better protected than most consumers. But the kind of information about health plans and how they function could be perhaps improved and expanded upon and provided to beneficiaries.
There's a second element, and that is information necessary for patients which we don't hear discussed as much anymore but is obviously important. Because whether or not you get to choose your health plan, you will be a patient, everybody's a patient at one time or another. And the relationship between a patient and a provider and a patient in health plan is fundamentally different than a relationship between a consumer choosing a health plan. Because a patient is someone who is not in necessarily an equal relationship bargaining, but in a relationship of unequal skill and knowledge. For a provider for that reason, have made rights for patients.
I mention this because the kinds of information that patients may need and Medicaid will need for evaluating whether patients' rights are protected and whether patients are adequately taken care of is different. That has to do more fundamentally with questions of quality, with questions of the type of care provided, personal care levels, the kinds of quality preventing caring for and monitoring care.
Now, as Barbara has mentioned, things that patients are concerned about when they're no longer consumers but patients is worry and fear that somehow by going into managed care, they're going to be deprived of some type of care that they want or need. And one concern in collecting information is that one of the primary things you're going to be looking for is, what didn't happen, things that aren't there. And that's not the kinds of data that is easily collected out of annual reports. I think you also need -- you will clearly need to look for information that is not collected ordinarily as part -- yet. I know there's recognition of the concern, but I don't see it collected as easily yet. There is not part, for example, NCQA, accreditation standards.
I teach, of course, and in academics there's always the fear that standardized exams will limit the information that you can provide students because you may be forced to teach to the exam. If your students are going to be judged on a basis of a standardized exam, you want to prepare them to do well on the exam. There is a tendency to do that on any accreditation process and any kind of information collection process so that if your data collection instruments looks at only the number of immunizations and the number of child visits, you will get very good information and excellent performance on well child visits on the number of immunizations. I think most patients and I think the government wants to know what happens when these kids get sick. What happens when the patients need to get an appropriate physician. Do we have information on that. That's much harder to collect on standardized exams, I recognize.
But I encourage you to act on what you already know and that is to get the information that you need. How are people treated when they're sick. I think people can be extremely well treated in managed care with they're well. We need to know how they're treated when they're sick. I even envision using mock patients running, going from health plan to health plan with a validated condition and diagnosis and recommendation and see how they're dealt with from plan to plan. I would also encourage you to look at records of grievance mechanisms, all plans must have them internally in this field. Take a look, it may not tell you what is happening with every patient in the plan, but it will give you, may alert you to some problems that may arise in a plan.
Grievance data is something that plans maintain, it can be expanded a great deal. It can provide you with some very interesting information about, at least what the patients thinks didn't happen that should have. And there may be information about things that did happen that should not have had. And it may also give you an idea of the range of patient complaints that are founded more on fear than on quality of care.
I know there is some related concern for adding patient surveys in connection with physician incentive plan, information that would be added to the CAHPS survey. That is an interesting opportunity to see if there are different responses in the ways in which providers, physicians in particular are paid and if that makes a practical difference in the way patients are treated. We don't have very good data on that. And, actually, the Medicare slot of regulations don't apply to many people in practice. But it is, again, an opportunity to try and take advantage of the population that you have and the obligation that you want to carry out to collect the information to make a real difference for the population. I think it can be a wonderful driving force to set standards for not merely a Medicaid population, but for other -- for populations across the country to the extent that it sets an example and provides a gold standard, and plans that have obviously both Medicaid and other populations may find it more suitable to do the same for all, particularly when the data collection is simpler when one instrument is used for all. Thank you.
DR. IEZZONI: Thank you for your very thought provoking comments. We'll be back with some questions, I'm sure. But Nora Wells, why don't we hear from you.
MS. WELLS: My name is Nora Wells from Family Voices. Well, I too was wondering why I was here and I'm delighted to be on this panel. I think I'm in the right place. I'm a national coordinator with an organization called Family Groups, it was founded five years ago when there was a possibility of national health reform. And the people that founded it were all peradacta [phonetic] activities who were chomping at the bit to be sure that there was some kinds of entitlement for adequate health care for all children in the country. And as we all know, that opportunity passed. But at the same time, we realized there was an enormous change going on in health care in this country and that it was really vital for what I call consumers, people who are recipients of the physicians to actually know how it's affecting people. And the particular angle on that group of people speaking up that we try to represent, our kids who have special health care needs.
One of the issues, of course, is who are the kids with special health care needs. We have so many ways in this country of trying to describe people who might have more than the usual needs for health care. Our particular description is that all children deserve quality health care. We tend to represent kids who need more health care than normal. Families self-identify as being part of our organization, so we are a little bit leery of too many divisions of who should be or shouldn't be described as a child with special health care needs. But among our membership, we have families who are very active with mental health issues, who are active with the most severe kinds of medical issues and the whole range of things in between. We have ten national staffs; we're located in four locations, New Mexico, Massachusetts, Iowa and Illinois, those are our paid staff. We have volunteer coordinators in every state, and we're very active with policy and control issues with every state in the country.
Basically we try to work at the state, regional and national level where committees, advisory boards, part of projects. We work with state health departments very actively, with Children's Hospital, organizations that represent Children's Hospital, AAP and NAGRI. We work a lot with health care all around the issues of how can program and policy really think about what families are saying they need from good health care, quality health care for kids who might have more than the usual kinds of needs for health care. We've done a lot of commenting on HEDIS. We work closely with the CAHPS project. We think it's very important to be collecting data on what's happening in the health care system.
But let me tell you a couple of the specific areas we thought there needs to be some extra things done. We are funded by maternal and child health, we've gotten projects through Robert Wood, Johnson, Packard, Casey, a number of foundations. Basically what we'd like to see is that all children in the country get the kinds of health care that's needed. And we feel that in order for that to happen, consumers need to have a role in figuring out how to collect data, how to report it and what ought to be done with it.
I wanted to start with a report that we did a couple of years ago on a project through the Casey foundation, and I gave you a little summary right here. What we did was since managed care is for kids receiving Medicaid and for kids in private coverage, we wrote a small consumer survey and we were funded to collect information from families in four locations. The locations were strictly based on our funder's desires. It wasn't done in any other way, but it gave an interesting kind of snapshot at the time two years ago. The sites were Des Moines Iowa; Atlanta, Georgia; Denver, Colorado; and Seattle, Washington. And we -- first of all, we asked the familes what kind of services their kids received. And if anybody's interested there's a much longer report I could share with you. And we found that all of the families who answered our questionnaire, and there were about 353 responses that we could use, were families whose kids had quite significant needs, all of them saw pediatric specialist, they had significant needs. We asked them questions that ranged from what did they think of their primary kind of cares did; to, what do they think of their specialty services. We had a way of looking at that care; we asked them to tell us whether they're in managed care or not managed care.
All of you know it's very hard for the consumers to know whether they're not are. Two years ago we got some responses that told us whether they're in managed care or not in managed care. Basically what we found was whose families that were answering our really felt pretty good about routine primary care -- (MICROPHONE PROBLEM) -- but the families that answered our questionnaires felt that they did get good primary care; that their primary care provider had knowledge of special needs; that they did good in hospital services; prescription medications and laboratory services. However, there was a very strong concern that families were not getting services that were needed by a much smaller group of kids, but which are very important, I think, to an indication of quality in a plan such as home nursing care, personal care attendants, hospital services and particularly mental health services. That was a screaming, screaming, screaming need in that small survey that we did initially.
Almost all of the families who said they had mental health needs felt their services were not being provided adequately. One of the things that we felt was that a lot of the kinds of things that families were saying in that survey were things that would not be that difficult to correct. They felt they didn't have enough information about a lot of the basics of the plan; how to file a complaint; what to do about it.
When we hear from Barbara that being a consumer hot line in the Attorney Generals office, how many families on Medicaid know that that line is there that they can call into. I think that sometimes when we collect certain pieces of information, we don't know -- as Wendy has said, what we didn't ask, we don't know. We found that families who were not in managed care, at least the way we asked the question, were almost twice as likely to be highly satisfied with a number of services that are pretty important for kids' special needs such as durable medical equipment, adapted equipment, medical supplies, nutrition products, home nursing, personal care attendants and emergency care services. Almost twice as likely to be satisfied. That seems to me to be kind of a marker that says, we need to be looking at how managed plans are to be working with kids who have special needs. Is it because they are not set up in additional ways as in the population as other kids, families not in managed care.
Now, we lumped HMOs and PPOs together in the managed care quality. But, again, families not in managed care were twice as likely as families in mannaged care to be highly satisfied in aspects of delivery service such as; access to a case manager or care coordinator access to opinions; the kinds of appeals and grievance procedures; and particularly the flexibility to use cost effective providers. Big issues with families with kids of special needs is the kinds of restrictions that certain managed care plans put on certain providers you could use.
And the families are as frustrated as everybody else. They want cost effective use of services. Medicaid families very often tell us, for example, they want to know what's billed in the name of their child. They don't want bills going out and not knowing whether this kid is getting -- supposedly being billed for oxygen or whatever services. They feel very uncomfortable with that Medicaid systems in many states is not set up to give explanation of benefits to families, and that's a big mistake in the way the system is used.
I guess similar to what both Barbara and Wendy mentioned, the families that responded to this small questionnaire that we did definitely showed a trend that as choice was restricted, they became less and less satisfied with the plans they have. Now, what do we needto do about that. I mean obviously our system is turning toward a direction which is going to limit people's choice, so we do need to think about what does that mean and how can it appropriately complained and how can it be appropriately utilized by the people who need the services.
The reason we think surveys of families with kids of special needs can be particularly helpful to a system is that these are kinds of children who are markers for all kids in a system. I love the statistic where people collect satisfaction from the general public and the general public says they're satisfied with their plan but they're very worried about what would happen if they become sick. Our families are all children -- our families are all children. Our families are all families of kids who are already, quote, sick. The families don't describe the children that way, they consider their children very healthy. My child has cerebral palsy, is the most healthy of my three children but they do need services through a health plan. And your goal as federal and policy makers is to really ensure that our system is providing the care that we want to our families.
Let me make a few recommendations. Let me tell you one other thing that we're doing. From that small little pilot, we were able to put and get -- successfully get funded to do a broader national study of families with children with special needs. We developed a tool which is hot off the press. I have just a few copies, but you can send it around and we'll take a look at it. We are doing this project in conjunction with Brandeis University. The first group is going out to 6,000 families, but we've just heard from California that they want to expand it to another 9,000 or so in several counties to see if we can get it going in California. This tool was built on the first tool that we used. Again, it asks a lot of kinds of questions that HEDIS isn't asking right now, even CAHPS isn't asking. We feel questions that are very important and I'll tell you a little bit about what they are.
But I would say that one of the important data collection pieces, I think, is information that comes from outside any particular kind of system. Because children like ours, children with special health care needs, for example, are served by many different systems. Many of them have Medicaid and they have private insurance. Many of them have two kinds of private insurance; the mother's plan, the father's plan, the stepfather's plan, the step mother's plan. And the reason is they're not getting adequately served by any of those plans, that's why families have so many different kinds of plans; through their school, ed plans through their special ed plans, special health care needs programs. The reason why all these programs exists is because we have no good quality of health care delivered to kids in this country, it's completely piecemeal.
And children with special health care needs is a marker of how the health care system is. One of them is, who are children with special health care needs. If we do think that they're kind of an important indicator of what our good health care system -- (MICROPHONE PROBLEM) -- national census that went out that helped trigger kids into certain ways of collecting information, but it's a very important question. Many of the tools that are out there are only collecting data on things that aren't really good indicators in what I think is quality health care for all kids of whether good care is being delivered.
Access to first line drugs, very, very important for kids with special needs. My own son had a serious two-, three-year reaction when the drug that he'd been taking for many, many years switched to a generic. I didn't even know it switched, he was the one that figured out the color of the pill was a little different. But it took us a while after that why he was having a bad reaction to his medication. What care is being provided by families in their homes to care for kids. Huge amounts of care has been switched from hospital-based care to home care provided by families themselves. And we're not measuring what's happening, who's doing that. How does that match with what the HMO is providing, what the Medicaid agency is providing, et cetera. We have no national standard of medical necessity.
Extremely important issue; who is deciding what care a child or anybody gets. And how are we monitoring what kinds of medical necessity decisions are being made. I love that idea of the same child or the same person going from plan to plan and seeing what people give because we know how inequitably care is delivered in this country. You know, we know that -- well, another big issue I think is, what role do families have in decision making; how do families understand the mechanisms within managed care for how decisions are made; do they understand who is actually deciding whether the child gets the visit to the pediatric specialist or gets the visit that has the real expertise in that area.
I would make a couple of recommendations. I don't know completely how your committee works, but it seems to me there is some data that is out there right now that's not being analyzed in the best ways. Whenever I ask about data on SSI for children, not all of SSI because, of course, SSI runs a very large gamut, but if we want to look at what is quality care for children, we've got a huge collection of data, which in terms of privacy doesn't have to be linked up to any given name, but it certainly could be measured right now as to what's happening with kids on SSI and fee-for-service because a lot of those kids are still in fee-for-service.
What are we asking state Medicaid programs to do in terms of monitoring some of the more prevalent conditions. Whenever we talk about monitoring at another level, and someone mentioned it this morning, people get when they say there's too small -- the populations are too small, but there are some pretty broad categories. There are the category of kids who are on early intervention kinds of plans, those are kids -- some of them are at risk but some of them have pretty important special needs.
Anyway, I think in terms of looking at HEDIS, it's very important that we look at some of the conditions and what's going on with the Medicaid system and how those kids are being served. I loved the persons suggestion this morning that we really monitor some of the services being provided. You know when we passed our special education law in Massachusetts, we said, It's no longer to live just by a diagnosis, we have to live by what does the child need, and our plan should be based on what the child needs. It might be a child with mental retardation, spina bifida, something else and they all might need some physical therapy. And to monitor what's going on with physical therapy in our Medicare programs across the country would be a very important indicator, especially if it's done over time about how our systems are meeting our children's need.
The last point I want to make is the whole role of what families roles are in data policy collection, data design and data reporting and then the use of data. I'd say to you, I think families can be very sophisticated users of data. When we've met with all the other players that we have about how tools are designed; how are you asking the question; who are you asking it of; why are you asking it; what are you going to do with the information. I think we've built some very good relationships with researchers as to what role families could have in helping design tools, think about how data could be collected and how it could be used. I would urge you in a federal role to think about how that could be institutionalized. , how we could require people who are being funded to collect data, to really think about using the voice of consumers in the direction of collect -- of designing tools, thinking about projects and following through on them.
And then I guess I loved Wendy's comment, No data is important unless it's used.
In my mind, nothing is good sitting on a shelf. f you don't use it, it hasn't kind of fulfilled its promise and I think that's where consumers families and other adcrowel [phonetic] will take the data -- doors and we will walk the halls of Congression to say these programs are not meeting the needs of the people they're supposed to serve. And so another very strong point to include as policy makers, to include families and consumers in setting the standards for what data ought to be collected. Thank you.
DR. IEZZONI: Thank you. I think you've all proven why we asked you to speak, even though prior you didn't know why. We did start a few minutes late. I don't want to go much longer because we need to break for lunch. But are there questions from the members of the committee. Elizabeth and then Paul.
MS. WARD: I think this is to Nora but if Wendy or Barbara, you can also answer. When we were in Phoenix having some hearings, we heard from a mother who gave us some very nice personal testimony. And her concern was that in responding to these multiple surveys, she actually was feeling fairly threatened. And she felt that this was not herself, but that her friends and dealing with friends who also had children with special health care needs, that there was a question of who was going to use this data as long as it was identifiable; and could it be used against her essentially because of things she was saying about the services. Do you have any similar experience from your association?
MS. WELLS: Yes, it's a very important issue. A lot of experience that I've had with it is anecdotal. But when we were trying to change laws in a lot of different places, getting families to testify about it is very threatening because our system is fragile and we are not providing the services that kids need as it stands right now. And to get somebody, you know, to go out on a whim, but nobody wants to loose what they already have in order to try to improve the system.
I guess I would say to families in collecting data, we have no intention in reporting any data that's connected in any way with any given person. And we always collect it and try to report it in ways that's really more system-based, that's really intended to try to look at how systems can be changed. I can't answer to some of the people who were talking this morning, and I may not have heard it all as to tieing specific pieces of information to specific people. But in all the genetic works that we've worked on, the national GNO project and things, you know, that whole area of what you have to report on in order to get information and what people know about you, and what it's going to be used in order to discriminate against you in the future is a very important area. I think you know -- what I think is important is that families understand more about how data is used, how it's being collected and how it can be used to improve systems, and what it's connection to a given name and address.
And I would say to all researchers, it's extremely important that the privacy of families be protected and that researchers, even when they're keeping stuff under lock and key, it's a little distressing as a family to know that because you're on the Medicaid program, your name might be passed on to blah, blah research group who would then have it. Families don't always understand that, but when they do find it out, they get very nervous legitimately. So they need to know more about, Well, what does that research actually do with that information and is that a legitimate use of it, and it is --
MS. WARD: I think rather than the research question, the real question I had was health plan in servicing their enrollees.
DR. IEZZONI: And that mother actually didn't want to report the health status of her child.
MS. WARD: It wasn't the satisfaction measures, it was saying; my child cannot walk, my child cannot hear, my child cannot hear.
DR. IEZZONI: We do need to move on.
DR. NEWACHECK: Paul. My question is for Nora. I know family voices helped -- that's included in the CAHPS instrument and I'm wondering if, at this point, does family voices consider the contact to be appropriate and adequate --
MS. WELLS: I love the CAHPS project. But my answer has to be no, not quite. And that's only because I think it -- well, it got as far as -- our last involvement was last December when we ran a couple of focus groups and helped to think about it. We've had a wonderful relationship with the CAHPS project and it's terrific to be part of developing a tool which has such a big possibility for impact. But I didn't hear Jack this morning, so I don't know exactly what status. I heard about the conversion of HEDIS and CAHPS. And we've been knocking on the door of HEDIS for the last couple of years as they ask for public comment. So I don't know exactly what status it's at right now. Our opinion is there definitely needs to be more questions about children, this is the HEDIS. And what the combination of HEDIS and CAHPS is, I don't know. And I think that the CAHPS tool that I last saw is really going in the right direction but needs more development.
DR. IEZZONI: Okay. Hortensia.
DR. AMARO: I have a couple of questions. I think the first one to Barbara. I was wondering whether the Attorney General's office has played any role specific to the definition of data collection requirements in its role to ensure that with -- especially with managed care organization's efforts to reduce costs, there's still maintenance of quality and access. Has the Attorney General's office played any role in trying to articulate what kind of data would be needed to evaluate performance with respect to quality and access?
MS. FAIN: I'd have to say, not really. We've -- okay. I'd have to say, not really. We often will look to what other people have collected to try to inform our policy recommendations as to what should be done. I'm not sure if -- part of what you're getting into is the confidentiality issues. Is it more quality?
DR. AMARO: No. I was just wondering that it just seemed within the role that's described, and I just looked briefly at the report, would very much encompass that in order to assess some of the issues that are in here, you would need to look at some indicators. And to look at those indicators, you might need to -- or the Attorney Generals office might be interested that they are collected so that there can be oversight of -- you know, and that issues of equity and access would be sort of natural to the Attorney General's interest and so --
MS. FAIN: Right. That is not -- that type of work, that type of data collection, data analysis, is not work that the staff of the Attorney General's office would be qualified to do. What we have recommended, and you can see there are certain areas in the report where we either borrowed from what is already out there or made recommendations as to what needs to be collected before informed policy decisions can be made.
DR. AMARO: Oh, okay. That's what I was asking.
MS. FAIN: One particular area off the top of my head is in the area of -- there is a bill, a manage care bill pending in the state legislature here that would regulate, it's sort of -- I refer to it as an omnifes [phonetic] manage care bill, and one of the big fighting points has been over the issue of physician incentive, physician compensation arrangements and how that affects -- there's a concern that how physicians are paid affect the quantity and quality of care that's delivered. And some people have called for prohibitions against certain kinds of incentives or any incentives. And we, for instance, in this report took the position that not a lot is known about the affects of incentives. And we made some recommendations as to what information should be collected and how it should be monitored in order to make an informed policy recommendation or decision. There were also recommendations that we made regarding more of the HEDIS kinds of quality data. Some people have advocated that the states do serve this function in many ways, duplicate or the efforts of HEDIS and other organizations out there that do this kind of data collection and analysis.
And we made some recommendations that basically the state work with the existing organizations to improve those kinds of quality measures and to focus NCQA and other organizations on collecting data in areas where they might not have been in terms of serving children; the other underserved; at-risk populations. Because there's obviously data missing, but that's not something, as an Attorney General's office basically a staff -- which is staffed by lawyers, not analysts, policy analysts, it's not something that we ourselves would get involved in.
DR. AMARO: My question wasn't rather you analyzed the data but whether you've been involved in the discussion of trying to articulate what data are needed in order for the Attorney General's office to carry out its mission of guarding the issues of quality and the issues of equity cross-populations --
MS. FAIN: We often do that. We'll either look for the information or request that an outside organization do that. In the report there are specific areas where we have made those suggestions.
DR. AMARO: I'll look for it. Thank you. I just have a quick question for Nora. Thank you for coming and it's really great to hear the work that your group is doing. I was wondering if in your survey you found -- you used in any way race and ethnicity, and you found any differences in satisfaction with services across families of different, either interracial, ethnic background or other economic indicators?
MS. WELLS: The first survey that I talked about, we didn't really see any. But we had so few numbers, you know, in certain categories that it was really hard to say much. This next one we certainly asked questions about race, ethnicity, area of the country, kind of educational background and income and things. And I think we may have many more numbers in order to see that on a national level. We won't really have enough at the state level. It's in 20 states but it's smaller numbers in each state, but bigger on the country. Probably in California, we might.
I think families feel very strongly that -- well, one of the pieces in our first survey we asked a competent question, something about how they felt they were respected. The answer to that was kind of positive actually. Families basically reported in that survey that they felt culturally respected. But I don't know what will happen in the next one, which is a lot broader.
DR. IEZZONI: And our last question for the, quote, morning session is Kathy.
MS. COLTIN: One of the things that we've heard is the risk of oversurveying particular populations and what that does to response rates. I'm aware that there's a group called New England Search --
MS. WELLS: Right.
MS. COLTIN: -- which also focuses on children with special health care needs, and they have developed a survey of families with children of special health care needs. I know that we are one of the pilot sites for their survey as is the KIESER plan in New England. And I wondered, first of all, whether your efforts are in any way coordinated and whether you see any problem with that? And I also have a follow-up question to the first one.
MS. WELLS: Sure. I worked with New England Surveyors, I was a consultant with them for five or six years, I'm very familiar with the project. It's a wonderful project, it always worked collaboratively. One thing we did when we started our voices, we pulled together the people just in this area, so that was Jack Fowler's group and New England Search and us and some people in several other projects and we really sat down and tried to think about some of these overlaps. And we made some conscious decisions to not serve in the state New England
Search was, things like that.
If I could get to the heart of the question. I think it is absolutely critical that in your role as kinds of guiders of a system, that we think about the overlap in projects and how we can get people to coordinate with each other and collaborate with each other and share what's going on. As an outsider in the research community, it's a lot easier for me to share because I don't have quite as much at stake possibly, or the same history of how people relate but I think it's a an important issue.
As a family member, I would say, asking any question is only as good what we're going to do with the answer. And we need to assure families that we are refining our process to the point that we are going to ask questions that are going to be useful, and that we are really going to use the data. What we say to families in using this questionnaire is, we want you to know this information. We want you to use it. We're saying that to the individual responders as well as to all the family advocates in the country. I'd like to see that principal be used in all kinds of data collection though. Families ought to really know what we're saying about the information we collect on them. And that ought to be part of the actual tool. It ought to say right inthere, you know, this is what we're going to do with this. Not specifically you, it's not specifically going to have your name on it.
MS. COLTIN: Follow-up.
DR. IEZZONI: Okay. Make it quick.
MS. COLTIN: In working with New England Search, we've constituted a focus group of families of children who have special needs in our health plan as well as providers --
MS. WELLS: Great.
MS. COLTIN: -- and one of the things that we were sensitive to is the reaction to families to receiving a survey that has, quote, identified them as having some particular special need and whether they would see that as a breach of confidentiality. And there were very interesting perspectives raised by the families and by the providers both. And we ended up with a methodology where the survey is actually going to go out from the providers with a letter. Although, the health plan is going to identify protective families to the providers and say, Do you believe that our list is accurate that these children are, in fact, you know, children of families who have special health care needs. And they can take families off the list or add families to the list. But the letters will actually go out from the providers. But it was interesting because families said, it's not so much that we mind that the health plan, we know they have our claims. It's just that we would prefer to make sure our provider knows and that the approach to us is coming from our provider, and that they believe that this is a valuable information and that we should provide it to the plan.
MS. WELLS: One of the things I love about that kind of anecdote is that it shows, you know, if you can ask families some of the things about the way you're developing, the tool, the letter that's going out, you can get a lot of information.
MS. COLTIN: So my question is, How are you identifying --
MS. WELLS: Right. We have two methods in each of the 20 states. Originally, we have family from the family voices membership list, it's a mailing list and we've randomly selected a half a sample from that group. And in order not to totally bias it as people who are interested and trying to help a system move along, the other half -- actually, two-thirds because we weren't sure of our return rate of our sample in each state, is through the children with special needs health care needs department. We know that varies enormously from state to state.
Those families -- that list we were being careful to develop a plan of sending out the questionnaires and request in a letter that introduces the questionnaire which shows those families that, they are not known to us as family voices they are only known to a children special health care needs department in their state. And, in fact, it's the children's special health care needs department who is sending out the survey and keeping everybody's names confidential. And we're only going to have an ID number. And it will be the health care department that sends out the return questionnaire. It's a very important issue.
DR. IEZZONI: Thank you very much. Sorry to make poor Nora talk so fast. But Wendy, Barbara and Nora, thank you very much for an informative session. I'd like to ask the subcommittee to return for 1:30, even though we've run over because I want to be respectful of our next panelists and our time since they were booked to come at 1:30. So you can feel maybe free to bring in food in here. I think it's okay to bring food in here, if you'd like.
(Morning session adjourned at 1:00 p.m.)