[This Transcript is Unedited]
Hubert H. Humphrey Building
Room 305
200 Independence Avenue, S.W.
Washington, DC
Call to Order and Introductions, Review of Agenda - Bob Hungate, Chair
Overview of the Issues - Subcommittee and Staff
Review Preliminary Findings (per revised Background paper) and further revise as needed - Subcommittee and Staff
Continue Review and Formulation of Findings - Subcommittee and Staff
Agenda Item: Call to Order and Introductions, Review of Agenda - Mr. Hungate
MR. HUNGATE: I have two work arounds that I know exist, Steve Young needs to be gone from 10:30 to 12:00 and Marjorie needs to leave at 3:00. Are there any others that we need to incorporate in today’s plan? Ok, well, I would like to spend a liberal amount of time in kind of getting to know each other, now I don’t know who knows each other well and who doesn’t, and so I’ll give you a brief history of who I am and what I believe in and so on and why I think I’m here --
DR. STEINWACHS: Can you tell us what our chances are of changing your beliefs?
MR. HUNGATE: Well, I could. My father was, he’s now 97 and no longer able to function well in the way he did but he’s a well known microbiologist, and he forced me at an early age to never give up an argument unless I was convinced, and so yes, I think I can change my mind, but I have to be convinced. And it used to drive me crazy because I’d ask him a simple question and he’d give me a 75 sentence answer, which was far more than I wanted, I was a teenager, I wanted it simple.
But anyhow, I come to this role honestly I think from 31 years of working in Hewlett Packard Company where I worked in a lot of different functions. I took an MBA from Harvard after going to Washington State in general studies and concluding I needed some calling card in order to get a job, and so I went to the business school and ended up at Hewlett Packard in California for five years, worked in the controllership function so I learned numbers and the characterization of businesses in numbers. And then I moved to the medical electronics division which was in Massachusetts, and was in the numbers game, but soon thereafter I moved over to the marketing leadership and ended up leading our worldwide sales effort, marketing effort for patient monitoring equipment, EKG machines, ultimately fetal monitoring, early computerized interpretation to EKG’s, so that I got into the technology of things that are supposed to the profession for use in the delivery of care.
Then I ended up moving to a general manager job where I led a hospital supplies business, and this was when quality improvement was coming in and I became a believer in quality improvement. We had a division in Japan which was started as a mirror to a division in Colorado which was our best division, it has the highest margin of any U.S. division. In the first two years the overhead rates in Japan were way above, the profitability was way below the level in division. At the end of five years, the profitability was above the level in division and the failure rate was one tenth the failure rate in the other division. And it took a long time for us to understand, and Hewlett Packard has always been known as a high quality organization, and we believed it. You better believe we believed it. And so I was leading a division and trying to get that division to improve its quality, and I learned that people do things for their reasons, not for mine. And so that was a very important experience for me and it has affected how I look at most everything.
After that, I ended up wanting to do some things in the division that my boss didn’t want me to do so I said look, if I can’t do what I want I’m not going to do it very well, and so I came down to Washington on health policy for Hewlett Packard. I was sent by the electronics group because we had been blindsided by DRG’s, our capital equipment business fell, just tanked for six months as people dealt with the uncertainty and didn’t know what to do. And we said we can’t afford that, we should have figured that out but we didn’t, so I came down to understand what was going on. But when I got here and started to talk about my mission, I said Hewlett Packard and people said oh, you’re here to talk about health care costs. Well, I really wasn’t, but I ended up having to.
And as I studied it farther I ended up in my job in quality assurance, quality improvement method, I had three customers. I had the medical products group for product access to market. I had the corporate leadership for the cost of health care, because it was a big chunk of our bottom line. And then finally I had the employees for their health, because I found that the personnel department people, when somebody got sick, would advise them to go out of the HMO and go into a fee for service so they had control over where they got their health care, contrary to the policy that the personnel department was pushing. And the only way I could reconcile all three of these agendas was around the quality improvement, so I’ve been focused on the health system ever since, and that was 1986. I was 51 at the time and believed that I could get the health care system straightened out by the time I reached Medicare age, optimistic. So I think I’m here because I take a system view of health care, understand information, understand organizations and how they work. And so I think that’s what’s brought me to this position, and look forward to working with all of you.
I like to work in a collaborative team based manner where we all kind of know what we’re trying to get done and I think we need to spend enough time, and I’m delighted that Kathy’s been able to join us, because our first try at things was a little hard because we didn’t have much background and I think we can build a base with her presence that will help us a lot in going forward.
So that’s enough from me. Marjorie, let’s go on around.
MS. GREENBERG: I’m Marjorie Greenberg from National Center for Health Statistics, CDC, I’m the executive secretary to the Committee. I’m really here in that capacity I guess and as a cheerleader in a sense. I really appreciate Bob taking on this challenging group and mission, and I really appreciated your jumping right into it but also really wanting to kind of step back and say ok, where have we been and where are we going from here. I really appreciate Kathy coming. Kathy we thought had a lifetime membership on the Committee but she finally left us, but I guess Kathy has a lot of, in addition to all of her obviously expertise and responsibilities in the real world, she has, if the Committee is not the real world, she has a lot of institutional memory about this Committee and of course this particular workgroup which started with her leadership.
Just as kind of an aside but to maybe put into perspective I think the importance of what this group is doing and what the national Committee does, yesterday I spent the morning at the Gallup Poll Organization, which had contracted with the General Services Administration, to survey all of the advisory committees. And they, actually they thought they did great, they got about a 50 percent response rate but for a web-based survey that probably is good, and I’m going to at the next, at the June NCVHS meeting I’m actually going to present the results for the NCVHS. I thought it came out looking good, definitely, it didn’t quite make the highest cut of best practices but we were up against committees that have very clear missions like we will make recommendations on which vaccines should be approved and then CDC accepts those recommendations, and obviously we’re in a much more complicated world in the National Committee and in this particular workgroup. But some of the best practices included obviously being very clear about what you were about and your mission.
But I think what came through in that meeting yesterday was really what an important role these advisory committees play in linking public and private sector, in advising the Department, and in building trust in government, in communicating, all of those functions really came out very clearly in the whole kind of range of advisory committees, and they said it was almost unheard of to get as positive sort of feedback as they had gotten both from the members of this wide range of advisory committees and 30 departments, as well as their executive secretaries and the policy makers whom they advise.
So I just want to kind of share that with you because I think that, of course. the single biggest frustration was that they didn’t feel that their advice was always taken, or sometimes they didn’t even know what happened to their advice, and that’s a problem that is universal. But I think actually with the National Committee that has changed somewhat, particularly with HIPAA, and there’s much more focus and much more recognition of the importance of the advice that the Committee provides. So I think this is an incredibly important area and my love in life professionally is standards, so anything that we can do to bring greater standardization to the measures and to the data so that we have more comparable information that will help improve our health care system and the health of the population, obviously I’m a public health person as well, I strongly support. So I appreciate the opportunity to participate and look forward to your deliberations.
DR. HOLMES: I’m Julia Holmes and I’m from the National Center for Health Statistics as well. I’m in the division of health care statistics that conducts the provider based surveys. And my primary job at NCHS, I’ve only been there a little bit over a year, is to actually be kind of the point person for providing NCHS and CDC measures that will go into the National Health Care Quality Report and the National Health Care Disparities Report, so I have kind of original experience with the limitations in terms of the availability of data as well as the comparability of data that we’ve encountered in producing these reports.
Prior to that I worked at Blue Cross/Blue Shield of Michigan in the Center for Health Care Quality, and we conducted studies about quality of care for some major groups, such as for Ford, General Motors, Chrysler, and based on that experience I have kind of an understanding of the limitations of trying to measure quality using administrative claims data and the difficulty of getting access to patient medical records. So that kind of provides an overview of my background in terms of quality measurement, but obviously it’s an ongoing interest of mine. I look forward to working on this Committee.
DR. STEINWACHS: I’m Don Steinwachs, I’m at Johns Hopkins Bloomberg School of Public Health, and I chair the Department of Health Policy and Management and currently director, but have been part of the Health Services Research Center there for about 30 years. Kathy and I started our careers sort of similarly probably in some ways, but I was an operations researcher who didn’t know quite what to do and got involved in public health and health services research, still didn’t quite know what to do. Sam Shapiro said to me, Don, have you thought about data systems? Maybe this would be a focal point for how you could build a career. Actually I’ve been hoping to become chairman of the board of General Motors but I was halfway through my Ph.D. and so I discovered who’s going to believe I can manage anything with a Ph.D., so I figured that took care of that. So Bob, I would have liked to have been in the private sector with you.
And so out of Columbia Medical Plan we were involved in the Health Services Research Center in the early days of trying to develop an encounter based system and issues of taking text and coding into, I think it was ICD-8, 7, 6, I don’t know. Not 6, probably 8, something like that. And then I remember we had one of the early MPROBE’s(?), Al Mushlin(?) led it, and it was the first, it was the one MPROBE that looked at, it was an experimental medical care review organization, looked at trying to measure quality in the ambulatory arena. All the other ones were inpatient, much more sort of utilization review where the predecessors to PSRO’s, so started trying to test the encounter data system for addressing quality issues.
Number one study which was my sobering beginning of quality was we, clinicians in internal medicine developed a set of criteria for how they wanted to manage people with high blood pressure, and part of this was frequency of return. And so one thing they used the data system for was to identify all the people who had been diagnosed with hypertension and so we came back to the chief of internal medicine and said well here are your criteria, here’s a report that shows, it was less than half made the three month return interval and it didn’t do all that well six months either. And he said well, that’s very important, I appreciate it, and I sort of said well what are you going to do with this. It’s awfully busy and we don’t have any resources, and so on, so I knew that I was already in trouble in trying to change the world. So Kathy knows how to interface these things and put them together. So it did convince me that information alone wasn’t likely to do this, but I do agree that we need to be able to measure it, and if you can’t measure it, I believe you can’t manage it.
Over the years I’ve done research in a whole variety of areas, but have sort of at this stage probably two agendas, one is I very much am enamored with the evidence based movement trying to say well we ought to be trying to do what we can to move practice to evidence based and answer the question, how many people ought to receive evidence based care, what are good variations versus bad variations. Also very interested in how in the broader community we can do the translation process, and so we’re doing some things now. Have a survey going into the field of mostly psychiatrists in Maryland and a sample nationally trying to assess not only their readiness to adopt evidence based standards, the pilot suggests they aren’t, so that takes care of that problem. But the other is learning style, trying to think about what are the different kinds of learning styles, there’s been some research in family practices to suggest different, and then go to focus groups and then try and get research funding to test interventions that are tailored to a physician’s learning style to see can you make more of a difference there.
The other commitment I have, which is a very large one, is a view that we really need to bring the consumer and the family into the health care system, so I’m a believer that if people have information they really see as relevant they can help make part of the change. You can’t say they can’t be the change force, but it’s always striking to me how often physicians in practice will respond to what patients want. So if you sort of have a sense of what you ought to be getting, if you have the four or five things that someone who has schizophrenia, congestive heart failure, whatever else ought to be having, you have a fighting chance of at least probably having some of those key elements.
So I would love to see things that we do that move about into the consumer arena and the public so it isn’t just held by those of us within the system. And those things that can really help the transformation process which we know otherwise is a very slow thing if you’re just waiting for osmosis to occur out there.
DR. EDINGER: I’m Stan Edinger, I work at AHRQ, and currently in sync I’ve worked in probably in every part of AHRQ at one time or other in my checkered life. I came to the health care field in a very strange way, actually as a graduate student. I was actually doing art conservation, fine physical chemistry to it, and it kicked me out of the field --
PARTICIPANT: Did you make a big boo boo?
DR. EDINGER: Actually not really, I think I actually got kicked out for two reasons, one of which is I tended to blow things off and they didn’t want to leave me around the Mona Lisa. And I think when I disproved that you needed scientific methodology and knowledge of art to find a fraud they threw me out. There was one day at the NY Institute of Fine Arts a very famous painting was there for authentication and many of the world’s great leaders of the art field, the art historians and conservationists were all looking over the painting and arguing whether it was real or not and they spent weeks arguing over it, they were doing a little testing on it. And then one day I walked in and said it’s a nice, I hadn’t been up there for several weeks, I was on vacation, I came back and said oh, it’s a nice fraud, and they were all very angry with me because it took them months to figure this out, and they couldn’t figure out how I knew and I told them it was very obvious from the stroke technique, and when I actually admitted how I really knew they really wanted to kill me. It was actually simple because one day everybody had gone out to lunch and there was nobody guarding the painting, so I assumed it must have been a fraud or they never would have left it with one guard -- that was the end of my career in the art world.
From the art world I went to Mt. Sanai and actually worked initially when they were doing, in the hey day of, this is almost 30 years ago, of automating the lab automation and hospital automation in New York. Tom Charmas(?) was there in New York and we were there with the 1199 in the midst of the strikes in New York trying to automate the computer system and it was actually a very good experience in learning how to automate patient records and medical records and lab techniques and what the problems are, interfacing with the medical staff who didn’t necessarily like doing it, the nursing staff and medical staff would disagree on what kind of paper the patient would be printed out or what color they liked.
And during one of the strikes, I think there were two reasons I left, one of which is during one of the strikes Tom Charmas decided to cook for everyone to show on national television everybody was rolling up their sleeves, and he cooked breakfast for all of us, and I think when I found the cockroach in the cereal I decided that it was probably time to leave because I don’t think I could take another strike and Tom Charmas’ cooking. But it was actually a very interesting experience on both sides of this. Seeing the problems of putting an automated system into place, the problems in dealing, interfacing with the medical staff, what went into the medical rec, what went into the automated record, how to disperse it amongst the various parts of the hospital, what data elements you needed to collect, and arguing with the computer company about back-up systems when New York’s power surges went about, and interchanged thousands of medical records, all the patient data was interchanged, so nobody knew what was on any of the records any more. That was one of the worst nightmares I think we ever had, along with the city health department who didn’t know what to do either, it wasn’t just our hospital that had that little problem.
From there I actually went to what was called the Bureau of Quality Assurance which probably nobody ever heard of --
MS. GREENBERG: No, I was there, I forgot to mention that. I worked for the PSRO data, I developed the PHDDS. That was one of my previous lives, too.
DR. EDINGER: And then I went to CMS when it was formed, and actually worked in the HCVA, I was there the first time they had a hearing on HCVA with Javitz(?) and all that, it was the first HCVA hearing, and they phoned the place and everybody was wondering what the hell it was. I worked in a lot of the labs, I think unfortunately one of the worst nightmares was working on the CLIA(?) stuff which was when I got to know her husband George. That was interesting, too, not just the regulatory part, but one of the interesting things was, we went into court hearings, one of the things, I got to do the glorious court hearings for all the labs that they tried to terminate, but one thing administrative people would always ask was why are you throwing this person out, this lab or facility out? What harm have they done? What is the evidence that they’ve done harm? It was very interesting to sit on the other side, and I’ve listened to well, he has 14 check marks that they didn’t comply with. I mean legally, legally yes you’re right, but I want to know also, what have they done wrong, what harm have they done to the patient, what did this result, this glucose that they supposedly never resulted into the quality, what did it actually do? And that was a very good question, and you couldn’t really prove it did anything. It was inferential arguments, it was never real proof that it had done anything wrong to any human being by the fact that it was 106 instead of 105, probably really hadn’t done any harm.
So I worked at that for a long time, I did some of the payment issues by default, I did some of the fraud and abuse stuff, didn’t know what to do with it so I got stuck with it. Actually once went on a raid and kicked a door in for the RIG boys, they went with the guys from the FBI and a few others and they kicked the door in, it’s here in Rockville because they didn’t believe in asking the guy for a key, that’s not how they did business, so that was an interesting experience.
I worked on the Hill, I’ve worked on actually the opposite side of Tom Skulley, he worked with Darmin(?), I was actually working with John Dingle, it was an interesting experience sitting across the table from Skulley and Bill Roper and other people who I worked for. That was a very big career move, too. And George actually who got involved in providing tons of data, probably made him very happy. And actually did work on a lot of the payment in Medicare and Medicaid issues, including payer issues. And then I went to AHRQ and I worked on the patient guidelines and I worked on some of the informatics issues in the floor days of Helen and Eduardo, and have been the, I guess I’ve had the distinction of being the first project officer for Mark Leclella(?) and David Bays and a whole bunch of others. After being Mark’s first project officer he never did research again, became a bureaucrat.
And basically like the rest of you I’m also interested in a lot of patient aspects, a lot of the data aspects, including patient record, the personal health record which we were talking about before. And the problems of coordinating your care, I mean as a, I don’t basically use the military hospitals being the public health service, and it’s very interesting to be in the position of both coordinating your care and trying to get your own medical records together and keep data, when somebody asks you what was your result two years ago and you have to go searching through your medical records, some of which times which are lost at Bethesda Navy, and being in places where computers don’t communicate with each other, you go to Bethesda Navy and then you go over to Walter Reed, and the same provider who happens to be working in both places can’t access your data because he’s in a different computer system. So it’s interesting being on both sides and trying to learn more about how to help in developing these systems so just maybe other people can benefit from them.
MS. KANAAN: I’m Susan Kanaan. I am a freelance writer. NCVHS is one of my main clients, or is my main client. I’ve been with NCVHS since 1990 and have written about, written or helped to write about a dozen of the reports of the Committee, including several that are very relevant to this report. I don’t have a long history with this workgroup because I think I started working with Kathy maybe last summer or so, and I was sort of brought in to help pull together the findings into a report. So I don’t have a lot of institutional memory of this workgroup although I certainly do of the Committee as a whole. And I’m happy to be part of this process. I’m very interested in this transition into perhaps a somewhat wider perspective on the topic, interested to see how you will handle that.
MS. CARR: I’m Willine Carr from the Division of Planning, Performance, and Evaluation at HRSA, and new to the group in part as the informal liaison from HRSA to the group. My background is in public health, largely around some health service research, policy analysis in the governmental relations, graduate level teaching around health administration and health policy, etc. I’ve worked for the Social Security Administration, the New York City Health and Hospitals Corporation, the United Hospital Fund, Meherrey Medical College, and a lot of other places it seems. But in each of those places did work around, tangentially related to quality issue, particularly for disadvantaged populations. And I guess my particular perspective and interest around the Quality Workgroup’s measurement concerns relate to being able to measure not just quality but quality disparities. I look forward to bringing perspectives in information from an agency to this group. I am particularly excited to be here at this time when there’s a chance to learn where you come from and where you hope to go, so I’m just very glad to be here.
MS. JACKSON: Debbie Jackson, I’m staff to the National Committee, NCVHS, been with the NCHS for about three years now. Out of all the committees I’ve staffed, I think the one that’s probably most pertinent would be the Accreditation Council on Graduate Medical Education where we helped to institute the accreditation of internal medicine residency programs, that was one of the first things under my tenure in the early 90’s when the programs were kind of going off and they were existing on their own and we had to get some sense of reigning them in, getting them into this green book, the Directory of Accredited Programs. So we set up all the requirements, the general requirements, special requirements, and it was my job then to get that organized and put together. So I’ve been drawing a lot on that experience in looking at assessment, whether it’s programs, whether it’s measures, to just see what are the comparable sets, measurable sets to see where something falls and an array of an analysis for a way of quality assessment that way, so I’ve drawn on that.
And I, too, am anxious to see the transition. Having seen the background as we’ve all read background documents that Susan prepared and seeing division of where we are, where we want to go, we’ve got new members on the Committee, we’ve got new perspectives and the world is changing, we all know that, we see that, it’s right in front of our faces, so this is a great place I think to be at a crossroads to really get some things done, under our belts, so I’m happy to be here.
DR. YOUNG: My name is Scott Young, I’m a senior clinical advisor at CMS in the Office of Clinical Standards and Quality. Like many days in my life I think I woke up today and heard these introductions and am honored and amazed that I am sitting in a room with you guys, this is really quite something after hearing about these backgrounds. Let me just tell you a little bit about my background. I’m a family practice doctor, I’m a kind of a medical mutt, proud to be that way I suppose. After I finished my training here in town I spent just a little bit of time at ODPHP, which I don’t think really exists still. And was hired by Intermountain Health Care and went to Utah and started some programs and kind of went to the Brent James School out there and learned some things, and was drawn in Utah to do some things like I loved doing away from Intermountain.
I started an organization called the Utah Health Care Institute and we worked to bring together collaborative projects around the poor and the uninsured and the disenfranchised and really tried to merge local, county, state, and hopefully federal programs in collaborative manners, just to try and flex solutions out. One of the ones we’re very proud of is a series of school based health centers, we used Department of Education money and were able to provide health care services for those, including reproductive services, and in Utah that was a significant thing to do. So it took a little bit of merging medical practice and policy and politics and some business acumen, you have to make payroll and these things at the end of the day, and what is the right thing to do. We learned some lessons in doing that.
I ended up doing some work at the state level while doing this around trying to flex information technology into rural and frontier disenfranchised areas. I don’t know that we came up with any great successes but we learned out that’s a really hard thing to do, and that there are a lot of barriers and a lot of, and that things that work in hospitals don’t work in clinics and don’t work in ambulatory settings. The business models are very different and the business models at the end of the day rule the decision making. And for a young doctor that was an amazing lesson, and a humbling lesson at times.
I ended up kind of giving some notice and came to Washington as a Robert Wood Johnson Health Policy Fellow and went to work for Jeff Bingerman(?) after my, spent a little time on the Finance Committee and spent a little time on the Health Committee, and actually spent a little time on the Energy Committee doing work on workers compensation issues, so I got a kind of a broad introduction. And I’ve tried to bring a lot of the lessons learned, both on a practitioner level on the level of trying to provide basic services to people, and the local and the state level to the federal perspective as well. Did payment policy, went to the prescription drug debate last year, the bioterrorism bill, and a host of other issues.
I wasn’t ready to kind of step out of the circle at the end of that, a lot of Medicaid stuff last year as well which is kind of my first love, and Medicaid policy, and actually the waiver program and some things I’ve done a little bit of writing I work on as well, went to somebody who I had maybe been on the other side of the table from during the last year, Tom Skulley, and said I’m thinking about, you seem like an honest broker, tell people what you think, and what should I do and he said come to work for me, and we’ll see if we can do a few things. And so I did.
Since I’ve started with Tom one of the things that interested me around was in the ambulatory setting, and we just really didn’t know what was going on. We knew that was where the vast majority of care was occurring, wasn’t the vast majority of cost, but the vast majority of care was occurring out there. We knew we had cost drivers out there, chronic disease and those sorts of things, and we really just had almost zero information that we could really credibly say, we had little studies and we had some systems studies and we had other things but to be able to say on a day to day basis, I know what the return to normal on hypertension is among physicians in New Mexico. Couldn’t say that. I could maybe say some things like that about hospitals through some of the powerful engines we have at CMS. So he said so let’s figure out a way to do this and we quickly came around to some of the work that was being done, NCVHS and others, and maybe we needed to try and design some demonstration models, which were our flexing information technology out there, and that’s what I work on principally. How do we cobble together standards, some sort of a way to capitalize infrastructure and payment incentives or financial incentives around the ambulatory care setting. The President said something about it about a month ago, Secretary Thompson has talked about it, Thompson talks about it a lot, they want to do it yesterday, and so that’s what I do.
So obviously to me information is important, if you’re going to talk about, you have to know what’s going on out there before you can make any kind of decisions. I’ve learned that on a personal basis. One of our clinics we had a capitation contract with Pacific Care which nearly put us under water, I had absolutely no idea what was going on in my own claim. So I bring it from personal levels as well as a national level to the debate here, and look forward to contributing a small amount. Thank you.
DR. JANES: I’m Gail Janes, I come from CDC in Atlanta. Like it sounds like most of the people around the table I have a varied background. My academic background is so long ago now it seems moot. I was in biology, biochemistry and in statistics and have done a variety of things over the years. Did a little bit of bench science before drifting off, and worked in cancer registry actually in Europe, which was an interesting experience in the Netherlands at the time when the Netherlands was beginning to play with the idea of a computerized patient record, actually before it was really being seriously looked at here.
Worked in cancer registry with the VA for a while, or a clinical trials for the VA for a while and then came to CDC about 11 years ago. And have worked on CDC, and CDC in Atlanta I think particularly has become increasingly involved and I think engaged in the issues of standardization and health data systems. In Hyattsville, NCHS I think has always been engaged on these topics but I think Atlanta has with its more traditional infection disease orientation I think it has been a bit slower to come to a sensitivity of these issues, both in terms of the date that we create and put out there, and how those data are used in the form that they take and then also the data that are out there with the combination of our own and others and how we can then consume that and use it to achieve our goals.
I’ve always had a chronic disease orientation, particularly diabetes and cancer, and so have really been involved in a variety of different capacities over that ten years at CDC, and I think two big initiatives. Worked with a small group of folks in Atlanta who I think we all kind of roughly think of ourselves as health services researchers, probably with small letters, because we all tend to be generalists down there. But have worked around two issues, one is this issue of data standardization which has started off as a small issue and has become a larger and larger and larger issue over the course of the ten years, and as we’ve tried to with great gnashing of teeth and tearing of hair to standardize our own data systems and just be able to bring those together, and then to align those with the growing group of national, international standards, and that’s still going on.
And then the other issue which is of course tied to this is the importance of preventative services and quality measures at meeting our goals of trying to encourage people to get preventative services, to ensure that preventative services are offered to people regardless of income level, to track the effectiveness of those services. And of course this is all very data related and again brings you into the ambulatory care world, and has brought us into settings where we did not traditionally live, sitting at the table at places like NCQA in fact, and so I have always worked in a sort of supportive capacity but have sort of followed that process along with some people like Jeff Harris who was very much involved in that for a long time before leaving CDC.
And actually worked with Kathy on an initiative that CDC sponsored, small initiative in which we had given some money out to a number of states to look at the feasibility of taking their own data and making innovative, thinking about innovative ways to use those data, and in Massachusetts, I say we, they and I sort of followed along, looked at their vital statistics data and administrative data and looked at the feasibility linking those and creating an enhanced data set which could then feed back and speak to the needs to both private sector and the public sector.
And so this has always been the flavor of the issues that I’ve been interested in and the projects that I have worked on. At least seven or eight years ago it sort of brought, I sort of wandered into an NCVHS meeting because somebody in my group couldn’t go and they said you know, this sounds like the kind of stuff you’re always talking about, why don’t you go and take some notes, and I came and I came back and said you’ve got to support me to go to this group, these people are great, and I’ve been here ever since so it continues to be an extraordinarily engaging group and we certainly haven’t run out of interesting topics to look at, so that’s where I come from.
MS. COLTIN: I actually appreciate the opportunity to do this, we never did this before and it’s really helpful I think to know something more about people’s backgrounds. I’ll actually tell you some things about me that you may not know. I started out as a nursing major in college and I went through almost two and a half years of it and I actually had two experiences with medical errors that convinced me that nursing was not the place that I wanted to be. And when you’re in training you do things that actually as a practicing nurse you never have the time to do, which is you’re assigned one patient and you develop a very detailed care plan and you know everything that should happen for that patient the next day. And on two different occasions things that should have happened weren’t happening.
In the first situation I had a very supportive instructor who helped us really track down the doctor and finally get an order to change something so that we could correct this and the error was prevented. In my junior year that was not the case, we fought, we tried to get something done for a patient but he was on an OR schedule and the surgeon said no time for this, bring him up to the OR, and the patient ended up dying. And that just said to me as a nurse I can’t write an order, I can’t fix this, my hands are tied, I don’t want this. And so I think that planted the seeds for me way back when about issues of quality in health care. I graduated as an English major.
I got back into health care when I started, my first job actually out of college was working at the Columbia Medical Plan, which is where Don and I crossed paths, and at that time I worked for Cliff Gouse(?) who went on to become director of what was then AHCPR(?), and learned a lot about research, which I really didn’t know a whole lot about although I had done research for the litigation department in a Park Avenue law firm during summers, so I learned a little bit about working with numbers, but I learned a lot more at Columbia, and they taught me how to do a stopwatch to do time motion studies and things like that, and got into looking at quality of care around issues like waiting time in the exam room and the waiting room and so forth.
And went to work in the pharmacy and at that time they were automating the pharmacy, and the pharmacists were too busy to work with the programmer so I got assigned to do this, knowing very little about data systems, just a little bit about how to use what came out the other end. And that’s where I really learned a lot about data systems, and began to kind of marry my interests in quality and data and information systems. And left there and went on to Harvard Community Health Plan in Boston, and they had one of the earliest automated medical record systems in the country, they were funded by what was then NCHSR, the predecessor of AHCPR, as the site for Costar Medical Systems back then, Community Health Plan, and started to think here’s a wonderful place to really be able to look at quality of care and information systems and medical records.
And one of the things I was astounded to learn from most of these medical records systems, and this is still true today, they’re on their third medical record system now, the systems are designed in many ways so that if you know who the patient is you can cut their data 20 different ways. But to look across patients and take a population perspectives, that was never a real priority, and I think now as systems are getting designed they’re starting to look at those capabilities. But back then it was very difficult to do that and so when we’re trying to do the kinds of work that I was trying to do which was looking at quality of care and so forth, the tools didn’t exist even though we had this wonderful system. We’d print out the records and then sit and do chart reviews, and it just didn’t make sense.
And so I ended up being fortunate enough to be given the job of creating the capability to do these kind of cross functional searches. And so we hired programmers and we designed systems and we built the system as kind of an add on that would allow us to do case finding and do some statistics and so forth. And I started to see the power of this kind of information and we had another grant from at that time I think it was still NCHSR to look at the impact of an automated medical record on quality of care and we did about 18 different projects there, a lot of which involved computerized reminders to doctors, care comparison feedback, things like that. And what we learned is that there’s tremendous power in this information to motivate doctors to make interventions, but it didn’t tell them always what intervention to make, so that if in fact your screening rates for colorectal cancer were not very good, how did you fix them, how did you motivate the patients to follow through on the order.
And that got me into doing research around effectiveness in the plan we have, what were the best practice tools out there. Now I had a group of motivated doctors to change and I didn’t know how to help them do, and so that got me into effectiveness research and the limitations of data for doing that became apparent. Even medical records are not very good for measuring outcomes. Great for process, especially automated records, but not really for outcomes.
And so this is a long kind of way of saying that as I progressed I began to see the real importance of knowing something about how to measure quality, knowing something about what tools were out there to improve it, and needing the evidence base for that. And then all of that resting on data and being able to have the right data and to be able to get access to that data and to manipulate that data.
And so that has led me in a lot of different directions in the course of my career because there aren’t a lot of people I’ve learned out there who know a lot about how to measure quality and a lot about data systems, and so I get called upon a lot to be on a lot of committees, and I’ve spent a lot of time on committees, not just this one but was on the NCQA Committee on Performance Measurement for almost eight years as well. And so that has gotten me very deeply into quality measurement and understanding a lot about the limitations of our existing data and data systems for supporting the measurement of quality and the use of that information for improvement as well.
And so that’s really where the report and a lot of what we focused on in this workgroup came from is we brought in people who were trying to measure quality in all different sectors for all different purposes and asked them what the problems were that they were running into. What were the barriers, what kinds of data was not there that they needed, what kinds of capabilities like data linkage were not available, and to try to catalogue all that and begin to then say well what could be done to try to improve it. So that’s kind of how my career has led me into a lot of the work that was done on this committee as well.
MR. HUNGATE: Terrific. I must say it’s an impressive base, this really is well qualified. I thank you all for sharing that information. I must confess a couple of things. I’ve neglected to say I was on the board of AHQA, American Health Quality Association, which serves the PRO’s, QIO’s, and just came off of that after about a nine year stint, and was a reviewer of Brent James’ early structure of quality improvement as applied to health care. I was at Hewlett Packard and he’d heard that I was talking about health care, quality in health care, and so he sent me a set of eight slides which I sent him back comments, and we talked for a long time ever since. And then Don Berwick was the other one that I spent a lot of time with.
MS. GREENBERG: What strikes me is that there’s been a lot of involvement here in all of the iterations. You can take, I thought when I said that I worked with the PSRO program that dates me, but then you mentioned the MPROBE, actually I can’t imagine that I started at CPHA, which is no more, but which was the original medical audit activity in the professional activity study, so I think we’ve sort of tracked all of these same activities over what a 20, 30 year period really, and we’re in sort of a good position to say where have we gone and where do we need to go. And now more into the information technology standards, etc., so it was really interesting to get that whole perspective. We don’t look that old.
DR. YOUNG: I’ll tell you I taught medical students for the first time a couple days ago at GW and I was struck in fact, I just wanted to say I need to see your college diploma.
MS. COLTIN: I had a similar experience. You were mentioning Mark McLellon, I precepted him when he was in the Harvard MIT Joint Program, and he did an internship with us around medical informatics because he had an interest at that time.
MR. HUNGATE: It’s a fairly small world when you get right down to it. You’ve got to run to your 10:30 --
DR. YOUNG: 10:30, yes.
MR. HUNGATE: I don’t want to drive you off, but let me make one little quick statement before you go. I think that I’ve got a tenure, and I want to make sure that when I step down at the end of my term that I have an accomplishment. And I’d like to kind of think of that accomplishment this way, that the Committee had a common sense of where the system needed to get to, and had enough sense to do meaningful things in the short term that helped HHS live to its role to achieve that. So that’s kind of the charge that I see that we have. I think we’ve got tremendous resource in your collective, in our collective knowledge.
And the other thing that I think of is that I think of us as creating a virus, we don’t have a lot of bureaucratic power but if we believe in information then information has its own life. And so I think that that’s the way I view what we’re trying to do, we’re trying to generate ideas that have a life that works in the system and makes some change in the way we want it to change. So that’s my charge. There’s more to it than that but that’s kind of what I think we’re trying to do.
Agenda Item: Overview of the Issues - Subcommittee and Staff
MR. HUNGATE: Now in terms of where we are now, I’d like to start really by leaning on Kathy to help us get up to where we are now, because I looked at this charge as it came out originally and couldn’t do much with it. I didn’t know where to go with it, and I think you’ve got, you shared a letter that shows more behind that than that. So why don’t I ask you to bring us up to date, give us a sense.
MS. COLTIN: Well this workgroup was actually formed in 1999, early 1999, and one of the practical considerations, and I can appreciate that it may be even more important now that you only have two members of this workgroup. At the time we had four but the other three were either chairing or co-chairing other committees or workgroups and everyone had very limited time to invest in this particular workgroup, we really were not able to have separate meetings. I think we had two over the course of the four years. And so we tried to accomplish our work through the full Committee by instead of having meetings to take testimony separately we would put panels on in the course of the full Committee meetings. And that obviously took a lot longer because you’ve got two hours at each of three meetings a year, and so to get six hours of testimony that would take one day took you a year. And so what we tried to do is focus on a very narrow kind of question and say well what could we really try to accomplish and to identify other areas where we could pursue our goals through the work of other subcommittees.
So the first two items that are on this work plan about the Medicaid agencies and the delivery of post acute care are really reflections of the input from members of this workgroup to the work of the Populations Subcommittee on the Medicaid Managed Care Report and on the Functional Status Report, which really kind of took the place of post acute care and the continuum of care, the real issues there that we identified in terms of data needs in that sector or around tracking functional status and being able to measure functional status.
And so while they’re listed on this work plan they really were accomplished through the work of the other two committees, of the Populations Subcommittee rather, the other two projects, whereby we made sure that within the testimony that was being taken that the issues of quality were brought out, that we specifically asked about data limitations of Medicaid agencies to measure quality of care. And so while testimony was taken on a broad range of issues about data we definitely took testimony focused on quality as well, and we did the same thing with functional status and its applicability to measuring outcomes in particular. So that really was not intended to produce a product from this workgroup, it was for this workgroup to contribute to the work of the other two.
The area where we really focused most of our time was on numbers three and four, and to a lessor extent I think five. And I honestly don’t think we really did much with six at all, we did a little bit, but not very much. Three and four were really the areas of identifying the gaps in the data that were needed to measure and improve quality, and we used the recommendations of the President’s Advisory Commission on Consumer Protection and Quality in Health Care as a framework for doing that. They had made a lot of recommendations about what needed to be done to measure and improve the quality of health care and cutting across all of those recommendations, or almost all of them, was a need for better data and information to support those recommendations, so the implementation of many of those recommendations depended on having robust information systems and having the right data and the right ability to link data and to analyze data and so forth. And so while there were many other groups out there that were developing quality measures and while one of the recommendations of that Advisory Commission was to create this national quality forum to actually review and endorse measures and move us toward greater standardization among some of these measures so that we weren’t placing an undue burden on the health care system to measure the same thing five different ways, that our unique area to contribute to that effort was to say ok, you’re out there trying to measure it, but you’re hampered by the unavailability of the data that you need or the inadequacy of the systems to support analyzing the data that do exist. So that really was I think what was imbedded in number three.
And number four really would be the outcome of that investigation which is to make some recommendations about what needed to be done to enhance the data that were being collected and the capabilities to use those data. So the report I think that we envisioned really was intended to address these items three and four in the work plan primarily.
One of the things that became apparent early on was that one of the best ways to do that was through what’s listed here as number five, which is to say ok, we have a lot of work going on in the area of standardization around administrative data, what can we do to sort of inject ourselves into that process to make the needs of those who are trying to measure quality of care known to those who are designing standard transactions, for example, as well as making sure that the standards that were at this time sort of out in the future around electronic medical records or PMRI would adequately address concerns about quality and the ability to measure and improve quality. So back in 1999 most of the work of the Standards and Security Subcommittee were on the administrative transactions and then more recently they had moved into the PMRI, but the idea was they would be making recommendations to the extent that we could get their recommendations to reflect some of what we were learning in our needs, we would try to do that. And we tried to do that in the Medicare Managed Care Report where we had some recommendations about what they should do and we tried to do it in the Functional Status Report where we had some recommendations about what we thought Standards and Security should do. And I think we need to do that in our report about where we think things need to go in terms of the next iteration of the administrative transactions as well as making sure that the PMRI recommendations adequately reflect some of what we see as the needs that need to be addressed. So I think that three, four, and five kind of go together in many ways.
Six, which was assessing the strengths and weaknesses of the national public and private surveys, we did take some testimony in that area, particularly around the private surveys, we heard a lot about caps and the Medicare Health Outcome Survey and some of the other private sector surveys. We didn’t do as much around talking about the national public surveys. I think the only areas in which we address that at all were probably in the letter that we wrote regarding the National Health Care Quality Report in some of the gaps in areas that needed to be addressed which reflect back on the adequacy of some of the national surveys to support that kind of reporting. And I think that’s an area that you might want to think about in the future, that really is an area that I think we didn’t address as completely as we could have or should have over the time period. I think the provider surveys are a great opportunity that we really haven’t looked at carefully in terms of what could be done there, particularly even in thinking about data linkage capabilities and so forth around some of these.
So that’s a little bit of history in transition. Do you have questions about that?
DR. STEINWACHS: Where do you think we should go from here?
MS. COLTIN: Well, I think that if, I think we at least want to try as much as we can to, I think we’ve closed out one and two here, I think we want to try to close out three, four, and five by putting together this report. Six may or may not be on your agenda for the future, I think we can say a little bit about it in our report, but that’s something that I would put to you as a possibility of something you might want to pursue in the future. I noticed when I read the summary of your discussion at your last meeting that you brought up the distinction between looking at quality of health care and quality of health, and clearly this charge and this particular report and the testimony that we took were focused much more narrowly on health care. And obviously there’s a lot more that needs to be done to improve outcomes and health overall than just focusing on the health care. But that was the focus of this particular set of work and so I think that number six in particular is a nice platform as a starting point to think about broadening that perspective, because the national surveys obviously provide a lot of information, not just about health care, and now I’m going beyond the provider surveys to consider, the NHIS and many others, to think about other quality issues in health and how we know or learn about what some of the other barriers are to improving outcomes besides the problems in the health care system itself. So that would be sort of my recommendation is that phase two might be to take it from just health care to a broader concept of health and that maybe this number six is a nice bridge to doing that in many ways. Not that you should limit it to the surveys but that that might be a starting point.
MR. HUNGATE: Another related question. I don’t recall from each of your individual comments how many of who are from your agencies were part of this discussion. We don’t have any carryover in terms of the NCVHS members, what carryover do we have in --
MS. COLTIN: Well, Stan has been involved in the workgroup for a long time, and Gail has been for a pretty long time, too. I don’t know if it was all the way back to ’99 but I think it was, wasn’t it? Were you involved --
DR. JANES: Actually this workgroup relatively recently but I came from a populations group --
MS. GREENBERG: Of course I go back to 1949. Well I’m trying to think who, Stan really has been with this activity, who else, were there other staff involved?
MS. COLTIN: You know, not really, we didn’t have a lot of staff support from the other agencies in this workgroup. We not only had members who were heavily committed in other areas but we really didn’t have a lot of staff support, and then Stan correct me if you think I’m wrong, but it was Stan. So that was another limitation in how widely we could branch out into other things, we really did have to be pretty narrow.
MR. HUNGATE: I’m familiar with these kind of committee arrangements. I am inclined to make no distinction between committee members and staff who are present in this committee, to say that as far as this committee is concerned everybody is co-equal. Is that inappropriate or is that entirely appropriate as far as these --
MS. GREENBERG: I think given the fact that it’s really unlikely you’re ever going to take any votes, perfectly fine.
MS. COLTIN: And you can’t break a tie.
DR. STEINWACHS: We go outside and duke it out.
MR. HUNGATE: As far as voting, that’s the full Committee. Anything that we put together, that’s to the full Committee.
MS. GREENBERG: Exactly, and in fact the more recent rules around federal advisory committees really only apply those rules to the full committee provided that recommendations and everything of a subcommittee get aired at the full Committee which they always do, and I have always felt that the confluence or whatever of interested and knowledgeable members with interested and knowledgeable staff is what makes the Committee work. So I think that’s great.
DR. JANES: Has John Lumpkin stepped away from the workgroup at this point?
MR. HUNGATE: No, I asked him specifically about how he felt and he said he cannot come, but he still I think wishes to be a part of it.
DR. JANES: Well we at least have three spots for committee members at this point in the workgroup.
MS. GREENBERG: Well, I think it would be great if we could get at least one more member and we have several, I don’t know if any of the current or new members are interested but we do have several new positions that will be coming up, and whether any of those will be filled by people with interest and expertise in this area --
MR. HUNGATE: I have some experience in recruiting, and will use the full Committee meetings to advertise our activities a little bit. We’ll see what happens.
MS. COLTIN: It would be I think really wonderful if you could find someone who is on the Standards and Security Subcommittee who would be willing to also be on this workgroup to create that kind of link, and perhaps someone who’s on the Privacy Subcommittee as well, because a lot of what you get into here when you start talking about things like data linkage and even surveys, because depending on the type of survey, how you identify the sample frame for the survey can get very difficult in terms of thinking about protecting confidentiality. One of the surveys that NCQA, for example, is trying to promote within the managed behavioral health care community is this ECHO(?) Survey, and there are a lot of issues around, how do you identify what sample frame of people who have used mental health and substance abuse services and yet stay within the HIPAA privacy regulations. We know that that’s an area of potential weakness within a lot of managed care plans, it’s an area when you start looking at where are some of the complaints, or where are the issues around access that mental health services are an important area, and yet when you start thinking about how do we look at that, how do we evaluate quality of care in that area, how do we get the member or the consumer or the patient, whatever term you want to apply for the context, involved in that. You get into issues of privacy and whether it’s identifying the sample for a survey or whether it’s doing a medical records audit for quality measurement. So having someone from the Privacy Subcommittee who has an interest in quality issues would be wonderful as another linkage to that committee I think.
MS. GREENBERG: I mean I will say honestly, well I mean it’s Privacy and then you also mentioned Standards, that one of my frustrations throughout the HIPAA process, Gail certainly knows this as well as others, has been that the focus was much more from, has always been much more on kind of the message format standards than really the content standards. And as Kathy said one of the goals was to identify some of the key information, we asked and I saw that as we went through this but I remembered a lot of it too as I went through Susan’s summary, but always ask people, Kathy always religiously asked people when they came to testify about administrative data, if you could have just some, a few additional pieces of information to make administrative data more useful for quality, what would those be. And we got some fairly consistent responses, and yet to try to get that into the agenda of the standards setting was difficult, and that’s really what led me, Don, to kind of working in, developing the Public Health Data Standards Consortium as a vehicle for trying to improve standards that were important for public health by using the HIPAA standards but finding a way to kind of enrich them. Because it really hasn’t been part of the mainstream standards work, and I haven’t given up on it, and I think now as we’re looking more at electronic medical records as well and has really has as Scott said has really gotten the attention of the Secretary and suddenly there are like 24 domains that have been laid out here by the Consolidated Health Informatics Initiative, and the Secretary would like standards adopted in all of them in the next six months. So there’s a motivation there but again, the extent to which people are really looking at do these terminologies capture some of the issues like functional status and other things that we have identified are important, still for quality assurance, is still not really the focus. So I think that any role that we can play in that is --
MR. HUNGATE: Well, it seems to me that that’s one of the functions the report has to then do, is stress those issues.
MS. COLTIN: I think also, I was telling Bob before the meeting about a conference call I had last evening with Arnie Milstein and a woman named Kathryn Brown, who is with the National Partnership for Women and Children, and they have really decided this is the next, Arnie as some of you may know is one of the major driving forces behind Leapfrog and knows how to get, how to mobilize the support of the purchaser community around key issues in quality. And the issue they have now chosen is getting additional data elements into the HIPAA standards, and they have said --
MS. GREENBERG: Did you tell them about the --
MS. COLTIN: I told them, I didn’t tell them all of it and I put them in touch with Bob, but I gave him some advice about how to deal with some of the committees, and in particular the Content Committee is, the NUBC and the NUCC. Because I said to him you know there’s a major disconnect, I served on the NUCC for two years so I feel pretty comfortable saying this. The people who attend that are the people who are dealing with paying the bills, and yet they’re there representing organizations that have other interests besides paying the bills. So the representatives, whether they be from United Health Group or from Blue Cross/Blue Shield Association, are there representing organizations that actually spend a lot of time using their data to measure quality and try to improve quality.
MS. GREENBERG: You wouldn’t know it, though.
MS. COLTIN: People who are sitting on those committees don’t even know about a lot of that stuff, there’s disconnects within those organizations and those associations, and so what I basically said to him is here are the people in each of those associations you need to talk to and you need to tell them who their representative is on that committee and what they need to do. So I’m kind of helping him from the outside but I think also I pointed him here because I said that some of what they’re asking for is absolutely in line with what we heard in our testimony and we could make recommendations that could be quite supportive of what they’re asking for. They have suddenly woken up to the fact that with all of the self insured employers out there, they are in fact as purchasers paying those bills, and they have not had a seat at the table as payers in the same way that the health plans with which they contract have had a seat and a representation. And they are basically saying we want these data and we’re not going to pay the bills unless you give us these data, we need to be represented as payers with needs. And so they’re going to try to start injecting their way into the X-12N committees and NUBC’s and NUCC’s to try to represent what they need from these transactions and they’re principle focus is quality right now. They’re saying we’re paying all this money and we want to know what we’re getting for it and we want to be able to measure quality and so --
MS. GREENBERG: We’ll suggest that they connect with the Public Health Data Standards Consortium --
MS. COLTIN: And who is the person --
MS. GREENBERG: -- representatives on both the NUBC and NUCC. I’m actually the federal representative on the NUBC and Bob Davis from New York State is the state representative on the NUBC. We have actually got agreement from, we’re supposed to represent public health and research on these committees and we have agreement that we can use them, the condition codes for some of these purposes. So --
MS. COLTIN: And what’s it called, the Public Health --
MS. GREENBERG: The Public Health Data Standards Consortium. I can send you something about it, and then --
MS. COLTIN: Yes, if you can do that and who the members are --
MS. GREENBERG: And then we have two representatives on the NUCC, too, Steve Steindel from CDC --
MS. COLTIN: I gave him Steve’s name because I knew Steve was involved with that.
MS. GREENBERG: And Walter Suarez(?).
MS. COLTIN: From Minnesota?
MS. GREENBERG: Yes.
MS. COLTIN: Did Steve take Denise’s slot?
MS. GREENBERG: Yes. So, in fact a group came forward to the NUBC at the last meeting and they wanted to add like hundreds of data elements which was dead on arrival, but they were told to work with the Consortium representatives. But it’s true, the people who are making the decisions about the content have, they’re very, very reluctant to add anything that doesn’t, that isn’t absolutely needed to pay the bill.
MS. HUNGATE: Yes, I understand that.
MS. COLTIN: And I can tell you that the perspective of a lot of the purchasers that I heard from, and I know this was true of those that participated on the Committee on Performance Measurement, was that you’re never going to move, whether it’s health plans, hospitals, providers, to adopt electronic medical records unless they can make the business case for it. The easier you make it for them to not have to report certain kind of data on these standards the easier it is for them not to adopt those systems. So if you start adding some of the clinical data to the administrative standards, they’re going to have to come up with cost effective ways of collecting that data and populating those claims, and that will create an incentive for them to adopt electronic systems within hospitals and ambulatory settings. And so they said you’ve got to push really hard on this not just to get those data and to get them into those administrative transactions so that we can use them, but because doing so helps to create the business case for investment.
MS. GREENBERG: That’s interesting because you often hear kind of the opposite that well, these are data that you should get from the medical record, they shouldn’t be in the administrative, but how do they think they’re going to get from the one to the other. You’re not going to get all these people’s electronic medical records.
MS. COLTIN: You can’t get access to those electronic medical records very easily, so by putting it into the administrative transaction they have to create the capability to pass the data from the electronic medical record to the administrative record, which many of them have already done, those that have it have done it because they recognized the efficiencies that they can achieve in their coding and billing offices if they can populate these --
MR. HUNGATE: Let me interrupt and try to help my own thinking, clarify. I need to have buckets to put things in so that I can stack them. Is my structure of saying there are survey data, there’s administrative data, and there’s clinical data. Is that complete? Have I left anything out?
MS. COLTIN: Those are the ones that we focused on. It depends on how broadly you want to cast those buckets or whether you’re a lumper or a splitter, so like where do you want to put data and health statistics. For example, birth records, death records --
MR. HUNGATE: There’s going to be some overlap. I’m comfortable with overlap.
MS. COLTIN: You can make them their own category or you could lump them.
MS. GREENBERG: And there are clinical data, it also is a question of whether you’re trying to put data or records, because administrative records have clinical data in them already, I mean they have diagnoses, they have some real clinical data in them. And then administrative data have clinical uses, and of course need to be in the clinical record, so it’s a question of whether you want to come talk about the records or the data.
MS. COLTIN: I tend to think of that third category as patient medical records, so that you’ve got survey data, administrative data, and patient medical records. And then I think of administrative data pretty broadly to include not only the types of transactions that are covered under HIPAA but a lot of the other types of data like birth records and --
MR. HUNGATE: Is there a good set of definitions that is universally used for those separations?
MS. COLTIN: I’m not sure there is.
MR. HUNGATE: Well then I think there needs to be, and so I think, to keep me straight I’ve got to have something there. And one of the reasons I think I want it is that I think in order to make the administrative data meet all its needs is that it’s going to need some clinical things so that you can risk adjust admissions, so that you can risk adjust payments, but those things are needed in administrative information to do those functions.
MS. COLTIN: And some of them are there and some of them aren’t, and some are there but not required, so they don’t get used --
MR. HUNGATE: To me that’s something that we should define and if that’s something that people say that’s right then let’s put it down and say it.
MS. COLTIN: And I think when you go back over the testimony, that’s the area, we took testimony on all of them, I mean we took a lot of testimony about the inadequacy of patient medical records, too, but the area where I think there were the most deficiencies and the most barriers identified were in the administrative data. I mean I think there were certainly some around surveys, particularly around lack of standardization at the time that we began and actually a lot of progress has been made in that area since 1999 around standardization, particularly in the area of caps, for example. But the administrative data are still a problem. On the medical record side there’s a lot of progress being made and the recommendations --
MR. HUNGATE: It seems to me that we ought to be in our minds saying we think this is where administrative data should go, this is the limit. We shouldn’t try to push it beyond that because that’s going to compromise some other things. Is that a fair objective, at least to articulate?
MS. COLTIN: I think we want to identify the areas where it could compromise and the area where it could stimulate, because I think the example I was giving before where you can say if you’re going to require the incorporation of some kinds of clinical data, additional clinical data into the administrative transactions, that will stimulate the adoption of these electronic systems because it will be too burdensome I think for hospitals to try to get that data into billing records using manual types of processes or systems.
MR. HUNGATE: I agree. But I’ve come up with one label, those things needed to risk adjust would be acceptable, are there others? What I’m trying to, maybe we can’t come to an answer in this discussion, but I feel like maybe there’s --
DR. STEINWACHS: -- risk adjustment sort of on the front end, sometimes some of those same things are sort of the outcome indicators on the back end. Find a disease and then you would think of some of those lab results, some other indicators as things that will tell you about ongoing way, and so I would probably put them on both ends so that the third is really the appropriateness of the care and to what extent are you looking for certain clinical indicators to tell you whether or not this person got the right sort of service even though if it’s diagnostically, but does it fit in that person’s, that’s a kind of a risk adjustment.
MR. HUNGATE: And that gets you into the whether you’re using the information for external assessment or whether you’re using it for internal improvement.
DR. STEINWACHS: The concept ought to be useful for both.
MR. HUNGATE: But there would be a lot of things that will work for internal improvement that are not necessary for external assessment, that in fact it gets too burdensome to do that much, it seems to me.
MS. COLTIN: One of the big problems and struggles you’re going to have is that most of the research has indicated that the biggest problems in quality occur at the intersection of hand offs from one provider to another, one setting to another, and that the systems, and that if you said for example well we don’t want to put lab results into an administrative data set because that’s something that really, if you had an electronic medical record you could use that for internal improvement. But in fact it’s these bridging from one setting to another, that the internal system isn’t going to help. The only electronic mechanism we have right now that bridges is the administrative transaction where you can put the data together that’s coming in from all of these different settings and get --
MR. HUNGATE: Well, that’s today’s limitation, but if we’re talking about an NHII --
MS. COLTIN: True, then everybody’s connected --
MR. HUNGATE: Then we’re visualizing a different, so it seems to me that we have to define what we’re expecting from NHII and what we expect from the administrative data, and be clear.
MS. COLTIN: I agree with that, but I think you need to think about a short term and a long term goal and strategy, and I think that the NHII is wonderful, but I guess as someone who’s been listening for 20 years or 30 years about the electronic medical record being around the corner, I’m a little bit --
MR. HUNGATE: See I have a belief structure that says the reason that it isn’t here is that the profession that practices medicine is not accountable for its results. And if you began to create accountability modules, outcome measures that were valid, that you’d flip the incentive. Now that’s a premise.
DR. STEINWACHS: I tend to be, I don’t disagree with Bob, I think accountability is key, but I’m down on sort of Kathy’s skeptical side that legally accountability is defined pretty narrowly for health practitioners. And one of the hopes of managed care, one of the things I loved about HMO’s is the idea that you have an organization that has accountability for an enrolled population. But as soon as you move away from that enrolled population accountability then the drivers force seems to me this sort of what I have in my computerized record. So I think we ought to keep both of those in mind.
MR. HUNGATE: I’m not uncomfortable with that but I think that we have sub optimized information by the focus of the plan as opposed to the provider. And we have tended to create multiple conflicting information systems which are plan driven in demand on the provider. And that if we were clearer in terms of what we expected from clinical information at the provider level and made the good transition to the administrative level, that we might speed that, the muddle slows the development, and I believe in getting venture capital interested in --
MS. COLTIN: I think, I mean I agree with you about that but I think, I don’t think anybody who started measuring at the plan level saw that as the end. They saw that as the means to the end, because it would motivate the health plans to take that data, in order for the health plan to improve, they aren’t the ones in the office delivering the care and that the health plan had to start collecting data down at the level of the provider. The problem, and I thing this goes to what Don was saying about the legal definition of accountability, is that at the provider level its who’s responsible for whole person care for the patient. And we’re in an environment where patients are more and more moving even out of the HMO or at least the HMO can say well you got this from the specialist, this and that, they didn’t talk to each other, you need this, to a mode of a PPO where there isn’t even a primary care provider, there isn’t, it’s like you go do it and then each provider you see is accountable for what they did but not what the other guy did, and so I think you end up being even more fragmented with more silos of information. And the administrative systems at least provide some vehicle for pulling that information from all these different silos and looking at it. But who’s going to do that? Who’s going to pull it together and look at it? If you’re an indemnity plan, there’s no accountability, there’s no motivation to do it, except maybe Medicare as the biggest indemnity plan out there in fee for service --
MR. HUNGATE: The current issue of health affairs argues that the government is the biggest purchaser, that it has the power to set the way in which these things are to be done, we’re supposed to advise that particular organization --
MS. COLTIN: My point is they’re the only ones doing it. They are at least trying to do it, they’re pulling the data from the various sectors and trying to do it, but they’re only addressing the problems in one population segment, and unless you get some ability to look across the information systems for people who are not 65 or older or not disabled, you’re not going to improve care at a time when you can prevent greater expense down the line for Medicare.
MR. HUNGATE: Well, I understand, I agree with that.
MS. COLTIN: So I think that the issue I’m concerned about is not even so much that these data systems exist at the levels of the individual providers in their offices with wonderful electronic medical records, but who is looking across that data at the patient level, not the visit, encounter, episode level to say is this patient getting what they need. And by putting pressure on an HMO or point of service plan to do that, to say you own responsibility for this population, they’re at least looking across, they’re saying this patient has diabetes, are they getting what they need. I don’t care if they’re getting it from the PCP, from the specialist, are they getting what they need. I think in a fragmented system you don’t necessarily have that, it’s like who’s accountable.
MR. HUNGATE: Well, I think we need, am I on the right track? Is this a fruitful and important area of discussion?
MS. GREENBERG: This is a system wide problem that she’s describing, and that’s what you said is your goal.
MS. COLTIN: Right, I mean we have an NHII, it still begs the question who is going to look at all that wonderful information that’s now linked, and are we going to say it’s going to be the patient, that the patient is going to have all that information, they’re not going to be able to interpret it, they’re not going to know what should have happened and didn’t, I mean there’s got to be some mechanism for once all that wonderful information is available to be exchanged, who’s looking at it across the whole patient. That’s part of the reason I focused a lot on administrative data is that at least, right now, if in fact you have an accountable entity like a health plan, you can do that.
DR. JANES: You know Kathy I think there’s, you’ve made a lot of interesting points but one that particularly resonates with me with my public health hat on, as most of us around this table, a hat most of us wear, is that one of the things that has alarmed me is that in public health we really glommed onto managed care and that group staff model, mode, when it came down the pike because for those of us who were data oriented it did give us that comprehensive package. And we have continued to adopt that focus even as that world has gotten smaller and smaller and smaller, and the point that you made is that the world is moving toward that PPO world, where the data systems are completely different. And yet one of the conversations we have internally at CDC all the time as we take our cooperative agreements and our programs and continue to send them out to the Harvard Pilgrims, the Group Health’s and the like, is that folks this is not the real world. But of course when we move over and we look at Blue Cross/Blue Shield and the like we’re so appalled at what we see we say well we can’t do what we want to do over here, so we’re back with the same little group.
But to the extent that, I mean Health and Human Services is public health and we are making recommendations to that group, I think there are a number of points to be made just around that issue that the data issues, the quality issues that we’re talking about are in fact becoming a problem or becoming more acute and not becoming better because in fact the market, the world is moving in a very different direction.
MR. HUNGATE: Which means your strategy is going to have to change, which is what I’m trying to say what is the next strategy. I think the staff model is only going to decline.
MS. COLTIN: Well, the staff model’s almost dead --
DR. STEINWACHS: It will be interesting to sort of not only this group talk through but think of is there testimony we want to get, the different strategies, I’m sort of the group, Kathy may and others in the same place, is sort of figuring you need to try, drive from the administrative to get all the other things like computerized patient records and so on, needs to come out of probably pushing through the administrative channel.
MR. HUNGATE: Well I probably can get the consumers involved in the clinical side. I probably can’t get them involved in the administrative data.
DR. STEINWACHS: And I wouldn’t suggest that, it’s like a war front, since we’re into war these days, you don’t describe with one attack, and so I would say that we ought to be pushing in several directions. But I think we ought to explore those means of push through the administrative.
MR. HUNGATE: I think we’re in full agreement.
DR. STEINWACHS: What does it mean to push from the consumer or patient side to say this is your personal record and we don’t know a lot about --
MR. HUNGATE: I think that does mean that we talk about a personal electronic record and the institutional electronic record.
DR. STEINWACHS: Well there’s a personal record and institutional record and then there are the administrative data systems. Just two things that maybe a little bit off, but it seems to me that when you talk about accountability, legally accountability for almost all the health care sits within states, and so it would be useful I would think for this group, my desire to get John Lumpkin to sit in a discussion about, is what are we thinking that helps take this to the level that states could do something with it. Now I recognize that states do --
MR. HUNGATE: One of the other hats that I wear is I’m the chairman of the Group Insurance Commission of Massachusetts, and we buy health care for all the employees beneficiaries, and we do have needs to make changes, so we’re a place that could do something.
DR. STEINWACHS: More importantly, you’re the home of the Medical Practice Act, you say what physicians can and can’t do, what nurses can and can’t do. You license the hospitals, you license the nursing homes, and so there’s nothing in that health care world in a sense you can’t say that the state really has responsibility for, because there are very few really federal overrides on that. And so it seems to me one of the things to think about is how do we enhance the capacity of state health departments and the state to do something. That doesn’t mean they’ll do it but at least --
The other one you were talking about, buckets, I like your buckets idea, so stacking buckets is difficult sometimes, sort of sit inside it. There is sort of a fourth bucket, just to point out at least, and that deals much more with the local environment, and that may be local health care resources, it may be that, or efforts to start surveillance of environmental exposures, some of the states are doing that with CDC leadership and support. And so there’s a context in which this is occurring that I think all of us know is important, it’s just --
MS. COLTIN: And that’s particularly important as you move from health care to health.
DR. STEINWACHS: Is it fair to think of those as risk adjusters in a way?
MR. HUNGATE: Well, they range from risk adjusters to the issue of capacity to provide, capacity response, so it’s a risk the population face, the risk adjusters, but it’s also the capacity and I think more and more if you talk about the diversity issues we were telling you some disparities in health care, you need to know something about what that environment is if you’re going to understand why it is that some people don’t get some things, and partly it’s because of those local structures.
MS. COLTIN: It’s not just dealing with capacity to address under use, it’s what’s the influence of over capacity on over use as well because supplier induced demand is a fairly well documented phenomenon.
MR. HUNGATE: We build it, we need to fill it up, if they don’t come we need to fill it up.
MS. COLTIN: We don’t like empty beds in hospitals.
MR. HUNGATE: Which leads me to another observation from my past, that when we sold EKG machines, and the separation of the responsibility for the product back to the manufacturer, and the use of the product by the profession, is a bad design system wise. You need to have those two mesh in some way, and it’s rampant through the system, pharmaceuticals have the same problem. All of that, I’m thinking of accountability for the result from the system, which is across a bunch. And that’s got to be priority for HHS, if we’re spending all our money in the wrong places, we ought to begin to understand it. At least that’s where I’m coming at it.
MS. COLTIN: I think that’s a really important point, that context and not just capacity, I think the other factors like the environment are important contributors to health, but we’re coming up against this in a very, I always like to think concrete, of generalized from the specific rather than, what is it inductive versus, but we’re dealing with an issue right now where a new quality measure that’s being implemented for health plans is one looking colorectal cancer screening. It’s been an issue for states, whatever, the main issue that we’re coming up against in terms of trying to improve is capacity. Capacity for doing these sigmoidoscopies and colonoscopies, that even in an area like Boston where we have so many doctors, the waiting time to get this test is like six months in many areas, or more, and what ends up happening is patients call and sometimes they’re told well we don’t have the schedule yet that far out, call back, and then the patient doesn’t call back --
So capacity is really becoming a major barrier in that area, so I think Don’s absolutely right, you have to think about those kinds of factors as well. Do we have the data to recognize quality problems in terms of our ability to deliver what we know needs to be --
DR. STEINWACHS: Just one other sort of piece which I’m not sure where it fits, you see it coming up mainly around issues of emergency medical services and designation of trauma systems and so on. It’s when we talk about the system, sometimes we think about the system that you and I sort of live in as a patient, but if you talk about a trauma system, that may be a state wide set of structures and designations, neonatal intensive care, and I don’t know where that sort of gets captured in the sense of this context. So it’s just worth again, maybe that’s coming back to this issue of is there a state level sort of context that eventually becomes part of when you say what do you need and understand what’s going on in America.
MR. HUNGATE: The other piece of definition that I want to get, another set of buckets, is the one of the function of whether it’s assessment, improvement, I use surveillance or you use tracking, to me those are different --
DR. STEINWACHS: Only during war time do they use surveillance, otherwise they use tracking.
MR. HUNGATE: You’re always at war against infectious disease, for instance. And that’s what I was thinking of, those things where you want fast response to minimize impact as different from those assessment things that are slower in their response. And so I was trying to think about what’s the structure, who’s the taker, who’s the user, who’s the actuator, and our task is to make sure we understand how those mesh. And I feel like it’s too easy to get muddied and think of one when you’re talking about, so that’s my concern.
MS. COLTIN: I think you’re right, I think as somebody who thinks in matrices a lot, I mean my sense is what you’ve identified is the cross cutting, that you’ve got the four buckets of the different sources of information and then you’ve got the uses of information kind of cross cutting, because I think any one of those sources of information can be used to support one or more of those purposes. And I can tell you that in Boston at least administrative data are playing a huge role in monitoring for bioterrorism, I mean we’ve got a huge grant to look at the uptake of symptoms in emergency rooms and whatever, coming out of administrative data to watch that. So there’s a use of administrative data that has nothing to do with paying claims but that is extremely important right now.
MR. HUNGATE: It depends on it being electronic transmission and not the old administrative claims process.
MS. COLTIN: Well most if it now, that’s what I think the standards are all about, is the electronic, it’s moving from the paper to the electronic. I think in terms of institutional claims, the percentage these days that are electronic is really high, it’s well over 80 percent, even in some of the areas of the country that are not known for technology or systems. It’s the professional claims where we’re lagging far behind because many doctors offices, even in eastern Massachusetts, don’t have the capability to submit electronic claims yet, they’re still sending in paper.
MR. HUNGATE: To me this is where we want to go a little bit, it’s getting ahead of our schedule but it seems to me it’s useful background, than to try to work more specifically on the what do we say now in this report so that we have the context.
DR. STEINWACHS: This also, there’s the more recent IOM report that identifies 20 conditions I guess is the, priority conditions, and it seemed to me it might be useful also for us to --
MR. HUNGATE: Which report is this?
DR. STEINWACHS: It just came out, and it’s a follow on to the Quality Chasm and relates to the report cards and so on, but a process went through, they went through with a set of criteria trying to identify the 20 conditions, they aren’t quite all conditions, but 20 conditions that ought to be given priority in quality monitoring and quality improvement as sort of the starting point, it’s not the end place but the starting place, and it seemed to me that would be useful to share in this Committee and has some discussion. Is that also a helpful thing as we think about these issues of what do you give priority to when you talk about expanding possibly the administrative data, or what do you give priority to that you try and build a personal medical record.
MS. COLTON: And a way to dovetail that with what some of the work we’ve already done and where you’re headed to is to think about the matrices that were created for the National Health Quality Report and say if you took all 20, some of those are already reflected there, if you took all 20 of them and laid them out where are the gaps in terms of having quality measures and why are those gaps there, how does that relate to the data systems that underlie them. Because most of the time when you look at why there are gaps it’s because we’re not capturing the data that we need to be able to look at those. It’s not always that there just aren’t quality measures there, it’s that people may have tried to create quality measures there, and they’ve run up against one of two things. Either the data systems don’t support it or there’s no evidence base for what you should be doing so that you can then create a quality measure based on a clinical guideline or some evidence base. And when you take that down a step further and say why is there no evidence base, often that gets back to the data not being available to measure the outcomes or to assess the process. So it comes up very clearly --
MS. GREENBERG: What is the name of this most recent report, do you know?
DR. STEINWACHS: I read the review on it, the intra view, but I don’t --
MS. COLTIN: I have the executive summary on the computer here.
MS. GREENBERG: And I feel that since your colleague Barbara Starfield is not here today, I need to ask, that looking at these individual conditions, what about comorbidity?
DR. STEINWACHS: Well, all of us love comorbidity, that’s right --
DR. HOLMES: Just as we love risk adjustment.
DR. STEINWACHS: The only reason I raise the report particularly was that I think it is going to try and help, it’s there to try and help shape where people focus initially, and it does then also make you think about examples, as sort of what Kathy was saying, take these against other frameworks and say well, how is this going to work and is this covering everything that should be there.
DR. HOLMES: And there are major gaps that have already been identified with respect to producing the National Quality Report and the National Health Care Disparities Report over against the framework, the original framework suggested by the IOM. It’s very clear where there are huge gaps in data availability or the existence of valid measures.
MS. COLTIN: And we had commented on that in our letter to the agency about that, and we did get a response to, very timely response, which is unusual.
MR. HUNGATE: I’ve made you all sit here and work kind of past the break time that was scheduled here. Do you want to take a short break now and come back?
[Brief break.]
MR. HUNGATE: Ok, let’s proceed. I would like to get as soon as we can to thinking about what recommendations we want to make in a report because I think my sense of watching earlier recommendations that came before the full Committee was that there hadn’t been enough work on boiling down to tight recommendations at the Subcommittee level, so I think that’s where we want to spend our time. And I think that that presumes that everyone has read the background paper and the data issues and obstacles in measuring quality of care and probably we ought to just open for comment, I don’t think we want to try to work through line by line, give you a minute to look through the background paper first and raise questions.
MS. KANAAN: And I would especially call your attention to the items on page three, sort of next steps, because there’s some questions that probably precede, I would like to see precede the discussion of the content of the report having to do with the intended audience and the objectives before we get too deep into the content.
MR. HUNGATE: Fine. I’d be happy to have you lead discussion in the areas here. Well, you’ll do it better than I would because you’re more familiar with it.
MS. KANAAN: Well, really, I think that the background paper speaks for itself, it’s just that I think, and maybe it would be stating what is already obvious to the group but you’ve already answered one of the questions I’d laid out there which is do you want to make recommendations at this stage, so you’ve got the answer to the fifth of five bullets. But what is the intended audience for this report, or is it if you’re making recommendations than that assumes this is not just going to be a background paper for the Committee, which was a little bit the way we talked about it in February.
MR. HUNGATE: I was very uncertain about what we could say.
MS. KANAAN: Yes, but now it just sounds like you’re talking about the standard report to the Secretary that would perhaps be the first of two or something like that, if you want to widen your purview of phase two.
MR. HUNGATE: Yes, I think we want to wrap up where we are today, position ourselves for what we think we ought to do next, which will include recommendations for HHS implementation.
MS. KANAAN: So then I guess the objectives are obvious, the objectives are bring about whatever changes you want to happen.
DR. STEINWACHS: So Bob, just to make sure I’m up to speed which I sometimes wonder if I ever get fully up to speed, but --
MR. HUNGATE: Me, too.
DR. STEINWACHS: So the sort of charge work plan that Kathy helped take us through, we’re really talking about a report that addresses sort of three, four, and five there, is that the gaps in data is what we’re saying and --
MS. COLTIN: Some recommendations, which are four.
MR. HUNGATE: Now is it practical within that content to put in the definitional buckets if you will of what we’re talking about?
MS. COLTIN: Absolutely, absolutely. And I think in this particular report we would probably want to talk about limitations of this particular report. We didn’t address that fourth bucket that you’ve just described in this report, but there’s no harm in laying it out and then just scoping the report somewhat more narrowly.
MS. KANAAN: And there are some limitations that are identified, if you go back to page one, the background paper, the third paragraph the work group can find the scope, and that kind of identifies the other limitations as well, the things that were really beyond the scope, so did you agree with those Kathy?
MS. COLTIN: I agreed, my note to myself here is that I agreed with number three, but not necessarily with number one and two, because I think the group certainly dealt with the data limitations around identifying race and ethnicity for the purpose of being able to then assess disparities in care. We took testimony about the fact that we can measure some aspects of quality with existing data systems quite well but we can’t measure it at the level of particular population subgroups, because we don’t have the data on race and ethnicity to be able to do that, nor do we have data on income, education, and other important factors that are needed to examine disparities. So we did take testimony on that, so my sense is that this was never a report about what are the quality issues in these different populations, it was about the ability to identify those issues given the data. And I think we did address that.
MS. KANAAN: I didn’t see that so much in, I mean certainly the Committee has taken a lot of testimony on that through Populations Subcommittee panels and so on, but I didn’t have a sense that that was really very much a subject in the --
MS. COLTIN: You have it listed here under incompleteness, race, ethnicity and the disparities, June 1, number one. And again, this is one of those areas when you say it’s not a focus of this workgroup, it was very clearly our intention as illustrated in bullets one and two under the workgroup, to try to work through the Populations Subcommittee, first the Medicaid Managed Care Report, anything in that report that has to do with quality can be brought into this report because we made that happen, same with the functional status and same with the race ethnicity, I mean I sat there at every, the two days of testimony in February of last year on race and ethnicity and asked specific questions, so even though that project and initiative was being done under the leadership of the Populations Subcommittee we weren’t going to run a parallel effort to look at that in this workgroup, we were trying to bring the quality issues into that testimony.
MS. KANAAN: You might want to think about either just appending the relevant portions of the report or even just integrating some of the key points then because that would enrich it.
MS. COLTIN: Yes, and I think that we were the ones that really prompted the panel session on race and ethnicity that we had where we brought in Janet Fowls(?) and others to talk about, on both sides of the issues, it was a wonderful session about the issues on race and ethnicity in administrative data, and I would clearly highlight that --
MS. GREENBERG: We tried to use that to influence the HIPAA standards process and we finally sort of gave up on HIPAA standard, but we have incorporated that into the Health Care Services Data Reporting Guide, so it was again as I said, we sort of used this public health consortium product to meet some of those needs. But again, that’s for reporting, it’s not the actual bill, and there are people who feel that, this is still this ongoing debate as to whether this should on the claim --
MS. COLTIN: Did you see the Wall Street Journal and the article that Aetna is going to start collecting race ethnicity on administrative data, so there are important developments --
MS. GREENBERG: How are they going to do it when it’s not in the standard?
MS. COLTIN: Well it actually is in the enrollment record --
MS. GREENBERG: They can do it in the enrollment?
MS. COLTIN: Yes, and it’s a conditional field and they’re going to make it a condition of their contracts and they’re going to start collecting it with new enrollees and over time, they’re not going back to the entire base of members they already have --
MS. GREENBERG: When was that article?
MS. COLTIN: I have the article, it’s been a while, that one I might be able --
MS. GREENBERG: And apparently this has come up again at the IOM I think, in the IOM discussions.
DR. STEINWACHS: It’s part of my ongoing education. In framing a recommendation, generally you rely on testimony or written evidence, is there sort of a culture here as to what makes a strong recommendation?
MS. GREENBERG: Well, it’s good if you have --
MS. COLTIN: If you can back up that it’s not just our collective opinion, but that doesn’t prevent us from injecting our opinion --
DR. STEINWACHS: Our opinion, and so it could be our opinion to we’ve gone out to try and document and --
MS. GREENBERG: But we’ve put you on the Committee because of your expertise and knowledge so you could actually make a recommendation without any testimony, but I think part of the, obviously you’re always up against push back of people not wanting to collect any more information or collect certain information, so the more evidence you have that a, others feel it’s needed, and b, that it’s feasible to collect and cost effective, feasible and isn’t going to break the bank always helps.
DR. STEINWACHS: I guess part of what you’re saying is that the more probably focus specific and maybe the more burdensome the recommendation appears to be the stronger the evidence ought to be to get it through the whole Committee no doubt.
MS. GREENBERG: Back in the core health data element recommendations, which Kathy and I lived with for a long time, there was this recommendation that there needed to be some SES measure, that probably the best one would be level of education. That’s not in the standard yet though I think, I’m involved with this Consolidated Health Informatics group and that’s what I have to go to at 3:00 in fact, I’m on the Demographics Workgroup, and we’re selecting vocabulary and I think education level will be one of the elements. So if that’s in the record it’s a lot easier to get it, you can’t get it into the administrative data if it isn’t ever in the health records, but the medical record or the clinical record, so apropos what Kathy was saying.
DR. JANES: Is there a position for that on the enrollment file, education at this point?
MS. GREENBERG: I don’t think there is, it’s not in X-12 but it’s in HL7, ASTM and HL7 both report it.
DR. STEINWACHS: What about English as a second language, is that picked up on --
MS. GREENBERG: Well primary language is also one of the elements that I think we’ll be recommendation, I mean this is really off the record.
MS. COLTIN: I thought that one was actually on the transaction now.
MS. GREENBERG: I don’t think it’s in X-12 either, but again, I think maybe HL7 and ASTM both, I don’t think it’s in X-12.
DR. JANES: Is this for your work on, I’m not following X-12 and I don’t, I’ll just say I don’t understand what those are.
MS. GREENBERG: There are two main standards development organizations, they’re known as SDO’s affectionately, that are most critical in exchanging health information, in developing the standards for exchanging health information. And they’re both accredited by the American National Standards Institute, one is called ANSI, although they’re both ANSI, Accredited Standards Committee X-12, ASCX-12, and it is responsible for the messages to exchange administrative data across banking, railroad, you name it. And then there’s an insurance subcommittee, which is N, so that’s X-12-N. Within the insurance subcommittee is where health resides, and those are the standards that were adopted for the administrative transactions, the most important one from our point, or the two most important from our point of view would be the claim encounter transactions, which is the 837, and the enrollment transactions, which is the 834. The limitation with the enrollment transaction is that it’s basically done by employers who are not covered by HIPAA, so there you have a standard, the people who are responsible for enrolling aren’t really mandated to use that standard but if you are using, if you are one of the covered entities, which is a provider or a payer or a clearinghouse, then you would have to use that standard. So those are the administrative standards and that’s X-12, and in this group X-12-N, and it’s the 837 and the 834.
There are other transactions as well but they are things like payment, they’re very payment oriented, did you send the payment, did we get the payment, things like that. They’re not ones that would be used for quality or for even public health or whatever. So those are the ones that we’ve focused on and I keep talking about this Public Health Data Standards Consortium, what we have done is, and then you have the standard but then you have an implementation guide, so the public health community made sure that the 837 standard, and actually the 834 as well, that’s the claim and the enrollment, that they now can capture race ethnicity as required by the new OMB standards, so all the categories and then you can ask for more than one race and all of that.
But the implementation guide that was approved under HIPAA, for the 837, does not allow you to collect it under the 837. It’s in the standards, you’ve got a standard, but an implementation guide says how are we going to implement the standard for this purpose, and for the purposes of the claims, the standardized claims under HIPAA, they took a subset of that standard, so they’re not implementing all of the standard, plus some elements they’re implementing but with conditions, only if, and that’s like Kathy said, like only if it’s required by the plan, although you’re not really supposed to have conditions like this because this is supposed to be a standard that everybody does the same thing. But anyway, the implementation guides are a subset of the standard.
So what we have done in the Public Health Data Standards Consortium is we have developed our own implementation guide, which is now gone all the way through the whole process, it has to go through all these X-12 committees and everything, and it is for, it’s a reporting standard because it was particularly driven by the fact of the vast majority of states, more than 40 I think, have hospital discharge data systems, which usually they get their data from the claim, where it’s still called the UB92, the uniform claim for hospitals. And that UB92 is really the paper version of this 837, just like the 1500 is the paper version of also of the 837, this ones professional and the other is hospital. But they had been collecting race ethnicity as a state field.
But with HIPAA, state fields go away so now if they start getting, and all of the hospitals are having to go from their UB92 format, which they had put into electronic format, to the 837 because that’s what HIPAA requires as of October 16th. There are no more state assigned fields, so here they’ve got an 837 that doesn’t have, and as implemented through the implementation guide that doesn’t allow you to collect race and ethnicity. And race and ethnicity is a critical field for the state discharge data systems. So we actually went, we actually went to X-12 and got it into the next version of the guide. But that version hasn’t been implemented and it doesn’t look like, I don’t know whether and when it’s going to be implemented, so there were other elements like this, too. They were state assigned fields that are now going away so what we have tried to do is migrate them to national fields and then if they aren’t in the HIPAA guide we’re putting them in our own guide, and that guide has now been balloted and approved. So the states will be able to say to the providers, you can send me the 837, you don’t have to send me some proprietary format or some old format, the same way you bill the payers you can send me that with these additional fields because we have a separate guide that will allow you to collect those fields. Probably this is gobbledygook to people but it is a vehicle for collecting additional information that we haven’t been able to sell to the content committees.
Now there are two content committees, the National Uniform Billing Committee, they’re responsible for the institutional content, and the National Uniform Claim Committee, they’re responsible for the professional content. And really the content of the 837 and these X-12 transactions is driven basically by these two committees. So the X-12 is kind of the format but they take their cue on content on the claim, not on the other transactions, but on the claim from the NUBC and NUCC, and that’s why we said look it, if this is what’s going to control the content we’ve got to have public health people on that committee. Up until 1999 there were no public health research behind those committee --
MS. COLTIN: And I can tell you, I was a lone voice for years.
MS. GREENBERG: In the early ’90’s, the National Committee through Bill Phelps(?), I don’t know if any of you knew Bill Phelps who was a, he just died, but he was a rheumatologist at GW and he was chair of the previous predecessor standard group, he became a voice to the National Uniform Billing Committee because of work we had done in the Committee, and I was his sidekick on getting external cause of injury into the claims, which is also a critical piece of information. So the HIPAA guide allows you to report one external cause of injury, our health care reporting guide allows you to report like ten. The standard will support ten but the implementation guide for the claim only supports one, our guide supports these additional ones.
MS. COLTIN: Does that help your reporting guide available anyway? That give it on --
MS. GREENBERG: Yes, it’s on the Washington Publishing Company, but I’m going to connect you with Bob Davis. But I’m sorry to be proselytizing about this but this is a vehicle really and I think it’s good for you to know about it. If it’s not in the standard at all, then we have to find a way to get it into the standard. But on certain types of elements we even have a way to do that because the NUBC, they’re UB92 format has a bunch of coded information, and they have given us a set of codes that we can use for these purposes, for public health and I would say quality purposes, even though they may not be part of the HIPAA implementation guide they could be in our guide. And it makes it so much easier for provider because they’re basically using the same format, they’re just filling in, which was what they were doing anyway. When they were using the UB92 they were then putting in state assigned fields, so it wasn’t necessarily the identical transmission to the payer and to the state, but it was the same format, it was just some additional fields in that format.
DR. STEINWACHS: Is there a set of briefing material, Bob could use and probably both of us could use it to sort of get us to know what’s in the X-12, in 834, 837, and in these --
MS. GREENBERG: We have, actually the, the Public Health Data Consortium is developing a web based resource center, and it has, it’s going live in June but we have documents and glossaries and things like that that we’ve developed --
MR. HUNGATE: Well for me at least you should assume ignorance on my part and send me anything that you think I should know, and if I already know it, that’s alright.
DR. JANES: Can I ask you a quick question before you abandon this? The one piece about this that I didn’t know was the part about this Health Care Reporting Guide which is interesting. And you also talked about, did I understand you when you said there are state --
MS. GREENBERG: There are not state assigned fields in the Health Care Reporting Guide, what we’ve done is --
DR. JANES: No, I knew that, but are there state implementation guides? You were talking about how the states are getting around the national implementation guides and the national standards and the fact they weren’t allowed using those guides, that guide to collect race ethnicity --
MS. GREENBERG: This is the guide, the Health Care Service Data Reporting Guide is the guide that states can use, if they can get migrate their systems to the new HIPAA standard. A lot of them are going to continue to have providers send in the old standard which is the UB92, so that’s really burdensome actually for the providers, they got --
DR. JANES: So when you talk about the Health Care Reporting Guide are you just talking about the 837 implementation guide?
MS. GREENBERG: I’m talking about an implementation guide for the 837, but it’s for public health and other data --
DR. JANES: Specifically for more public health entities, i.e., specifically for, for example, state hospitals --
MS. GREENBERG: People who want to collect information that is not supported by the HIPAA guide.
DR. JANES: Exactly. And were you talking about a group called the Consolidated Health Informatics Group, is this a product of the Consolidated Health Informatics Group?
MS. GREENBERG: No, this is a product of the Public Health Data Standards Consortium. The CHI is the group within the Department that’s developing message and vocabulary standards for exchange of information at the federal level, exchange of patient medical record information.
MR. HUNGATE: It’s probably a pretty important activity.
MS. GREENBERG: Yes, that was what two, three weeks ago Secretary Thompson basically went out to Detroit to say that the Department was, and VA and DOD, was adopting the standards that were recommended by the National Committee on Vital and Health Statistics, he didn’t say that, and the press release never mentioned the National Committee, but those standards that were recommended through the CHI Initiative were almost identical for the message format standards to what the Committee had recommended in February of 2002 to the Department, and they just were tweaked a little bit, but HL7, LOINK, whatever, they all were what the Committee recommended.
MR. HUNGATE: To what extent do the data gaps and so on that were identified in the testimony get addressed by these sorts of elements?
MS. COLTIN: Well, I don’t know, I haven’t seen the Health Care Service Data Reporting Guide, so I don’t know what additional data elements that were identified as gaps in the testimony are included in the guide, but even so it’s still not 100 percent clear to me how one would invoke that guide in terms of trying to get some entities to report using it. I can see how a state could do that by saying hospitals in our state must report to the state using, but as a health plan for example, trying to look at the quality of care provided to our members, I don’t think we could go back to the hospitals and say report to us using this guide. We’re really reliant on the claims standards.
MS. GREENBERG: True, although if you could contract with them for additional information and the guide could support it, it would be a lot easier for them than if you told them they had to send some completely, send you that information in some completely other form. Now one possibility is some of these stuff could migrate to the next version of the HIPAA guide, another element --
MS. COLTIN: And that’s where I think our recommendations ought to focus.
MS. GREENBERG: We’ll have some experience in collecting it. That’s the thing, we couldn’t wait for them frankly to, I mean the states couldn’t wait. Another element that is supported by the standard, that basically the Consortium got in but is not supported by the HIPAA implementation guide, but is supported in our implementation guide which is a very important element for quality, is that diagnosis indicator, whether or not the secondary diagnoses were present on admission. That’s the whole issue of comorbidity versus complications. And that is supported, California and New York are the only two states that currently collect it, and they collected it differently, so we got agreement, so now there’s only one way to collect it, I can’t remember which way it is but anyway, and it’s in the standard, it’s in the 837 standard, it’s not in the HIPAA implementation guide, but it is in our guide. So I think --
MS. COLTIN: And that I think would be an important one for us to weigh in on based on testimony that we took, that it is extremely important.
MS. GREENBERG: Yes, Lisa’s written about that. And again, it’s --
MS. COLTIN: I think race ethnicity, I think that particular data element is critical, and I think being able to get, if you ask people if you can have three additional pieces of information in there, what would you want. I mean the other one that was mentioned again and again was lab results, because it’s just so important both for outcomes measurement and for looking at continuity of care and handoffs, did they get followed up. What percent of abnormal tests actually get followed up? And you can’t do it unless you can identify the abnormal test results.
MR. HUNGATE: Is there a recommendation here that I should write on the board?
DR. STEINWACHS: I think so, I think we’re finally getting there. Now you just need a facilitating mechanism.
PARTICIPANT: Marjorie, do you expect that all 40 states that support hospital discharge registries are either have or will be using the Health Care --
MS. GREENBERG: That’s our goal, right now there are only probably three or so --
PARTICIPANT: How long has it been out there?
MS. GREENBERG: It hasn’t even been published yet, but it just was approved maybe a few months ago. Do you all know about the claim attachment standard? This is relevant to the lab issue, that’s why I mentioned it. Because there is, one of the standards that hasn’t yet been promulgated, it’s not even in a Notice of Proposed Rulemaking but is in line to be promulgated, is an attachment standard which is generally consists, generally it’s the attachments that providers have to send to payers to support services, but not for every single patient. If you need something on every single patient you’re going to want to put it on the claim. But if you only need it for certain patients, like those who came in an ambulance, or those who had certain tests maybe, or those who, whatever, and it’s usually clinical data because it’s not supported by the claim, and these are not, even where claims are almost totally electronic, attachments are almost all in paper, so they’re really burdensome.
DR. STEINWACHS: They could be electronic --
MS. GREENBERG: They could be but there’s no standard for them, so like everybody asks it differently, and of course the providers view is that we don’t need any attachments, and the payers view is we can have as many attachments as anyone would ever want, so somewhere between. So this standard, which has now been delayed for reasons I won’t go into, but is still coming down the pike, they’ve defined like six of them and one of them is laboratory, and one is ambulance, and one is emergency department, I can’t remember what they all are. But there is actually an electronic format that is being developed, working with HL7 which is the clinical group, working with X-12 which is the administrative group, to transmit electronically a lot more of clinical data than is currently supported by the claim. But it wouldn’t be for everybody, it would be for specific --
MS. COLTIN: It’s a two part, a discharge diagnosis is on the standard claim.
DR. STEINWACHS: But not present in admission.
MS. COLTIN: That’s on the standard claim, it’s just not a required field.
MS. GREENBERG: It’s on the standard but it’s not on the HIPAA guide, it’s not in the HIPAA guide.
MS. COLTIN: I thought it was actually, it’s not in the NUBC data set?
MS. GREENBERG: It’s from the diagnosis indicator, no three, which is the diagnosis indicator, that indicates whether a secondary diagnosis was present on admission, not present on admission, or you don’t know. You have those three options.
DR. STEINWACHS: So you’re assuming the principle diagnosis also.
MS. GREENBERG: And that diagnosis indicator is supported by the standard but is not in the guide, is not in the HIPAA guide, it’s in our guide.
MS. KANAAN: So just in terms of thinking about the structure of this report, or maybe one table or something, I can imagine some kind of a matrix or table really, that states what the recommendations, what the recommended data element is, and then has these other variables, the standard, the implementation guide, the whatever else, is it required or is it optional, so that it would be easy for people to track the status or where the --
MS. COLTIN: I think that makes sense in the recommendations section where you kind of pull them all out. I think in the text section what you want to do is put it in its context, so that you’re saying here’s the gap that was identified, here’s the data that’s needed, here’s the data source that it could come from and that can sometimes be more than one because you can make a recommendation that it be collected through a survey and through something else, and then what the recommendation is. And then in the end you’d pull all those recommendations from that and put them all together.
MS. KANAAN: And see where the obstacle is to the --
MS. COLTIN: -- it seems to me, because to just look at them in the recommendations section, people are going to look at it and say why, whereas if they’re built in you’ve got the rationale, here is the need, here are the opportunities, here’s the recommendation.
MR. HUNGATE: I guess I’d like to get the conclusions and then develop the support. Is that ok?
MS. GREENBERG: What about in patient safety, did they talk about better --
MS. COLTIN: They talked about that as being one --
MS. GREENBERG: But adverse events, how about adverse events, because --
DR. STEINWACHS: They would not be captures in diagnostics.
MS. GREENBERG: Either there are external cause of injury codes that indicates that this was a medical misadventure or whatever, they’re grossly underused --
DR. STEINWACHS: For obvious reasons.
MS. GREENBERG: But again, the standard permits much more extensive collection of external, of e-codes than the HIPAA implementation guide does. That’s certainly a patient safety related issue and --
DR. STEINWACHS: These are sort of the three highest priority categories --
MS. GREENBERG: What about functional status? The standard doesn’t support functional status.
MS. COLTIN: Well if I had to say where was there universal support, I think that in the area of functional status it would have more to do with the code sets recommendations that would enable the, what is it now called, ICF to be a standard code set that could be used. That’s the first step.
DR. STEINWACHS: So the capacity to --
MS. COLTIN: To even use that code set, because right now it’s not a standard code sets so --
MS. GREENBERG: It’s not a code set named under HIPAA, and actually what the Committee recommended was that we needed to do demonstration projects and pilot studies, which I am just waiting and ready to do but I have no money to do.
MS. COLTIN: Part of my hesitation is that on functional status, that was the main recommendation in the report, is that we look to that particular code set as sort of the great hope of where this could be realized, because even in the core data elements report, we had put a data element for functional status but we had been very vague about in how you would measure and what code set would you use and all of that. What at least that functional status report came to was that of all the possibilities out there this code set looked like it was the best possibility --
MS. GREENBERG: It’s not a measurement to it really, it’s a classification. But when we did the core data elements report it was still going under revision, and when we did the functional status report it had now been adopted by the World Health Assembly, the ICF, so we could say now let’s, this is something to point towards.
DR. STEINWACHS: Is that used in capturing, in nursing homes was it the MDS or the --
MS. GREENBERG: All of those can be mapped to the ICF, that’s, the FIM(?), the FDS, the Oasis, the work really is to map those and other assessment tools to the ICF.
MS. COLTIN: And I think that would probably be what I’d recommend in terms of making a recommendation, would be to support the recommendation that was in the functional status report and to reiterate again the need to do studies that will --
MS. GREENBERG: To do that research, because a lot of people did mention functional status. And if you’re looking at outcomes, and if you’re looking at health and you’re looking at chronic disease and all of that, it’s really more important than the diagnosis.
MS. COLTIN: And those codes, which are like diagnosis codes in a way, with a diagnosis indicator can also be powerful in terms of something that wasn’t present and later it.
MS. GREENBERG: That’s true, you could do that with functional codes, too, that they came in being able to walk and they left and couldn’t.
MS. COLTIN: Even if it’s temporary, like with a hip replacement or whatever.
MS. GREENBERG: And that’s where you need data, of course, over time.
MS. COLTIN: Over time. You need to have it in the professional claim so they can --
MS. GREENBERG: Actually, what the ICF codes do that the ICD codes don’t is they include severity, so they say whether you have any difficulty, moderate difficulty, severe difficulty, so they really do allow you to track over time improvements or deterioration.
DR. STEINWACHS: So Kathy, essentially what you’re saying is the report, the recommendations would essentially be around these four problems based on what the testimony was --
MS. COLTIN: And I think one more that would be useful is the linkage of the procedure code to who did it, the rendering provider in the field --
DR. STEINWACHS: So we have that for surgery but we don’t have it --
MS. COLTIN: On a professional services claim you get it, because you have a service line for every code so you know you have a rendering provider field on that service line, but on a facility claim right now for a procedure that’s done you have the procedure date and the procedure code but not who performed the procedure. There is a field, or there was a field I think for the operating surgeon --
MS. GREENBERG: There is a field for the --
MS. COLTIN: Operating surgeon --
MS. GREENBERG: Which should go with the major procedure, but there may be other services --
MS. COLTIN: That’s the problem, they may have had a cardiac cath done by a cardiologist and then they had a coronary bypass done by a thoracic or cardiac surgeon, and you don’t know and if something happened as a result of the cath you’re going to blame the surgeon when it should be the guy, of course we don’t blame --
DR. STEINWACHS: We don’t blame, in this society, Kathy, no blame.
MS. GREENBERG: You could link the professional claim with the institutional claim, the problem is they use completely different classification systems, so that’s a problem also which gets to the single procedure classification system which I’ve now decided I won’t see in my lifetime. But what about the whole issue of linkage, though, unique identifier? I mean that kept coming up, too. I mean I know we’re not supposed to talk about that --
MS. COLTIN: The issue, the data linkage is another big problem that came up over and over again, is the ability to link data, and that does get you back into that really touchy issue of the unique identifier for the patient, which we I thought were told to kind of keep hands off of --
MS. GREENBERG: Cease and desist.
DR. STEINWACHS: Well, why don’t you just give me unique identifiers exist for facilities, at least hospitals, for nursing homes, in terms of national --
MS. COLTIN: For individual physicians, the --
MS. GREENBERG: The national provider identifier will be for all providers, individual and facility, which is about to come out in final form.
DR. STEINWACHS: So if you’re a psychologist you can bill as a psychiatric social worker you would have a unique national, if you can’t bill but you render --
MS. GREENBERG: Even then I think ultimately probably they would enumerate.
MS. COLTIN: They were talking about giving codes to taxi drivers at one point.
MS. GREENBERG: Because Medicaid pays for taxi.
DR. STEINWACHS: And I think they should get those taxi drivers, quality performance is important in all dimensions --
MS. COLTIN: It really was service provider in a fairly broad sense.
MS. GREENBERG: Certainly nurses and others --
DR. STEINWACHS: Essentially you’re saying when you got licensed in the future, you would get a rendering code, a national rendering code --
MS. COLTIN: A national provider identifier, which would then be put in that rendering provider field when you are the provider who performed the actual service, and that you can do that on a professional services claim right now --
DR. STEINWACHS: But you can’t do it on a facility claim.
MS. COLTIN: But you can’t do it on a facility claim. And when the person is hospitalized they often, that’s when a lot of procedures are done and often they’re done by different individuals, different providers, and right now you cannot disentangle that at all.
MS. GREENBERG: Even if you can link the professional with the, is it because of the different classification systems?
MS. COLTIN: That’s what you’d have to do and that’s what I actually suggested to them, is that they think about doing that through linkage because there is a field on the professional services claim for admission date and discharge date for when you’re billing professional services that are associated with an admission, you fill in the fields for the admission and discharge date, and so using that they could in fact link those, but then that gets them into are you going to get the same patient identifier, is the doctor going to use the health plan ID and the hospitals going to use their own ID for the patient, and how do you link those two if in fact you’ve got different patient identifiers.
DR. STEINWACHS: And I think the other problem is, depending on how you interact with a patient, you may not really know the discharge date, so you maybe when you see the patient no one can figure out what the admission date is. The matching of those is --
MS. GREENBERG: These linkage issues which raise all sorts of flags, but we used to say well we can’t even talk about them until we have national privacy regulations, well we have them now, but even so it’s --
MS. COLTIN: I mean within an insurer, within a payer, you can obviously link them because they’re putting their subscriber number or patient, for billing purposes you have to know this is your member and something you should pay for. So in a health plan I can link these and even if they don’t know the discharge date all I have to do is look at the dates of service and see if they fall within the range of a discharge date and admission date that I have on a hospital claim. So once I’ve got a patient and I’ve got those two dates, they then go to the other database and I say find me every claim that came in with a service date of that time period. So we can actually do that, but it’s a problem in a lot of cases because the provider bill, the professional services bill comes in with a provider code that’s a group as opposed to an individual.
MS. GREENBERG: Now that’s going to end with the NPI --
MS. COLTIN: Well, hopefully.
MS. GREENBERG: I think.
MS. COLTIN: It will, because groups are going to get NPI’s, too, I thought.
MS. GREENBERG: No.
MS. CARR: May I get clarification? Are these recommendations being framed around a particular data set or data system? Or a particular users needs? Or a particular purpose? I can’t quite get the framework.
MS. COLTIN: I think it’s all of the above, I think what happens is that the particular purpose is measuring quality of care, that’s the broad. Then you get into subsets, things like what do you need to look for to do risk adjustment, and one would say these are the data elements that people said were really important for that purpose. And where might those data elements come from? Well, they might come from an administrative data system. And right now what we’re talking about are really just administrative data recommendations that pertain to administrative data systems, but we can move on to others as well. So the diagnosis indicator, for example, is if you’re trying to measure outcomes, or you’re trying to, and look at complications that might have occurred during the course of a stay in the hospital, that diagnosis indicator becomes very important because you want to know is that condition present when they arrived or did it emerge --
MS. CARR: I understand the specific indicators, but I’m not getting a sense of framing for this that will make sense to an outside audience.
MS. COLTIN: I think that’s partly because we really need to kind of lay out what are the sub categories under measuring quality of care, what are the kinds of things that people are saying they were trying to do. So we’re trying to look at follow up in care coordination, and what are the issues in trying to do that. They were looking at being able to risk adjust to compare outcomes, one of the biggest problems when you’re saying we’re trying to measure outcomes of care is the ability to risk adjust so that you can compare populations, so what are the data issues and gaps around being able to do appropriate risk adjustment. Another thing that we heard about was, on the outcome side was the use of the diagnosis indicator in two ways. One as a risk adjuster when it is present on admission, one as a potential outcome when it wasn’t, so it falls under two categories. I mentioned the continuity and coordination, the lab data are another example of one that would fall under more than one category that can be, they’re very important for looking at continuity and care coordination, and things like follow up of abnormal findings and whether it happened or not. But they’re also important as an outcome and a risk adjuster because if you were looking at classifying a diabetic population in terms of severity of illness, lab results can help you do that. Likewise if you were looking at outcomes, you might look at control of their hemoglobin A1C as an outcome and then you need the lab result to be able to do that. So some of these same recommendations will come up as supporting more than one type of quality measurement. But I think it makes sense to start out with some typology of what most people are trying to do around quality measurement and then what the gaps are. And we’re going to see these same things showing up again and again and where one recommendation could actually address multiple needs.
DR. HOLMES: Isn’t the big bucket here, though, data gaps, I mean isn’t that the big bucket that we’re talking about? We don’t have data that supports this type of measurement at this time.
MS. CARR: I understand that, I’m thinking more in terms of, when you have a recommendation it seems that you need a target for that recommendation either in terms of who can act on it or something, and that wasn’t coming through to me.
MS. COLTIN: Oh, I’m sorry, that part is these are all recommendations for data elements that should become part of this HIPAA standard transaction that we were talking about before. The implementation guides for the 837, we would be saying, for example, that the implementation guide for the 834 enrollment and the 837 claims, institutional and professional, ought to have a field for race and ethnicity, or fields, plural, for race and ethnicity. We may want to also recommend education as another area for looking at disparities, and it’s also a very important area in terms of targeting quality improvement intervention. I mean knowing what someone’s level of education is helps you in designing educational interventions.
MS. GREENBERG: And language, too.
MS. COLTIN: And language as well.
MR. HUNGATE: I have a real need to do this, because I do get totally lost in this and I’ll confess that. And I don’t know how to do that but I think we’re going to have to break for lunch. I hope that you can think about what organizing, what I want to come to is kind of six recommendations, just as an arbitrary, if you don’t have a reasonably tight list it’s hard to get it articulated. Now one might have --
MS. COLTIN: So can one recommendation be additional data elements are needed on the 837, and then you could list four or five, they wouldn’t be separate recommendations.
MR. HUNGATE: Yes, but I think in light of what you’re saying, the why has to be attached to it, the why and the what, and we have to have in mind who has to do it. And this is not purposefully unclear but --
DR. YOUNG: You also want things that can be operationalized kind in the world that we live in now?
MR. HUNGATE: I think we want things, my sense of the recommendations is that they should be operationally doable now, and move us in the direction of where we want to be.
MS. COLTIN: And we might want to divide them into short term and long term, so for example, in terms of even a recommendation about adding data elements to the HIPAA standard transactions set, short term, if the variables already exist but they’re just not part of the content data sets, that could be addressed in the short term. If they’re not even part of the standard and we think they should be added, that might go under a long term.
MS. GREENBERG: And it’s already in the standard, I can tell you that the only group that really came forward to make the case for race ethnicity was the Consortium, and the real use case that we were able to present the need, the business case, was the state hospital discharge data systems. We have now met their need through our guide so they’re saying it’s not going to be in the next version of the implementation guide, but it’s timely to --
MS. COLTIN: We’re going to start finding more demand for this.
MS. GREENBERG: And we need others to come forward.
MS. COLTIN: The reference that I have on the Aetna thing, it was in the Wall Street Journal but I actually have it from the Boston Globe, it was in an Associated Press stories, so it was picked up by a lot of different papers on March 5th, and what it says is Aetna has begun collecting data on the racial and ethnic background of some of its 14 million health plan members, in what insurers call an effort to narrow the gaps in treatment between whites and minority patients. The Wall Street Journal reported Thursday, so I don’t know March 6th might have been Friday and that might have been in the March 5th issue of the Journal. But if you have large payers with a lot of influence on both X-12-N and some of the content committees who want this, Aetna being one of them, it’s going to drive it because it’s the payers that are really I think most influential in what gets on these. But payers is a very broad term because the point I was trying to make before is the people who represent those payers on some of these committees tend to represent only one functional area within that payer organization. And unless the leadership within other areas of that organization connects and communicates with that representative about what the other needs of that organization are, they’re in there representing one very narrow perspective which is pay claims, process enrollments, and so a big part of this is making sure that we can try to foster internal communication.
MS. GREENBERG: At the end of the day these committees are more controlled by the providers, though, than the payers. But it would help to have the payers, the providers really are against this but nonetheless.
DR. EDINGER: Would another place be the personal health record because they’re race and ethnicity data could also be captured when you come in because some of the treatments and some of the diseases are based on your race and ethnicity, not social and economic data necessarily, but certainly race and ethnicity data would be important.
MS. GREENBERG: I think cultural competence and all that, we’ve tried to make the case but frankly --
MS. COLTIN: I think we needed to take this data source by data source, so we were starting with the administrative data set. I think we have recommendations around surveys that we ought to make and I think around electronic medical records, and I think the personal health record kind of fits under that third bucket, so I absolutely agree with you but I was kind of approaching it source by source. If you want to approach it gap by gap, I think that’s something you need to decide, how to organize this, you could take the gap and say ok what could you do with administrative, what could you do with --
MR. HUNGATE: Well, this is short term, long term, on each of them, so I have to spend some time Bill Yasnoff and talk about the NHII and how we judge whether we’re making progress or not over lunch, so I can’t, I’ve got to do that. I opened my mouth once too many times, so it’s lunch break. Reconvene at 1:30.
[Whereupon, at 12:44 p.m., the meeting was recessed, to reconvene at 1:44 p.m., the same afternoon, April 11, 2003.]
Agenda Item: Continue Review and Formulation of Findings - Subcommittee and Staff
MS. COLTIN: As I listened to your suggestion earlier, Bob, about the purpose, like the surveillance, I thought maybe that’s a better way to do it, and then you --
MR. HUNGATE: I feel more comfortable if we could do that.
MS. COLTIN: So what I was talking with Susan about that I thought might work, and I think we should debate this, is to take that notion of purpose and to say ok, the main organizing principle is that we’re trying to measure an improved quality with care, so we’re not looking at every purpose in the work, it’s a narrowly defined purpose, data needed to measure and improve the quality of health care. And under that, there are particular subsets like surveillance, quality assessment, quality improvement, and they have different needs, different data needs. And under those, there are also different categories of activities or types of measurement that are going on, so if you’re talking about quality assessment or quality measurement, you can put it that way, there are measures of the process of care, are things that should be happening happening and things that shouldn’t be happening not happening. There’s measures of the outcomes of care, there’s measures of the structure of care that’s being delivered. There’s measures of the continuity and coordinate of care. So you start looking at what are the major ways in which the activities of quality measurement tend to get organized, they tend to get organized within that kind of a framework. And one can then take those and say well of those people we heard from who are trying to measure process of care, what were the major issues and problems they were encountering in doing that, missing data, that’s where I think these headings then come up, that you could say well, because some of these lack of specific data apply to quality improvement, some of them apply to quality measurement, some of them, see where I’m going, so that was sort of one way or organizing this is to say ok, this groups intention was to look at what needs to be done to improve our capability to measure and improve the quality of care from a data standpoint, and what does that mean to measure quality of care and improve it, what are the data needs to do that, where are the gaps, how might we address those gaps. And that’s where I think this is a great way to categorize the gaps, the way this report was reorganized, missing data, lack of standardization, whatever. I’m not sure its necessarily the best way to organize the recommendations because lots of times the same recommendation can cut across multiple categories. But I think you’ll come up with recommendations within each one of these that will show up in more than one place. No matter which way I thought about cutting it, some of the same things were going to show up within multiple categories, because they’re cross cutting problems.
DR. YOUNG: Can I ask kind of a clarifying question? When you say data needed to measure and improve, I think of measure and improve as two separate events. Now I think they should be linked, I think there should be a, I don’t know, it’s interesting to measure things, sometimes, but I think that there should be a way or there should be in some way a bright line drawn to those people who are the provider, and I’d define that broadly, to be able to operationalize on that measure. Are you thinking in the same way?
MS. COLTIN: I think I am because part of what came out from this morning that broadened my thinking about that is that when you move from quality assessment to quality improvement, your scope enlarges, because some of the, when Don added the fourth bucket about the environment and capacity and things like that, those are all very germane to quality improvement. The example I gave about the colorectal cancer screening, that if the problem is the supply of gastroenterologists or people who are trained, allied health professionals to do sigmoidoscopies or whatever, you need to have that kind of data to support quality improvement. You don’t necessarily need that to support quality measurement. You know what I mean?
DR. YOUNG: I do know what you mean and that helps me think about that, thank you.
DR. HOLMES: Is quality reporting even a third tier of that?
MS. COLTIN: For accountability decision making?
DR. YOUNG: As a method for improvement.
DR. HOLMES: As part of the three tiers, one is quality assessment, then the reporting for accountability, and then third is quality improvement, for the purpose of quality improvement.
MR. HUNGATE: What’s the distinction between quality assessment and quality reporting?
DR. HOLMES: Well, quality assessment I tend to think of that as occurring more at the individual patient level, holding a particular provider accountable for the delivery of health care, as well as the environmental factors and so on impacting on a particular patient, whereas the reporting, the public reporting is more for accountability of a larger population.
MS. COLTIN: Actually I don’t, a lot of the quality measurement, in fact most of it, is at a population level.
DR. HOLMES: Well, we talked last time about the example that we ran into with the National Health Care Quality Report, and we ran into a difference of opinion about what should be in the report as to the level that constituted an adequate hemoglobin A1C level for individual patients. And if you were talking about a particular individual physician dealing with his or her patient, it should be for example no greater than 7.0. But that if you were talking about reporting about a population, you would be unfairly identifying quality problems if you went with that 7.0, then it should be maybe 7.5 or 9.0, because at the individual level the particular physician could take into consideration for example risk adjustment, comobidities, whereas reporting at the population level, which we are for the Quality Report, we were not taking those things into consideration, and there was very strong disagreement.
MS. COLTIN: That’s the argument that was made in the Performance Measurement Coordinating Council which was the group that brought the AMA and then CQA and JACO together to talk about this particular issues, and the way they divided that was actually the way we’re just talking about it. When you’re measuring for quality improvement, you’re reporting to the individual provider and that’s where you’d draw it at 7.0, because that’s the quality improvement goal. But when you’re reporting for accountability, that means that, comparisons are implicit in accountability, and that’s where risk adjustment becomes so important, and where reporting at this higher level was a solution to the inadequacy of the data systems to support risk adjustment. It wasn’t that that was the right level, I think we would have loved to report at the level of 7.0 but only if we could accurately risk adjust, so it points back to the problem of quality assessment for accountability purposes, that risk adjustment becomes a big factor.
The other thing that becomes a big factor when you’re reporting for accountability purposes is sample size, and that’s what drives you many times to administrative data because you can’t possibly measure at the individual physician level across thousands of physicians using medical record chart review, at any reasonable expense in today’s environment. So you end up having to rely on administrative date if you’re going to take measurement down to the level where the action and the improvement really often can occur, and that forces you to rely on administrative data so then you’re stuck with the gaps and the problems with administrative data.
MR. HUNGATE: A recommendation I would like us to make is a very general one which says that recognize the limitation of administrative data, but fix the things, limit what you expect from it but fix the things that make it not even do that, make it do what it should do but define the limits of what it can do, which says that it can be used for accountability, but you’ve got to make it so it will work for risk adjustment, you’ve got to make it so it has outcomes, which it seems to me reinforces the importance of functional status. So to me that’s a macro recommendation of the sort that I think I can argue strongly for. And then we could put some details underneath it that relate to it. Does that fit?
MS. COLTIN: That fits, but it also I think supports this notion of quality measurement versus quality improvement, organizing it that way because what you expect administrative data to do for one purpose you may not expect it to do for another. And risk adjustment may not be as important from a quality improvement standpoint where you’re just providing data to an individual provider about their patients and you’re not trying to compare them to anyone else.
MS. CARR: In terms of the structure you suggested, where do the different levels of care, do you account for the different levels of care, prevention, primary, old ambulatory, inpatient, long term, trauma, emergency, because --
MS. COLTIN: That’s where I think the gaps often appear in certain areas, but if I were talking about quality assessment and I was saying I’m trying to look at the process of care for any particular condition, that care can be rendered in multiple different settings, and what are the issues in trying to characterize process of care measures, for example, as opposed to outcomes of care measures, and I want to know did the person get this test. I may need to look in multiple medical records to answer that question because they may not have gotten it in this visit with Dr. Jones but they did get it in that visit with Dr. Smith, and I need to be able to link that data for that patient across those different doctors and across different settings, and that’s when you run into issues of incomparability of code sets across different settings as a data problem, you run across issues of interoperability of systems and the ability to link data across systems, and that’s when you run into problems of identifiers, unique identifiers and being able to even identify records that belong to the same individual across systems. So those are three different problems on which you could have recommendations, all geared to looking at the problem of measuring process of care for any health condition. They don’t have to be specific that its diabetes or heart disease or whatever, they’re common problems that cut across.
And so the care setting issue to me comes up in the context of thinking about how you’re measuring care for a patient, for a particular problem. When you start thinking about measuring care for a patient at more of a holistic level, that’s where the functional status stuff really often becomes very valuable. Because we know a lot of patients have multiple conditions, and when you start trying to separate them into this condition, that condition, you can look at care that way but then if you’re trying to characterize everything a person got, then the sum of all their conditions, did they get the right preventive screening, did they get the right care for whatever conditions they have, then the issues become even more complicated because you’ve got more settings involved in many cases and the common denominators, or the things that you were able to look at, often tend to be these kinds of functional measures that cut across. Speak up if you’re disagreeing with me.
DR. YOUNG: No, I do agree, I agree strongly with that statement. It’s at the end of the day when I look at different measures, return to normal rates on hypertension or hemoglobin A1C or kind of these other things that we use commonly, I kind of go back and see patients and think I don’t ever think about that. I mean that has never, that is not the model of care when I look at somebody, and so this gets into this, measurement versus improvement. It seems like a useful goal would be at the end of the day is data that’s either driven or derived in such a way that it gives me that functional picture, either --
MR. HUNGATE: You’ve just articulated the goal of the NHII. It says that what you do over here gives you the information that helps you over here.
DR. YOUNG: I would further argue, this is in the QIO’s and OCSQ and others that we have these dialogues, I mean this is what we’re striving for at the end of the day, so to the degree that the recommendations from this Committee can move towards that, can either discount data sets, which you may have to do and say well thank you we’ll be doing that for a while, that doesn’t work, we’re not going to do that anymore, or collapse categorizations of data sets. Just thinking purely as practitioners and people who run health care systems, it becomes useful and it becomes something that I can use in an operational sense, and I guess that’s my bias that I bring in --
MS. COLTIN: I mean I don’t disagree with you but I think the argument goes backwards the other way that says if I start there and I measure functional status, and this comes up for health plans in the context of the Medicare Health Outcomes Survey, where you may see you have a problem that people are doing worse than expected or whatever, in order to sort of know where to plug in, where’s the problem, why are they doing worse than expected, you end up having to tease it apart and to saying is it the people who have hypertension who are doing worse, is it the people who have this who are doing worse. In order to know where to fix the process because usually it is the process that is broken, you need to be able to measure at all those different levels. So I think while the Holy Grail may be let’s measure outcomes, the improvement happens at the process level.
DR. YOUNG: We tease it out, you’re right, we kind of validate this kind of global image then we tease out these other sub systems, but you’re kind of going to go all the way around the world here, you’re going to kind of come back to that global view and that’s kind of the part that we haven’t quite gotten to yet, or we’re good at dissecting the cadaver, but you actually have to be able to put it back together.
MS. COLTIN: But maybe that is the way to start, I mean maybe the way to do this is to do this sort of in the context of the story, maybe what we say is that ultimately what you want to be able to do is improve health outcomes, that’s the purpose of the whole thing. So if you start looking at measuring outcomes to see whether you’re doing that, what are the problems with measuring outcomes and our ability to do that, what are the data problems, next thing is ok, you’ve solved all that and you can measure outcomes now. And you find you have a problem, how do you begin to look at dissecting it and doing what Don Berwick and others called the diagnostic journey, root cause analysis, why do I have a problem, for whom do I have a problem, so segmenting your population to see if it’s certain sub populations where that’s a problem and not others, or sub setting your care settings and seeing whether the problems are occurring in certain settings. Sub setting conditions that the patient has and seeing whether they’re getting the right care for those conditions or whether the outcomes, you could take outcomes down to the condition level, too, and say were there clinical outcomes for the conditions that they do have.
But when you find problems there they usually then go back to process, and so you have to then say alright, so now we’ve got to go and measure process, and what are the data gaps and the issues around measuring process. So you could start organizing it that way and start out with the outcomes and then build down toward the other, and identify the data gaps at each level and how you would do it. I mean one of the things that comes up in terms of measuring global outcomes is that if you’re going to get into not only the issues of code sets and whether we have the right information on administrative data systems, but you’re going to get into surveys and our capability to measure patient reported outcomes, and our ability to do that and what do we know about that at a national level, we have the Medicare Health Outcome Survey and that tells us quite a bit about the Medicare population that’s enrolled in managed care and now also in recent iteration and fee for service comparisons but we don’t know that about most of the other people in the country besides who are under 65. So there comes a limitation around surveys and to what extent can we get the same kind of information from the, is it the disability supplement in the National Health Interview Survey, are they comparable, are we measuring the same way, can you look across the populations, and so then that may get you into the standardization issue of yes, we’re measuring it in both places but we’re measuring it differently in both places and so we can say one thing about the Medicare population and we can say one thing about the population in general, but we don’t have consistency sometimes in how we’re measuring.
So I mean that’s another way to think about organizing it but it still follows that framework of the purpose being to measure the quality of care and just starting with the outcomes and then working backwards.
MR. HUNGATE: When you start talking about outcomes you can’t really say you’re trying to measure health, and then you get rapidly into health care, what you did in the process and structure. So that’s consistent.
MS. COLTIN: But if the group felt more comfortable starting at that level with outcomes, because outcomes is sort of a Holy Grail --
DR. HOLMES: But isn’t that the case, Kathy, that not a lot of the testimony that the group heard had to do with outcomes simply because at this juncture there’s such a possity of good measurement tools as well as measures themselves?
MS. COLTIN: Well I think we did hear about it but we heard about it in the context of some of the other work. We heard about it in terms of the functional status report and the need to measure outcomes, and I think we did hear it from some of the other, in the November whatever testimony where I brought a group of people, quality measurement experts together to testify, people were talking about some of the limitations in trying to measure to outcomes.
MR. HUNGATE: It’s in there, it does express the limitations, and does say they’re isn’t much of it.
MS. COLTIN: That’s why there’s such a data gap, such a gap like in the National Health Care Quality Report in the outcomes area is precisely because of a lot of these issues of risk adjustment as well as the issues of being able to parse out how much of that is due to health care and how much of it is due to these other, this other bucket over here around the environment and genetics and all of that other stuff.
MR. HUNGATE: I think we have to articulate to some way the boundaries the administration data and the clinical data and the cross hatches in order to have any sort of recommendation that holds bait. Does that feel comfortable, does it fit with your mindset, that way in which you see it?
MS. COLTIN: Laying out what we think those boundaries are?
MR. HUNGATE: I think to some degree, allowing for cross hatch but you know I look at all the patient safety efforts and they’re asking people to report things that they’re not going to want to report in some cases, and so if you’re going to do that you’ve got to arrange a system that provides confidentiality, you’ve got to have process confidentiality and so you’ve got to have a transition from the clinical confidential to the publicly accountable quality assessment. And I think we have to articulate something and say that the NHII is going to have a big payoff in the clinical part, which will show up in the way we assess things in the administrative data, but we’ve got to have the right data there in order for that to work.
DR. YOUNG: Can I ask kind of a clarifying question? We’re kind of working on designing some demonstrations around information systems, that we’re thinking through these issues. And we think about administrative data, fiscal measurements, and we think about clinical data, but we also think about another bucket because that’s not the whole story. We think about clinical data is like blood pressure, things like that. We think about systems data as well, and should we not be thinking about that? You mentioned clinical data and administrative data, it would be breathtaking to think that we were cutting new trails, or we were delusional once again, that’s usually the case.
MS. COLTIN: This is an area that is a developing area of measurement right now, there’s a lot of instruments that have been in play just in the last year or two, the Larry Castelino(?) article in JAMA about a big national survey Steve Shortell(?) and others did, they measure a lot of these system aspects around their relationship to quality. And there’s been a number of studies done that are showing direct relationships between having some of these kinds of systems and outcomes and compliance with guidelines and so forth. There was a big analysis done by the Cochran Collaborative on that recently. So they are looking at things like do doctors offices have patient registries, can they link data, those kinds of capabilities. To me I would put them in that structure category that I mentioned. I think those people do fall into that framework and I think it’s a reliable framework to --
DR. YOUNG: Here’s something that we think was so important that we put it as the third classification of developing measures around this, that we felt it was just that big of deal and we started talking about fee for service demos or other demos that it’s going to be in there.
DR. JANES: I mean certainly it speaks to in chronic disease, one of the things people have been looking at for a couple of years now is some of the work that Ed Wagner has done around the chronic care model which has got a lot of different names now, but it’s referred to as a systems based approach to care and I think it comes from this idea that particularly in the treatment of chronic disease, with the requirement for long term follow up that you’d have to address changes at the system level, and then it does, as you said, it brings us again just sort of hewing to at least Ed Wagner’s model, some of the things he recommends I think you can easily track, as you said, does a physicians office have some sort of a medical record, is there information systems, and some of it has to do with much more a morph’s concept like the development of a team based approach and redefining responsibilities and encouraging responsibility sharing across physicians and paraprofessionals and --
MS. COLTIN: Linking into community based resources.
DR. JANES: You got it. These are tough concepts to measure.
DR. YOUNG: To the degree that we can see kind of a cross walk between those three larger, clinical, administrative, and systems, I guess the view we have is the way that’s coming at us and I would say the way that it was coming at much of the health care center is competencies, I mean that’s going to be certainly the driver that we see in Medicare, and again, I would argue that many other systems, private systems as well, and it becomes useful at that point. So if you were talking about that cross walk back and forth I thought of a third area --
MS. COLTIN: The one difficulty with that and I think it belongs in the report but I think it may need to be stated as a limitation of the report, in that we didn’t take a lot of testimony on that, I don’t know that it really came up, I mean we didn’t take testimony focused on it, and it may have come up here and there in terms of things people said but I don’t recall it being a major theme.
MS. GREENBERG: But if that includes like information systems or --
DR. YOUNG: Information systems, patient waiting time, there’s your FTE count change, I mean doing these things, either up or down, it’s one of three broad categories of the measures that we will use as we go forward.
MR. HUNGATE: And I think of those as all a part of quality assessment and accountability. They’re external views of the quality of care delivered and the capability to deliver that care within that system.
MS. COLTIN: And they’re extremely pertinent, maybe even more so than to assessment, to improve. Many of those kinds of structural features are the interventions that were employed to try to improve, well let’s put in a patient registry system or let’s put in a care coordinator or --
MR. HUNGATE: These pieces are used over here, and the other thing that I think will evolve over here is share data across registries in a different way, around peanut allergy and every single comobidity that you have could have a separate, I think the research field will move toward databases research drawn from collected data across the system. But then I see information systems feeding back to professionals, problem oriented record, the various kinds of systems that can improve and augment the capability. But you have to have these pieces meshed with it.
MS. COLTIN: The one difference I’d see with this, I mean I see surveillance and quality assessment and QI as kind of the top level. I would see administrative data, survey, and clinical as actually branching out from each of these, well maybe not so much structure but certainly from the outcomes and process is how is each of these, when you’re trying to measure outcomes, where the value group surveys, how do they contribute and what are the limitations, what are the value of administrative data, how does it contribute, what are its limitations, what are the value of medical records, how do they contribute, what are their limitations, what did we hear. Medical records often are not a very good source of outcomes data. If the patient doesn’t come back you figure out they’re fixed --
MR. HUNGATE: Or dead.
MS. COLTIN: That’s right, or dead. So that’s a big problem with medical records.
MS. GREENBERG: I completely agree with you, I think the type of data is a lower level type of thing. But I need to understand a little bit better about what the distinction you’re trying to make about, I mean obviously the administrative data they have their limitations and for a whole lot of reasons that have been documented, but what is the distinction you’re trying to make about that you can only, the boundaries that you’re trying to draw.
MR. HUNGATE: It has to deal with the orders of complexity of what the system can deal with. To do a good job in the clinical setting for each and every patient, the depth of information is going to exceed the capacity of quality assessment.
MS. GREENBERG: Yes, probably.
MR. HUNGATE: Because you know the clinical trials all have a bell shaped curve that shows the results and it gives you a percentage and so many patients get so much benefit, well there’s people at both ends of the curve that are different. And if they get the best result would be different in clinical environments, so just the complexity. The system will work better under chaos law than it will under cookbook procedure.
MS. GREENBERG: Well, individual care clearly needs --
MR. HUNGATE: Right. And our task is to mesh the system that makes individual care work better with the overall assessments of system performance. And I think that’s the architecture thing that we need to have a model for.
MS. COLTIN: I think where you’re going is how far do you go with administrative data before it becomes a medical record.
MS. GREENBERG: Before it becomes a medical record, exactly.
MS. COLTIN: And you don’t want to go that far.
MR. HUNGATE: And you don’t want to go that far.
MS. GREENBERG: Which is the conversation we certainly had with this guy at the NUBC.
DR. YOUNG: You want to start with the medical record, you emanate it from the administrative side which is what we’ve done now, we’ve kind of wrapped everything around it versus kind of the BA model, where they kind of start with the clinical record and wrap things around that, and I would argue the latter is a little more useful approach at the end of the day.
MR. HUNGATE: I would agree with you. And I think we want to recognize that and say that we don’t want to burden the development over there with the structure we have to maintain administrative consistencies.
DR. YOUNG: We should say that explicitly.
MR. HUNGATE: We should say explicitly say --
DR. YOUNG: I agree.
MR. HUNGATE: And that’s what I’m trying to get at, and then bound our discussion in the administrative data.
MS. COLTIN: And lot of it gets caught up in the what versus the why. With administrative data you can usually do a reasonably good job of figuring out what is happening and if there’s a problem. But why it’s happening requires going down to a much finer level of detail on data then you can usually get from administrative date, and I think what we’re trying to struggle with is how far down in answering the why should you be able to go with administrative data, and when do you say no, I need to go to medical record data and I need to address --
MR. HUNGATE: -- risk adjusted outcome --
DR. YOUNG: And I would also bring up something else --
MR. HUNGATE: As a test say let’s hypothesize, that’s one level. Are there other levels?
DR. YOUNG: Coming from the clinical perspective as opposed to the administrative perspective, I’m much more agile at the kinds of information I can get and as it changes downstream. One of the lessons CMS has learned is that we’re very powerful at measuring things administratively, we’re not terribly agile at measuring things administratively, and that’s become a problem. I mean as we’re now doing our public reporting, that’s now fairly plain. And it’s a large discussion and it’s powerful but not agile, and I would argue agile is better.
MR. HUNGATE: The current issue of Health Affairs argues for the return to public release outcome mortality data as a better approach I think then the current.
MS. COLTIN: The problem with only going that far down, I mean I would agree with you, you have to go, you should be able to do that. But then the question is who has the responsibility to fix it, where does the accountability lie when the outcome is not what you want it to be. So that unless you can down a level to look at not just outcomes but some of the broad process measures across the system, you’re not going to get people to take ownership of the poor outcome and --
MR. HUNGATE: I don’t disagree with you, but I’m kind of making the argument we have to put intensity behind the outcomes, the importance of outcomes, because one of the reasons, the clinical information systems hasn’t evolved is there’s no demand for it.
DR. HOLMES: I thought that was one of the things we were going to suggest in the forthcoming phase, that we expand and broaden this whole concept of quality measurement, that at this point --
MS. COLTIN: I agree, I think --
DR. HOLMES: -- that that was going to be a recommendation that the Committee work on that in the next phase.
MS. COLTIN: And that it recognize that outcomes are influenced by a lot more than just health care, and that even when they are influenced by health care, you need to go down a level to get the health care delivery system to take ownership of their part of the problem, so just going to outcomes broadly I don’t think is going to help because most patients are getting their care from multiple providers in different settings and it’s very easy to do this. And so I think you have to be able to go down to the level of profits.
MR. HUNGATE: But I don’t think you necessarily have to do it with administrative data, that’s the point I’m trying to make.
MS. COLTIN: How do you look at care across settings without administrative data, in today’s world and in the near term?
MR. HUNGATE: Probably can’t.
MS. COLTIN: So that’s why I think you need to use administrative data.
MR. HUNGATE: At least in the near term.
MS. COLTIN: And maybe longer than that, because I don’t know, even in a world with an NHII as I said before, a given provider may have access to data all over the place about the care this patient’s gotten from lots of different places, but unless someone has the responsibility for all that care, and can look at it and piece it together, who’s going to have access to that record, now I’m thinking with a privacy hat on, besides the people who are involved in delivering the care who can look at it and say there’s a problem here, this patient isn’t getting, they’re seeing six different doctors but something’s not happening that should. I don’t think any one of those six is necessarily going to take that on themselves to look at the care everybody else is delivering and not just their own. I think maybe under primary care model that might be more likely to happen, that the primary care provider would be accountable for that and for looking at the care that others are delivering, but as you move more and more toward a PPO where people don’t have to designate a primary care provider and can just go and see specialists for everything they want, I’m worried that model falls apart and then where does this happen.
MS. CARR: I wanted to step back a minute to the overall framing where you suggested surveillance, assessment, improvement. I want to suggest that from what I’ve heard and how others frame this, it’s all assessment and measurement is what we’re about, so that’s a this far end, then it’s form either for monitoring/surveillance or accountability, or quality improvement, and the data types that one would need to do that assessment would be structure, process, or outcome data, and the sources would be vital statistics, survey data, clinical data, administrative data or whatever. But I have a little problem with seeing surveillance, assessment, and QI because I think assessment is really the --
MS. COLTIN: I think you’re probably right, it’s interesting because I was sitting here and I had surveillance assessment and improvement and then underneath assessment I had external accountability and internal feedback and I think what you’re saying is maybe assessment is the overall branch and then you’ve got surveillance, you’ve got measurement for external accountability, measurement for internal feedback, and measurement to support improvement as well, which could be things like alerts, reminders, outreach, whatever, which are different than feedback which is sort of this is you’re doing on these four things.
MS. CARR: So it’s all assessment, but then assessment for different purposes.
MS. COLTIN: For different purposes.
MS. GREENBERG: But what about action? I mean for quality improvement you actually want to take action. You start with assessment, and then you evaluate whether the actions did anything.
MS. COLTIN: But even the improvements, when you start thinking about the data for quality improvement and you start thinking about things like whether it’s computerized reminder systems or whatever, they’re internally making the assessment about whether they need to send a reminder, so I don’t really have a problem with putting assessment as the big branch, as the over --
MS. CARR: Knowing that, I mean the purpose assessment ultimately is to improve health outcomes, ultimately.
DR. JANES: I don’t think of surveillance as anything other than a series of assessments over time.
MS. COLTIN: And I think what often distinguishes them is kind of the level at which they’re at. Surveillance is a really hard sort of population based broad assessment is usually often within a system of some sort, whether it’s a health plan or a provider organization, and improvement is often at the patient level. You’re getting a reminder on Mary Jones because she didn’t get something, or you need to outreach to her. A lot of times in order to do improvement you really have to take the data down to the level of the individual patient. And those aren’t as clean as I made them, they’re gray between each one.
DR. STEINWACHS: Just to add a couple of things. It seems to me part of trying to cast this I guess, you’re raising some of these issues, one is you’ve got me thinking about well how do we know it’s a quality problem, well it’s either because the outcomes are bad or unexpected or the processes are not consistent with scientific evidence that say these processes predict outcomes. And so one way might be to sort of cast that assessment function is on some combination of process outcome, what this Committee’s concerned with is our capacity to do both of those, and then it seems to me that the links between clinical and administrative, I think very clearly one could talk about that we can push administrative data but we probably can’t push it beyond sort of in the short term. I’m not convinced necessarily that in the long term they’ll be a lot of blurring between what we call clinical and administrative, it’s just, I’m sure there’s more than one encounter, you’re viewing one version of this data set, I’m sure they’ll be a different level, so I think the short term distinctions are useful, I’m not sure the long term are because the long term I would think are more successful, these are integrated frameworks and you’re no longer talk about, the administrative is not clinical and clinical is not administrative because indeed they intermix in certain ways.
MR. HUNGATE: I wonder if the issues of confidentiality and privacy do get into that issue.
DR. STEINWACHS: Well, but I think confidentiality and privacy get into what pieces of data pass through what walls.
MS. GREENBERG: That’s really what I wanted, because I think the vision is that the administrative data would just come out of the clinical data --
DR. STEINWACHS: Well we would hope that our current administrative, it didn’t get into the clinical data, but sometimes there are things that a clinician ought to know, a set of the administrative data set that just don’t show up over them.
MS. GREENBERG: -- the system, and then you might, the different subsets, so I think if we think of, maybe not just administrative data for the purpose of paying claims or supporting payment, but sort of reporting data, which could be quite, it could use a lot of clinical stuff if they were readily available. It wouldn’t necessarily have to go to the payer for payment but as Kathy points out, it’s got to go, I mean somebody’s got to be up there, somebody’s got to be looking over this.
DR. YOUNG: There’s a great example that brings all those together, I was thinking about it, Tyranni(?) and Ghandi(?) kind of did this work on how do you reduce utilization of lab tests, and there were three or four papers that all came together that showed that if you did these following things you would see like a 20 percent reduction in over under utilization. If you showed people their own lab results, the doctor explained your own lab results, so there’s clinical data right there, kind of streaming at them. You showed them what a lab test costs, and you can make an argument that this is administrative data and if you had a system in there that showed the probability of the lab being abnormal, that it improved quality so there’s some ambient intelligence that comes in to this. When I think about these things I kind of have to think about them in a real life, what is this really going to look like at the end of the day and that’s one prime example of merging all of those together for quality.
MS. COLTIN: Interestingly, the other thing that we’re starting to hear as a health plan, when you start thinking about the personal health record I think a lot of what people have been talking about is all this clinical data. It’s amazing how many patients want the administrative data to do their taxes in particular, so they want to get their claims data to see what their out of pockets and their deductibles and everything were that they had to pay so that they have a record to file in their taxes. So they want that administrative, too, which is quite interesting.
DR. YOUNG: The EOB is kind of gobbledygook for a patient.
MS. COLTIN: That’s right, just give me once a year, tell me for last year.
MS. GREENBERG: This is a long term vision for the medical expenditure battle study, if everybody had those, abstract them and go home.
MS. COLTIN: There was an article in the Boston Globe just either yesterday or the day before where Forrester(?) Research actually did this looking at health plan web-sites and your ability to get the kinds of data that consumers are saying they want to get from different health plan web-sites, and ours got dinged because we couldn’t give that exact example of a person who said I’m doing my taxes and I need to know what I spent on health care last year, can you give me all my claims so I can see what the out of pocket portion was, the co-pays and all. And we couldn’t do it at that point, we will be able to in October, we’re building it, but we weren’t there yet. So we’re already getting hung out to dry for not having it, that’s how, and it made the Globe.
MS. GREENBERG: Do some have it?
MS. COLTIN: Some do, but it’s new, it’s coming along, we’re actually I think reasonably competitive in terms of where we’ll be, we’ll have it in September, October, but to get held out as being poor on that says something about the state of that desire to have that kind of information, if you’re already pointing and saying they don’t have that capability, somebody’s saying that’s an important capability. And it’s not something that I’ve really thought about as part of this personal health record. Which I also don’t hear consumers clamoring for as much as I hear them clamoring for this.
MR. HUNGATE: Well, are we bogged down or are we making progress?
MS. COLTIN: I think we were ok with taking the last suggestion, moving quality assessment up or measurement, whether we call it assessment or measurement.
MS. CARR: Measurement is the proper word there because at each level some assessment is going on based on the information and the key here is good outcomes, quality contributes to that, you need to manage for quality, you don’t manage what you don’t miss.
MS. COLTIN: So let’s call it measurement.
MR. HUNGATE: So we’ll call it measurement.
MS. GREENBERG: My only concern about measurement is that some distinction between measuring and I guess I’m hung up in sort of a class occasion measurement dichotomy. I mean some tools you need just to sort of classify and categorize and they aren’t the same as measurement. Now they may come out of measurement --
MS. COLTIN: Can you give me an example?
MS. GREENBERG: An example is functional status and ICF. You need to measure functional status in some way, so you might do it as Lisa Azone(?) points out, you could do it in a very informal way, the physician watches to see if the patient, how the patient walks into the room, whether the patient can get up on the table, whether he or she can converse coherently, all those types of things. It’s not administering the Karnosky(?) scale or something, they’re doing something very informal but they are observing functional limitations. Or another way, or you could use a very detailed scale or you could give the person a survey which is another form of measurement. But then you need to capture that information and what ICF does is it classifies it so that you can then report it, just like ICD classifies, it doesn’t measure. And I don’t know, I mean that’s a very important distinction when you’re trying to explain to people what a classification is and isn’t, and administrative use classification. If you’re going to look at a population data maybe you don’t even need to classify it if you’re just doing one on one patient care. But if you’re going to look across populations or look at statistical data or whatever you need to classify or you don’t know how to, you don’t know what buckets to put things into.
MS. COLTIN: But that’s true, isn’t it, of any kind of measurement. I mean if I were designing a survey I’ve got to come up with a scale for capturing the responses and create them --
MS. GREENBERG: This may be a non issue it’s just that there’s so much misunderstanding between the roles of measurement versus the roles of classification, and we’re thinking of the reporting out and administrative data, I just think somehow we have to be, make sure that our recommendations don’t get caught up in that conundrum.
DR. STEINWACHS: Can we go back to the name of the Committee like health statistics?
MS. COLTIN: I think we’re ok because the term quality measurement is so widely used that I don’t think, in fact quality assessment has actually kind of died off, it used to be a big term back in the ‘80’s, it really now is just called quality measurement, so I think that is really --
MS. GREENBERG: Just be aware of these distinctions because sometimes they can be problematic when people are mixing up what they’re trying to accomplish, so I’m ok with that.
MR. HUNGATE: Let me express my concern in this as we go through that there isn’t, I think of the quality improvement as those things that you measure for your self worth, self improvement, the things that you do internally to your functioning, whereas measurement from external is another thing. I don’t want to get us into the position where we’re trying to establish guidelines for internal measurement. We have to establish guidelines for what administrative data is, so that’s where I get the same what is the boundary of what we’re going to specify versus --
MS. COLTIN: I’m not sure we have to set those boundaries because I actually think like patient medical record standards should consider things like time specific entries so you can look at cycle times and delays in care and so forth. I think we need to expect individual providers to have those capabilities so they can do improvement and if the systems can’t support them to have those capabilities they’re not going to be able to do it. So saying what those systems should be able to do to me is important. I mean the example I gave before of having worked under three different electronic medical record systems, which were great at the patient level but didn’t allow you to look across patients, to me is something medical records should have that, medical records systems should have that capability for a provider to look across their panel of patients.
MS. CARR: And I think our proposed framework also has for accountability allow us to deal with that external look and at a different level, and what’s exciting to me about this potential is if we can lay out that framework and say this is kind of the quality measurement needs in general, so far the Committee will have recommendations about A, B, and C, so some of them can be more fully blown up and developed in this report. But at least it’s in a context and a framework that also leads us with some pointers to where we might want to go for the next duration.
DR. YOUNG: If you think about what you just said about EHR’s and PHR’s, there’s kind of these couple of levels that are very timely and the things we’re talking about here, we’re starting to do standard setting for the CHI, interoperability standards and we’ve just announced the first five, and I think we’ll get 18 before the end of the year. And then looking in how do we operationalize EHR’s, there was another set of standards above that that needed to be set, and you’re right, they’re not set now. Stu Gutterman’s(?) going to do a demo, a demonstration with electronic health records, we don’t know what that is. We can’t say that’s electronic health record. That’s what it’s going to be today or tomorrow, and that’s what, an SDO, Standards Development Organization is not as yet blessed that nor has somebody like the IOM, or NCVHS said here’s what we see as these functionalities that an SDO needs to go and lay out, whether that SDO is HL-7 or whoever, something that gets created. And that’s really a stopping point, I mean in a very world way before we can go forward with these things, or before we would be willing to go forward with those things I think. And I think providers, as Kathryn just said, run into that. My system can talk to my system, some of the time.
DR. STEINWACHS: Bob, your concern about not wanting to be too intrusive in the organizations, I’m trying to understand.
MR. HUNGATE: The more standards you have to set the longer it takes to achieve something. Standards are by nature a committee process, they’re slow, they bound some things --
MS. COLTIN: But I think fortunately what we’re going to find is that sometimes a very few areas of improvement will impact a lot of different areas, that we may not have to ask for a lot in order to impact a lot because so many of these problems show up in multiple areas and if you fix it you fix it in a lot of different places.
DR. YOUNG: You’re saying there might be different kind of suites about you ahead of this. I mean first iterations, you’re right, standards are very slow, either adopting them or developing them, but you kind of have to do this, and whether a first iteration you can say we have these kind of core issues and this should be the first iteration, that’s useful. It’s a start.
DR. STEINWACHS: Standards can then contribute to efficiency. At Hopkins we do take pride in the fact that none of our systems talk to each other, so we can not violate this --
MS. COLTIN: How does this work? If the overarching is quality measurement and the three categories under it are surveillance, which is mostly a population, performance assessment, which can be an organizational or an individual activity, and then quality improvement, which often really when I’m thinking about it I’m thinking about it more at the patient level. Under performance assessment you have external accountability and you have internal analysis and feedback. That internal analysis and feedback is part of the quality improvement cycle, but the way I’m thinking about quality improvement is really real time point of care information, concurrent reminders, things like that and what can a system do to support that as opposed to the more internal analysis and feedback, which could happen off an electronic medical record or health record, or it could happen off an administrative data system or it could happen through a survey that’s done about what your patients experience of care was in the setting. Does that work for people as a way to think about that?
DR. HOLMES: Does survey research then go under surveillance would you say? Where are the data sets for surveys, like big national surveys?
MS. COLTIN: I think for the most part they go under surveillance, yes, I would put them there because they really are at a population level, and if they come down to the state level we think they’re really great.
MS. CARR: Well surveys per se could go under, potentially under any of them.
MS. COLTIN: Surveys in general, but you were talking about the national surveys --
MS. GREENBERG: That would make CDC happy if we actually admitted that we do surveillance at NCHS.
PARTICIPANT: But NCHS wouldn’t be happy.
MS. GREENBERG: You can’t make both of them happy.
MS. COLTIN: I labeled that the three big buckets are surveillance, performance assessment and quality improvement, which could be called performance improvement or something else if you wanted to distinguish it. And then I had this further breakdown under performance assessment, which was for external accountability, which really requires some external party to have access to the data and raises different privacy confidentiality issues, then when the data are being used for internal analysis and feedback.
DR. YOUNG: Say what you mean by external accountability just so I can stay kind of definitionally on track here.
MS. COLTIN: CMS measuring health plans or doctor office quality project --
DR. YOUNG: That’s what I thought you meant, I just wanted to make sure.
DR. JANES: Do you have it external to the system?
MS. COLTIN: Yes, to report to consumers, to report to your stakeholders whomever they may be.
DR. YOUNG: But you’re thinking both fiscally and in a regulatory manner as well. You’re thinking of kind of a whole world --
MS. COLTIN: Some states require health plans to report HEDIS data to the state, even though it may be voluntary to report it to NCQA.
MS. CARR: -- your internal analysis and feedback from quality improvement, the distinction you’re making.
MS. COLTIN: I’m not, that was what I was trying to get at before. Quality improvement kind of spans that side. I was thinking of quality improvement more as real time improvement. The data that you need at the point of care to improve that interaction with the patient, so it’s a real time transaction as opposed to, that’s why I was a little hemming whether it should be called performance improvement or some other term because when people think about the term quality improvement it would span internal analysis and feedback and that point of care kind of information, so that internal analysis and feedback could be under either one of those two buckets.
MS. CARR: For me it works better under the third --
MR. HUNGATE: So it’s a little hazy back and forth here, you feel more over here.
MS. COLTIN: I have no objection to moving it over. I mean it stands both, it could work either place, everybody has their own way of thinking about it.
MR. HUNGATE: They are gradations, distance --
MS. COLTIN: See my sense is if I were drawing this I might actually say that surveillance is out here and it cuts a little bit into that external accountability, and that quality improvement is out here at the point of care, but it cuts over into this internal analysis and so if you were going to draw it that clearly there is overlap where you put these --
[Writing on the wall discussion.]
MR. HUNGATE: Ok we’ve got a structure that I think we agree has some overlap, makes good sense. Let’s use it.
MR. CARR: Under each there, potentially one would want outcome process and structure information.
MS. COLTIN: I was thinking of it almost like a matrix, these going across the top then I have outcomes process and structure in that order because we talked about starting with the outcomes, but in order to understand those you need to do this, and in order to sort that you often have to have the right structure. But if we were going to try and tell this as a story and we’re going to say that ultimately we want to improve the health of the population we might want to start here.
So then within each one of these sort of cells that’s created by this, and those become kind of the sub headings in a chapter, they could be used as the chapters and these as the major subheadings, within each of these cells you have what are the problems we heard about, and what are the potential --
DR. YOUNG: Did you bifurcate, that bifurcation of external accountability, you both look at that under performance assessment and surveillance as well, and then under those --
MS. COLTIN: And then the same thing with this internal analysis and feedback, I look at that as well as point of care under the quality improvement. I don’t think it’s as important which column it falls under but when you do start putting it into a chapter heading you need to address it in both chapters, so if you had a chapter on surveillance and a chapter on performance assessment you might be talking about external accountability in both chapters to some degree.
DR. YOUNG: When you were talking about CMS I was thinking about things we’re doing now, nursing home issues, actually we’re kind of set in both of those, you can argue and there’s what we think is a logical progression between those two.
DR. COLTIN: We’ll have to play with that. I think when we get down to specific that’s when it becomes kind of well where do I put this, where does it belong, and if it turns out that almost everything is belonging in one place we might want to really just discuss it all under that one heading.
MS. KANAAN: So Kathy do you, you were saying that in each cell then you’d talk about the problems that you could then, with respect to that cell matrix, would you also then envision some statement about the solution?
MS. COLTIN: Yes, and then that, those solutions that appear in each chapter under each one of these subheadings would then get collected and put collectively at the end as recommendations, so they would appear two places. They’d appear in the context of the discussion of that particular purpose and that type of measurement, so surveillance outcome measurement, what are the problems we heard about, what potential solutions are recommendations might we make for addressing those. They’d be in each chapter, there would be a recommendation, it could be underlined as a recommendation. And then you would pull those out, and whether they go up in an executive summary the way the IOM always does it, or do they go in the end at the chapter called recommendations, which is how we’ve tended to do it, or both, they’d be pulled out and somewhere listed separately.
DR. EDINGER: What would we do with the things that haven’t actually been covered in testimony, we decide there were certain outcomes or whatever category it is, there are certain things that are important but they’re not, they haven’t been covered yet in testimony, would we recommend, I guess the issue is do we make a recommendation or do we recommend that we have more hearings to come up with a recommendation or put it in the recommendation section and say --
MS. COLTIN: Well, I don’t think we can make recommendations about things we haven’t heard about, so we’re going to talk about the problems we heard about in the testimony we took as opposed to problems that may have emerged since or that we now see that we haven’t really looked there for problems, if we looked we’d probably find them and we would have had testimony. So I think that may help define your agenda for part two.
MR. HUNGATE: Aren’t those things that are published usable as testimony in effect? If they’re in the public domain?
MS. COLTIN: I mean they could be.
MR. HUNGATE: I’m thinking of Health Affairs articles --
MS. COLTIN: In general in the reports that are done by the Committee they don’t generally include a literature review.
MS. KANAAN: We never have and it’s not realistic to think we can do much of that between now and --
MR. HUNGATE: Well, I wasn’t suggesting that. But the current issue of Health Affairs has some specific suggestions along these sorts of avenues that relate to experience from people that are the kinds of people would have testify. It’s public testimony.
MS. KANAAN: The NHII report did, if you’re looking for precedence, that report did in fact cite some of the major things that had been done, information infrastructure, so --
MR. HUNGATE: Crossing the Quality Chasm articles.
DR. YOUNG: If information is peer review, out there and the Committee’s made aware of it, fine, versus an active process that you were talking about --
MS. COLTIN: I don’t think we have resources to do it, but I think if people are aware of very specific articles that are pertinent to something and want to bring it to the attention of the group --
DR. EDINGER: What about the future, instead of maybe having some of the hearings, there are a lot of things like that that are already out there. Instead of having somebody come in and basically recite their Journal article, maybe we can reduce some of that hearing time and basically use these as background materials for these future recommendations. If it’s in Health Affairs why get five people in Health Affairs to come and tell us what they wrote in Health Affairs. Probably read the article and save five people’s travel out there.
DR. HOLMES: Unless there’s an opposing strong view, then there’s a good reason to have a hearing to give different people an opportunity to state their position.
DR. YOUNG: There’s some technical articles, high technical articles versus things that maybe have other perspectives or viewpoints for which there may be different ways --
MS. COLTIN: Which is one of the problems with Health Affairs, a lot of the articles are expert opinion, and some of them aren’t even backed up by empirical data but a lot of them are expert and then you would want to have the countervailing opinion.
MR. HUNGATE: Right, you’d want to have both. Well I have the feeling with the NHII we’re really dealing with uncharted waters, that we’re seeing limits with what we can do now, we’re hopeful that that’s going to help, but unless we’re able to articulate the ways in which we hope it will help I feel like we probably haven’t done quite we need to do.
DR. YOUNG: Do you want to wrap this around the NHII or the NHII around, how do you want?
MS. GREENBERG: Unfortunately I have to go to this conference call but I was just going to say that I really like what Susan did, the way she pulled in both the, a lot of the activity, the NHII, the standards, and the 21st Century Vision Report also, which really talks about all those influences on health in addition to health care, all of which are very relevant to the quality I think. So I think to the extent that the report does that it will grab the attention of the other members because as we know the majority of the members are not on this Committee, but also it will fit in with this effort to get more of a population focus to the Committee’s work and to integrate which we don’t still do that well, I think, so I support that.
DR. EDINGER: The other thing is, if we wrap the report with the NHII, the NHII is still the vision of the future --
MS. GREENBERG: Well, you can’t wrap it in that --
DR. EDINGER: -- other way around, but we don’t have a health information infrastructure --
MS. GREENBERG: Right, and that’s why you’re making some of these recommendations.
DR. EDINGER: That’s what I’m saying, if you wrap this around something that may not come, maybe you should do it the other way around, wrap that into this as possibilities, but look at other short term possibilities, what you can do now, what you can do in the future, and wrap it the other way around.
MS. GREENBERG: I mean there are people who would say, who can’t do any of this without the electronic record. But meanwhile every day people are making you do a lot of it as Kathy can tell us, and we can’t really wait.
DR. YOUNG: I would say the electronic medical record might be transformational in many of these things, and whether that’s NHII or a regional version of that, I mean I don’t know --
MR. HUNGATE: What it really is will evolve. Are we ready to talk about recommendations some more? Or do we have all of them there, just not all --
MS. COLTIN: No, I don’t think we have all of them there, I think the ones that we had there really were those that related primarily to administrative data, the ones that are cross cutting there are the provider identifiers and patient identifiers, which I think pertain to other data sources besides administrative data. But the other recommendations that were all pretty specific I think to administrative data, so that’s the race ethnicity, the lab data, the diagnosis indicator, the linkage of provider to procedure.
MR. HUNGATE: These all link administrative data?
MS. COLTIN: Yes. The provider identifiers and the patient identifiers are cross cutting, doesn’t matter what data system you’re talking about, in fact you would want in some cases to link survey data with administrative data the way CMS links the NCVS with administrative data.
DR. EDINGER: That data though could be, not administrative, you’ve got an electronic medical record you could get lab data off of that, too.
MR. HUNGATE: Does lab data --
MS. COLTIN: I think the assumption is that lab, this is the gap, it isn’t a gap in medical records, medical records generally have lab data, this was identified as a gap in administrative data, they don’t have lab results data.
DR. YOUNG: -- electronic record, too is unlikely.
MS. COLTIN: It wasn’t lab data as a topic, it was the absence of lab data.
DR. YOUNG: I understand now.
MS. COLTIN: So that was specific. I think in terms of surveys there were, I have to go back to some of the specifics --
DR. HOLMES: Well some of the specifics if I recall correctly were low response rates and lack of standardization.
MS. COLTIN: Lack of standardization, which fed into the low response rates because, I mean one of the things we talked about is that if you were trying to measure a patients care experience at the individual physician level, you could have a situation where the same patient gets a survey from the doctors office about their doctor, you have to survey from the medical group or health integrated delivery system that that doctor’s office is part of, and get the survey from the health plan about their experience with that doctor, all asking the same kinds of things in slightly different ways, and because the patient is getting barraged with all these different surveys you get low response rates on all of them because they’re going to pick and choose. If they’re going to answer any at all they’re probably not going to answer all three. And so you end up increasing the sample sizes to compensate for it, which increases the cost, so it’s a real issue. So standardization I think of surveys and coordination around survey administration I think is also important. Because we could all be sending the patient the exact same survey but they’re still getting three of them.
DR. JANES: Or for the large national surveys you’re asked about functional status in many different ways and it’s not that it’s confusing to the person who is answering the questions, it’s confusing when you’re analyzing the data because the items are not similar across surveys.
MS. COLTIN: And the same thing is if you wanted to look at outcomes in patients who, Medicare patients who are in managed care, right now you can’t use any of the other national surveys to do that because they measure functional status differently. Fortunately CMS is administering the outcomes survey in a fee for service sample, too, so now we can do that. But if they wouldn’t have had to do that, if the national surveys measured it the same way, so it creates a redundancy in that area. So I think this question of standardization and coordination around surveys is an important one and it does create these response rate problems, which people are finding, then you don’t have valid measurements if you can’t get good response rates.
MR. HUNGATE: What would be the cost of standardization?
MS. COLTIN: Well, you know, it’s not all that hard. I mean when NCQA was looking to add a measurement of flu immunization in older adults to the CAPS Survey, what they found is that the NCVS asked it one way, the NHIS asked it another way, and the BRFS asked it another way. So they actually brought the federal agencies together and said can we do this, can we all measure it the same way, and everybody changed their surveys to one way of doing it. That’s one example but it could happen in other areas.
DR. YOUNG: So that’s the difference between, the answer to your question is adoption doesn’t cost much, development is a different thing. Standard development thing, then you’re talking about time and money. But adoption is --
MR. HUNGATE: So is this a process recommendation for HHS?
MS. COLTIN: I think it’s for everybody, to agree on the common ways to ask about the same thing.
MS. KANAAN: We don’t want a recommendation that says everybody should do it --
MS. COLTIN: We’re not even saying that it has to be the exact same instrument, what we’re saying is if you’re measuring something in particular for quality purposes that you measure it the same way, that you ask the question the same way.
MS. KANAAN: Is there a way to translate that into the Secretary should do such and such or the Department should do such and such?
MS. COLTIN: Yes, should promote standardization across national and federally supported state surveys in how the same aspects of quality are measured. And I think we could use the flu example, because it actually came up I think, I don’t know if it was Steve Klauser(?) or somebody who when they came and presented it to us used that example I think, about that happening. In which case it’s a good example of how it can occur.
DR. JANES: In which case, I mean the role played by the Department is really one of just identification of the issue and then convening of the interested parties. And then as you said, Scott, it’s not an issue of standard development in most of these instances, it’s just the standard is implicit in the three approaches and one of them is chosen and you can always use the argument of benchmarking or --
DR. YOUNG: This is CHI model, Consolidated Health Informatics model, and CHI model states that the federal government shall adopt X standards, and although they’re talking about IT standards, it’s useful here in the mode, is useful for what we’re talking about. The understanding is is that the federal government is going to be big enough that whenever they say this is the standard that’s reasonable that everybody else is going to fall in line about this. Because I don’t even know if CHI decision has been made saying we’re going to tie purchasing dollars to this yet. It’s just making that statement and you could have a similar model here for Medicare and Medicaid, that’s 80 million people, that’s a significant impact on the market. I guess the big question is whether you can bring agencies outside HHS into this, whether there might be a VA argument or not, and I don’t know that. I think that would be, but I think just a federal source --
MS. COLTIN: I think we have to recognize that there are limits to standardization and that we don’t want to promote it to the exclusion of innovation. But where in fact there is a good standard out there that it makes sense to the extent possible to adopt that and to promote consistency across the surveys, at least in terms of core items allowing perhaps experimentation with new items.
DR. YOUNG: You’re right, we have to get something out there initially, but I mean this is the perfect me and the enemy and the good, which is what in IT we used to get hung up on this for ten years, this is just going to be it and we’re going to start out here, and you’re right, we need to revisit these things and improve them and allow for that further dialogue to occur. And the industry actually welcomed that I think.
MS. COLTIN: There’s been a major ongoing effort around surveys within HHS trying to integrate them and coordinate and reduce duplication and whatever, and I think we should acknowledge that and support it but just identify that it could go further in some areas.
MS. CARR: Just thinking about the report, we would want to be and I’m sure tight, not too redundant, etc., and given that we don’t have potential recommendations in all of those little cells, we might talk about as much as possible the issues that arose and the potential thinking around solutions in the little sections. But then hope our recommendations until the end, maybe put them in front, but build the major ones around given these issues and potential solutions, we think it can be attacked in these more narrow more specific ways in the short term, focusing on HIPAA, focusing on national survey coordination, focusing on X, Y, and Z, which don’t necessarily fit in any little box well, and I would want to --
MS. COLTIN: The other thing we’ve done in some is make a recommendation like the example Marjorie gave about asking that they do more demonstrations or conduct more analysis to evaluate these things. That can be a recommendation, too, it doesn’t always have to be a recommendation for a solution, it can be a recommendation for a process to arrive at a solution.
MS. CARR: I’m just suggesting we needn’t necessarily have them in every cell in terms of what our recommendations are, because our recommendations are being more --
MS. COLTIN: No, I think you’re right.
DR. HOLMES: One more issue I’d like to raise with respect to the surveys, which that might be a possible recommendation, has to do with some things that we found out with the National Health Care Quality and Disparities Report, which is that we’re having a terrible problem with the various surveys that are in this first iteration of the report with having sufficient sample size for certain minority and ethnic groups. And it’s not because of low response rates, I mean that could, the main issue is that these are groups that have low representation in the population overall so that you either have to, you have to increase your sample size or you have to over sample in areas where you will find these minority ethnicity groups. And I think this is something that will become increasingly publicized once the quality and the disparities reports come out because we have a lot of measures where we can talk about, for example, white/black differences only, and if you want to talk about black/white differences getting down to age you can’t talk about it because you don’t have the sample size. And if you want to talk about Asian Americans Pacific Islanders versus other groups we certainly don’t have enough sample size for that.
MS. KANAAN: So the recommendation is more over sampling?
DR. HOLMES: Well some approach, to promote some approach to increase the sampling ratio of under represented minorities and ethnic groups, whether it’s over sampling or whether it’s increasing sample size of certain surveys.
MS. COLTIN: And I think the other strategy that has come up as a recommendation sometimes is this rotation idea, that this year you over sample this, next year you over sample that, so that over time you’re able to put a picture together that’s more reasonable. And that concept also came up in the context of the provider surveys, the national provider surveys as potentially a vehicle to do more in the way of condition, deeper condition specific measurements, that it might be that this year it’s asthma, and next year it’s diabetes, and the year after that it’s something else where you could actually get much richer, more robust data about what’s going on, particularly in the ambulatory care survey and the hospital ambulatory, about specific conditions. But that the way, without making the survey unwieldy you could just have condition specific modules that rotate as a way to get national benchmarks, because that was one of the concerns is that for a lot of these quality measures that are out there right now, health plans are reporting publicly but nobody else is and with the exception of what I really applaud as CMS’s efforts to get fee for service comparative data out for a lot of these measures, where even they’re struggling with trying to do that for some of them. There aren’t good national benchmarks and to the extent that some of the national surveys can serve as a vehicle for that I think that would be helpful.
MS. KANAAN: A question that I have is that while the Committee as a whole has looked at a lot of this kind of problem and made recommendations and so on, I’m not sure that we can, that we have much input through the testimony on this question, particularly the sub population issues around surveys and what do you --
MS. COLTIN: It’s in our letter to AHRQ on the National Quality Report, actually, it was in our recommendations to them about taking the data down to the level of the state in many areas where they don’t because it needs to be actionable and that that would require increasing the sample sizes. So I don’t know --
DR. HOLMES: On page two of the themes under lack specific data we have, it would be under incompleteness gaps related to, we list a whole series but one is race ethnicity and those disparities.
MS. KANAAN: There wasn’t a lot said out it, if you look at the summary, but I think --
MS. COLTIN: I don’t know if it was specific to surveys, I don’t think it was, I think it was brought up in the context more of administrative data, but it’s an issue that was identified and maybe it’s our prerogative to say the solution may not lie entirely in administrative data, it may be that in order to look at these things we need to think about the surveys as well.
MS. KANAAN: I would think that rather than holding ourselves to a strict requirement that we only address those things on which we have a reasonable amount of testimony, that we may need to want to cover the subject and put in testimony wherever it’s available --
MS. COLTIN: But the other thing, Susan, and this was the point that I was making before about in our charge and work plan where we said we had rolled our interests into the testimony that was taken about functional status, that testimony was held by the Populations Committee, not this Workgroup, but we had specifically decided we were going to do it that way just as our panels were held in the context of the full Committee. So drawing from testimony that was taken to the Populations Subcommittee is entirely appropriate and there was lots of testimony about race ethnicity and survey sample sizes to the Populations Subcommittee. So I see no problem in drawing on that testimony at all. Why would we mount a separate set of testimony on race ethnicity when they were doing it? Our goal was to try to work through them to minimize effort and people’s time.
MS. CARR: And I would also think that members of the Committee are on the Committee because of your expertise and things that you brought to the table and agreed on are equally valid support.
MR. HUNGATE: I share that feeling.
MS. COLTIN: I agree. And that’s always been the case, however, to the extent that we can quote testimony it strengthens our report.
MS. JACKSON: One entity, NCVHS, the entity itself, so you’re free to pull whatever material and as you’re saying, speaking of your own area of expertise, so you use your testimony as background guidance direction and then kind of go from there, so you’re fine.
MS. COLTIN: And I would particularly, Susan, refer to those two days in February of 2002, there was two days of testimony in February of 2002 where we brought in just about every lead person for every national survey to talk to us about race ethnicity and the issues, and adequacy of sample size was a common theme across those two days of testimony.
MS. KANAAN: That was to the Populations Subcommittee.
MS. COLTIN: And I sat there as the chair of this Workgroup probing these issues.
MR. HUNGATE: You suffered through that.
MS. COLTIN: I think that is something reasonable to include in our recommendations. And I think I would also add sample size issues around being able to look at the state level because most of the solutions in the quality improvement interventions that need to occur around these things are really happening at the state level and if they can only get the data at the national level it’s not actionable. So we had made that recommendation in our letter as well.
MS. KANAAN: So I presume that those are three sub categories of recommendations, of a big recommendation related to surveys, standardization, coordination, or greater specificity one way or another about ethnic subgroups, and this state level data, is that how you see it?
DR. HOLMES: Well it has, both of those last ones have to do I think with sample size, or sampling method in order to increase the numbers of minority race ethnicity groups, either at the national level but also at the state level.
DR. JANES: Now I’m not sure I would phrase it as specificity, it depends on what you’re talking about. The specificity of reported data is poor because we have, not because we have insufficient categories on the surveys but because we have insufficient data to give robust estimates.
MS. KANAAN: I was more, not so much checking the wording with you but checking in the way we’ve bundled these, these are all parts of the --
DR. EDINGER: Was there something on SES data, about the lack of SES data, the need for more SES data?
DR. JANES: We’ve talked about that in a lot of different contexts, we’ve talked about it in the context of administrative data --
MS. COLTIN: I don’t recall talking about it in the survey, most of the surveys I thought had pretty good, I mean the one --
MR. HUNGATE: They don’t necessarily standardize it though.
MS. COLTIN: The other problem we did hear though about the national surveys was in many cases, this relates to your bucket four, a lack of contextual data about, if you were breaking the data down to a certain level, county or whatever, what did you know about other things that were indicative of the county, because you’re collecting information at the individual level, but those individuals live within the context of an environment that has things going on that can help you interpret the data, and that most of the surveys didn’t seem to have a good flavor of the context around a lot of them.
DR. JANES: Which you could get from linking in the Census data, that sort of --
MS. CARR: I would just not want to, I’m not saying it would happen in how you characterize that, in how we bundle to make something seem subservient to how you get the data versus what the data is, what the data are. I mean types of data, we certainly I gather want to emphasize race ethnicity functional status, and maybe some other things which you all know, but it almost gets hidden if you say in surveys let’s make sure we have enough, I mean that’s a sub of that issue, I would put the issue, data type issue, highlight it in some way, and maybe that comes out somewhere else.
MS. KANAAN: Free standing recommendation.
MS. COLTIN: I think it may come under quality improvement because I think it really is an issue in large times about making the data actionable. If you don’t have it at the level where the improvement can occur or the resources exist, you don’t have it at the state level in the case of some of these surveys, I think even in terms of CAPS, I think you can make a wonderful argument for a lot of the measures in CAPS that the health plan isn’t the right level to measure doctor communication, it’s at the doctor level that you’ll measure that. But we don’t have sufficient sample size to look at it at the doctor level in those surveys, the sample sizes are based on comparing plans, not comparing doctors, and the sampling frame is not constructed in a way that allows you to look at doctors. So I think these issues are relevant to a discussion of their usefulness for quality improvement, too. So which column you put it in, I don’t know.
MR. HUNGATE: You’ve got to run. Thank you for joining us. Are there any last observations on recommendations that you want to make?
MS. COLTIN: Well, I think we hit the highlights under surveys and I think we hit the biggest highlights under administrative data. The one thing that may be a little weak there that isn’t quite clear is the issue of linking records across settings and identifiers are one part of the problem but interoperability of the systems and some of the other things that support linkage privacy issues, I think just the whole concept of facilitating linkage of records is an important one, even the notion which I’m not opposed to saying again even though it falls on deaf ears, is the issue of a common procedure coding system. When you’re trying to look at people who got a particular procedure and you start looking at combining people who had inpatient surgery and people who had outpatient surgery, the same thing is coded differently, and it does create issues for measurement. So even though nothing may happen with that recommendation, it doesn’t hurt to go on record and it reiterates a recommendation that this Committee has been making for years that has never gone anywhere, but it’s worth saying one more time.
MS. KANAAN: So you’re suggesting a recommendation, a general recommendation around various ways of facilitating linkage.
MS. COLTIN: And like I said, we had common identifiers down there as one but I think there are other recommendations we can make to support record linkage across care settings and providers, which would be sort of the overarching thing. What we didn’t talk about today is recommendations that relate to medical records, and overall I certainly would see us wanting to support promoting the adoption of electronic medical records, I mean that’s the obvious one, and drawing on the work of the Standards and Security Committee around recommendations for common standards for medical records and terminology and so forth, all of that. The issue I raised earlier about the function of being able to look at populations as well as individuals, that record systems should have that capability.
MS. KANAAN: But does the Workgroup have recommendations for the Subcommittee on Standards and Security about data content that needs to be addressed through the medical record?
MS. COLTIN: You have to go back through the specifics that you pulled out in the testimony and try and refresh my memory about which of those were specific to electronic medical records but --
MS. KANAAN: I was just thinking that this group might have some inputs for that group, it’s not just a matter of supporting their work but maybe --
MS. COLTIN: Well I think that issue, that functionality issue about being able to look across patients is one that they may not be thinking about and one that really is critical for quality purposes.
DR. YOUNG: Well on the surveillance front, having kind of a common definition of an EMR or EHR, which kind of combines this clinical aspect as well as the administrative aspects into one functional record becomes useful because I’m not looking at different systems to gather data from and so forth, since we’re saying surveillance is important it’s important to say that we’re going to have this, if we go there this patient record should contain these different aspects of it, administrative as well as clinical data. So I don’t know where you put that in, process, structure --
MR. HUNGATE: Does that relate, I guess not to provider, as being admin data, EHR and EMR being mutually compatible and amountable?
DR. YOUNG: Our dialogue has been but that becomes one thing, we call that an EHR, that we have ambulatory or CPOE, EMR and admin thing under this one data set that’s a discreet data set that we call an EHR. And to the degree that we know what that is and know what we’re talking about when we say that, that we’ve standardized that definition is, we’re not there yet. It’s a stopping point for us in going forward with demonstration projects and other projects, I mean it’s a stopping point.
MR. HUNGATE: Well, I think we have to think about how our recommendations can help remove the stopping points, but that’s part of what this function should serve, so suggestions in that order are --
DR. YOUNG: -- talking about when I had to leave here this morning, I went to a meeting on these demonstrations and it was vision and design demonstration around EHR’s but if a doctor comes up or a health plan comes up and says well we have one, pay us for that, we would just assume be the people who define that on this end and be able to gather data from that, wind up being able to report quality date and performance data downstream. So that’s kind of a very real time need, but we’ve not necessarily seen NCVHS address explicitly, so if this Committee could do that, or maybe they have and we just haven’t found it.
MR. HUNGATE: So you’re thinking of the patient center electronic health record?
DR. YOUNG: No, at this way we think of the iterations of this as being an electronic health record as being kind of the first iteration and seeing those diffused and adopted broadly. And then the second iteration, or maybe the third iteration being a personal health record, and that’s that very portable and adaptable thing that the people will have. And then the next iteration above that becomes regional connectivity, and interoperability between those systems, I mean you have to kind of build the centers first and then connect them up.
DR. STEINWACHS: But the electronic health record you were talking about one for facility, one for -- administrative and clinical data.
DR. YOUNG: Right, and people could kind of bring their personal health record and plug it into that, but I mean at some point you connect those. But if you don’t plan it out, from this perspective, you suddenly get to that regional connectivity timeframe and you’re going whoops. Everybody kind of defined their own thing and now we’re connecting diesel engines and gasoline engines and they’re not talking around.
DR. JANES: I thought HIPAA initially included a piece which actually the Committee was going to weigh in on this issue of standards and that it was decided to take that off, at least off the front burner. Anybody remember? That was one of the things that was on the list when HIPAA first came out in terms of standards that were going to be recommended, I do remember that, and it came off.
DR. YOUNG: We went to CHI and said is this something you want to take on, and they said no, we really don’t think that’s kind of our thing, but it’s one of those kind of odd issues that become problematic for us to be able to tear down barriers. And I was just thinking of this as maybe the right place to at least as a call to action I would say more than anything else.
DR. JANES: Do you know why the Department decided to step back from that issue of standards for patient?
DR. YOUNG: Well I don’t know that they have, I don’t know that people actually got to the point where we’ve gotten to now and said this is really an issue, I mean now we’re at a tipping point and we’re saying we need to operationalize this thing and the Secretary’s saying this and the President and Tom Skulley, and saying ok, how do you do that. And then as you work through it, how do you do that, you get to the point of going oops, you have to have, as well as having LOINK and SNOMED and all these other things there, you have to be able to say what is an electronic health record. And what are the standards on that, and more than that where are the functionalities and interoperabilities on that as well.
DR. EDINGER: On the personal health record -- had this discussion a number of years ago, they started to do work on it, one of the problems they had was the lack of standardization and agreement as to what are the elements of a personal health record, because like your own, you go to five different hospitals and they ask you for different information, which is one of the major pains in the neck of having personal health record. Many said the lack of agreement and standardization and consensus as to what should be on there was one of the problems that they ran into when they were originally thinking of working in that area, so they just gave up after a while because they said that was a major headache.
MS. KANAAN: These kinds of questions are being addressed and have been addressed by the NHII Workgroup and the NHII Report, some of the different elements that you mentioned being integrated and being compatible, so that’s --
MR. HUNGATE: It’s there, but it’s not well evolved yet.
MS. KANAAN: It’s not operational.
MS. JACKSON: And will be explored further, personal health dimensions hearings in early June, is it going to be the cross reference for that?
MR. HUNGATE: Probably should be. I’m going to both functions, both activities, and in thinking about this workgroup I had felt that the personal health record might be appropriately worked on here. I’m not sure of that because I don’t know what I mean by that.
DR. YOUNG: We don’t either know what you mean by that, we think it’s an important thing, and we know a lot of different people have defined it, but --
MR. HUNGATE: But I come from over in the corner of the quality thing with this nagging thing in the back of my mind that says quality is meeting or exceeding expectations. And somehow that relates to my own personal characteristics and what the system can do to fix things that are wrong. And I can’t get right now good expectations from the system and so the system is not quality driven from that standpoint. So I have a personal wish to be able to get that out of the system, which requires a lot more information than is now there.
DR. EDINGER: [Comment off microphone.]
MR. HUNGATE: Doesn’t even come close.
MS. KANAAN: It feels to me as though that that would be an important part of the discussion in phase two when we’re getting into first and center, rather looking at health quality --
MR. HUNGATE: I think when we start talking about your terms as the measurement system for the system, it’s got to be in individual terms and it’s got to tie the personal health record which has to tie to the medical record in the system.
DR. EDINGER: I’m wondering, electronic medical record which basically both groups are interested in, maybe we should look into the idea, maybe like joint hearings in a sense, since both groups are interested but different viewpoints, or maybe even different viewpoints, or maybe the same, that that might be something both groups tackle together rather than separately and sort of feeding into each other, maybe there could possibly a joint group effort. At least on the areas where we’re looking at what the individual elements are and they may be looking at interoperability standards, they’re both important and different but they’re related and you really can’t separate, if you have one without the other they’re both useless. So maybe we can do some cross fertilization or maybe the staff and members here can go sit in on it or vice versa whenever we have hearings on that issue we both be there or something instead of doing it separately or doing twice.
MR. HUNGATE: Well now somebody could help me on this I’m sure. I’ve heard about the Markle Foundation and its funding, is it working on personal health record?
DR. YOUNG: Well, the Markle Foundation funded two entities, they funded Connecting for Health and the eHealth Initiative, and Connecting for Health is really kind of this collaborative body, there’s a lot of industry people in there, some academic folks, and they brought in government folks to really kind of talk about the barriers and different ways to lower those barriers, so that’s Connecting for Health, and medical providers are in there as well. And then there’s eHealth Initiative, that have really kind of produced this, as best I can find, is produced this thing called the Marconi(?) Project, which is some middle ware that helps translate data that we would use at CMS on the seventh scope of work and the FDA’s going to be using, I think CDC as well is involved, that’s our understanding. And it’s really trying to introduce some authenticity into passing data back and forth. Which oddly enough we have a problem with authenticity, because you can’t let people inside the data sets, inside CMS, kind of dangerous to open a hole up in our data sets we think, so it’s kind of a weird thing. We desire it but we can’t allow it. And that has a fairly significant IBM progents to create middle ware, but those things are, they’re staffed by the same people, Connecting for Health and eHealth Initiative, the same office, I sometimes get confused, it gets blurry where that comes in. But they’ve done a lot in driving the agenda forward and continue to do that.
MR. HUNGATE: Well, I think we should be closely linked to that, and how we articulate that in this structure is a little unclear to me but I think it probably moves over to the right hand side.
DR. YOUNG: It might be Carol Diamond, might be the contact for Markle, I don’t know Carol or Jen. Janet’s here in town, Carol is at Markle, which is New York, right?
MS. KANAAN: I was wondering if I could raise a process question. Not to throw a money wrench but to be reality based about where we are in our process. And I haven’t made my mind up about this but just to raise the question, we have a half hour left, we’re talking about a report, presenting a report to the Committee that would have to be ready in six weeks essentially to be able to get it out. I’m wondering if we need to push back our deadline to September and have another, this has been such a rich wonderful discussion that seems to me laying the platform not just for summarizing what’s come before but what’s coming ahead. Are you seeing it that way, Bob?
MR. HUNGATE: I don’t think it’s practical to bring a report to fruition in six weeks.
DR. STEINWACHS: Bob, Susan lives in the wine country of California, that wine can mellow you, I think six weeks --
MS. KANAAN: I had an opportunity to talk to Kathy a little bit about the process and development in this report and I think if things can be worked out with NCHS she is willing to be an author in this, which is wonderful and we will have a much better product, but it frankly makes it a more challenging process to produce a report like that when you’re collaborating closely with someone else. So you’ll have a better product but it will also take longer.
MR. HUNGATE: The thing I don’t feel we have in this discussion is enough strong provider per se, you a provider, a provider from your own experience, but not in current role. Kathy is very strong in the managed care understanding of the major issues as they face them, but frankly I have no faith that the managed care program is going to answer my questions up in the right hand corner. And so I don’t think we’ve addressed that yet.
DR. YOUNG: So what are you thinking of, are you thinking of association representative or --
MR. HUNGATE: Just content. It may be that we need it through the personal health record, maybe it’s, I’m not sure where to get it yet.
DR. YOUNG: The EHR as well --
MR. HUNGATE: You could get it through the EHR --
DR. YOUNG: For the provider, that’s what’s going to be the investment, that’s the belly button for the provider is the EHR.
MR. HUNGATE: I have the feeling for the information set that’s needed to answer the questions that I want answered as a patient, and they’re not dealt with yet in this data set.
DR. HOLMES: But maybe, I thought we were getting at that might be our phase two focus, that we would widen the work, we would make recommendations based on --
MR. HUNGATE: I believe it is. But we have to describe that transition adequately, that’s all I’m trying to express. It has to show up here in a way that it’s understandable, and I don’t know what it’s going to take to get to that.
MS. KANAAN: Are there people on the Committee, you may not know this Bob, I don’t know that I do either, but I keep thinking in different contexts as this discussion goes forward about the potential of engaging other members of the Committee who are not on this Workgroup in this discussion because we keep resonating and bouncing off other things that are going on and I’m wondering if there would be things to be gained from trying to engage possibly the full Committee in some, so if you get some more insight on some of these questions, if other people of the Committee, some of them are physicians, could participate, then you get a stronger populations perspective, etc.
MR. HUNGATE: Yes, I think that’s a useful suggestion. There are two pieces that work there. There’s an executive committee meeting which has a conference call here in another week or so --
DR. EDINGER: There’s a meeting in July or August, too, face to face, it could be July or August, it won’t be Chicago I guess.
MR. HUNGATE: Is that true?
DR. EDINGER: Yes. Maybe July, John was in Chicago to have a retreat so usually they --
MR. HUNGATE: Well I haven’t heard of anything.
DR. EDINGER: Marjorie told me it happens at the agenda time --
DR. JANES: I would argue with at least not initially limiting it to the executive committee. I guess what I would argue was just getting in on the agenda for the June meeting.
MR. HUNGATE: So what I was going to suggest beyond that is I’ll talk about this issue at the next executive committee meeting and articulate it, and say we need some time before, and what I’m trying to think about is things that are underway at CMS, current things that are working or being done as being a way to get the Committee tuned to the quality aspect of the information agenda. That was part of what I was thinking about with this request in Health Affairs on mortality data again. To bring that issue squarely before the full Committee to talk about what are the issues between public release of outcome data, because that ties back to risk adjustment information which we have talked about here.
DR. EDINGER: Actually it might be nice in September, the Disparities Report, the National Health Care Quality Report, the Patient Safety Report, all these reports basically, theoretically due to Congress.
DR. HOLMES: They better be out.
DR. EDINGER: Well, in any event, that might be a nice time to discuss the issue of the national survey, which is the individual health, the difference between the two and the fact that now what are these reports and the viewpoint of them is, but we still don’t have this half, your individual quality of your own care. You may have a national picture, you may have general state level, but don’t have you. That might be nice to show the difference. I’m not even sure the Congressional staff who wrote it had an idea which they actually wanted, they may not have actually realized there was a difference.
MS. KANAAN: So if you were going to talk to the Committee, try to engage them in a conversation in June, a fruitful conversation, would you want some kind of a background document for them to read or --
MR. HUNGATE: Well, we probably want to have a start, something, I don’t know what yet. Our framework probably.
DR. STEINDEL: I think if you took the framework and then a brief synopsis of what has been done so far, the work related framework and then what we’re starting to talk about as a phase two, and you can do that in a couple, three pages possibly, and it seems to me that gives the Committee something that actually says well are you guys on target as far as the Committee is concerned, are we going in the right direction or not.
MR. HUNGATE: I like the idea of giving people a little more settling time on accepting recommendations, so I’d like to give them a preview of here’s the direction the recommendations are going, it’s not final yet but here’s what’s coming, just because I think it’s a good process.
DR. EDINGER: And also give the other Committees who are overlapping time to think about that overlap and how much they want to participate or feed into it.
MR. HUNGATE: Yes, we’d like to stimulate that if it works. The other piece of that, and this is again a thought, HHS, CMS, are the sole system for end stage renal disease, ESRD. I would like to assess whether that is a good example of a quality driven program in my terms of how it does for each patient and how well is it measured, because there there’s nobody else involved. And so my inclination is to say let’s take this in-house program, pull it up and look at it in our terms before the NCVHS and see where we end up.
MS. CARR: I understand your statement with renal disease, but the care is actually given by others, right?
MR. HUNGATE: Absolutely, all it is is the payer.
MS. CARR: -- how that assessing their performance is more like park(?), what OMB is doing, I’m not quite sure I understand assessing CMS in that disease versus how that disease is treated in the world.
DR. YOUNG: I would say both, you can assess this model that CMS has very purposefully created around end stage renal disease, it’s not a fragmented payer stream, it’s a streamline payer stream so we can say presumably this is how we want to see care given and help shape it. It’s unique in that, but the only other thing that would be any different would be HIV, because CMS, other than Ryan White, we’re the number one of three payer, Ryan White being in the middle between Medicaid and Medicare on HIV. But end stage renal disease is fairly pure in that CMS, but you’re right, there’s this entire network out there like privates who are going --
MS. KANAAN: But you’re saying that in terms of external accountability there’s only one group calling the shots.
DR. JANES: And actually I thought there was basically only three providers now.
MR. HUNGATE: And I think there’s quite a bit of work to make that into a quality improving system --
DR. YOUNG: We just brought in a new kind of end stage renal disease czar --
MR. HUNGATE: Which should help us think about if we’re trying to do a measurement structure, if that’s where we’re thinking we ought to be working, then this is an example of the current state of the art by the only place that really exists. I’m not ready to say we ought to say we ought to do that, but I want to think about it and have people tell me that’s a dumb thing to do or that’s a good thing to do.
DR. EDINGER: It might be helpful to get some background information about what they’re doing and look at it first before bringing somebody in, you never know what kind of can of worms that could bring up, advocacy groups do it in open hearings but it might be good just to have some background on the program, we can look at it and then decide whether or not it’s something we want to do in an open type hearing.
DR. YOUNG: I think that that would be, it’s useful in that it’s kind of this sole payer, and therefore is presumably shaping the scope and quality of care. So to get something on it from that perspective, yes, that would be good.
MS. CARR: -- go about it politically as well, because when I think sole payer I think national health insurance --
DR. YOUNG: This is Reinhart’s(?) algorithm on who pays for health insurance in the United States and it says is the organ over 65, Medicare, unless it’s a kidney, Medicare, or is the organ under 65, somebody else, unless it’s a kidney then it goes to Medicare, it’s this great little, it was a Christmas card he gave actually, where he sent this out on how is health care paid for in America.
DR. HOLMES: And in terms of being a condition, it’s not one of the priority conditions that was identified by the IOM to be kind of in the vanguard of looking at quality measurement, and it also probably doesn’t affect as large a proportion of the population as some other disease entities might in terms of looking at quality measurement.
DR. YOUNG: But would you be looking at it from the size of the impact or for as a model, to learn things from?
DR. HOLMES: As a model with respect to what is going on in this committee.
DR. YOUNG: It’s a chronic disease which requires a significant technological intervention, it’s expensive technology intervention from that perspective, a lot of the chronic disease we have we try to not have expensive technological interventions, and so you would say maybe this is not a great model.
DR. HOLMES: If we’re supposed to be looking at the quality of the information to support quality measurement as opposed to the quality of a particular program.
MR. HUNGATE: That’s correct, and that’s what I was trying to get at, is how they’re using information within that and not judgmentally about the program.
DR. YOUNG: And I’d have to bring somebody in, I’m not that close to the ESRD program.
MR. HUNGATE: Given all the pitfalls that have been articulated in this discussion it would be inappropriate to do at the full Committee and probably better to do it here if it were appropriate, because it’s much more contained.
DR. STEINWACHS: Well, I think it would be good to try and figure out how to focus the discussion, at least this stuff that Neil -- lectures on, does all the work in that area, and there are differences in the for profit, not for profit in outcomes, there are differences racially in organs and receipt of organs and you can go down the list. Well, who’s responsible, it drives in accountability issues, it’s an interesting problem because it’s clear that the payer necessarily is even though you wield a big stick, and then there are different ways of doing dialysis, at home and so on and those have differences also. And so I think the trick is probably what you’re saying, we’d have to come into this with a very clear idea so what are we discussing here and what are we using this probe as an example to try and amplify, and I don’t necessarily think in this case that it is a lack of information, it’s who’s going to use this information how --
MR. HUNGATE: But I think we may have the same problem.
DR. YOUNG: I’d also have to wonder how generalizable it would be and I’ll tell you why, this right here, U.S. Congress. I mean there are a lot of tweaks and changes that go on in the ESRD program every year that come via statute, and they’re not necessarily driven out of, I shouldn’t say this, there’s different policy drivers that drive these decisions and so they’re there, and they’re going to contaminate how you’re going to look at this.
DR. STEINWACHS: Well, I guess the difference between the ESRD and maybe what we were talking about is that what we were talking about probably comes back to the idea of the states have most of the power and concept if they were exercised. ESRD doesn’t quite look that way, even though it still sits within a state context because there’s one major private company that delivers all these services. So it would be nice to see somebody take an example, I like your idea about trying to pick an example and sort of work it through to look at our framework and look at how we’re making recommendations that are helpful. Because it could make the report eventually come alive, instead of saying recommendation X, Y, & Z, well in the case of, if you look at people with this condition sort of means, or what a personal health record eventually means as you get down to the line.
MR. HUNGATE: I’m sure that if you could go far enough upstream on that program you could make a big difference on the number of people who got into the program. Some of those failures are preventable back upstream.
DR. STEINWACHS: Well, Medicare won’t pay for that.
MR. HUNGATE: But if you do the quality upstream you might save a lot of money, and it would all be in one pocketbook, and so the return for doing the system job is pretty high there, so I want to try as I say get out of the box and think about what are the things that are going on that we can work with to understand these issues better, and help us better in what we’re trying to do, and so I thought of two issues that came to mind, and you folks are closer to it than I am, probably have some others.
MS. KANAAN: Well it sounds like from the list of variables that you were mentioning, Don, it sounds like a good case study for sort of phase two for this broader thing because there are the political factors and the economic and racial and ethnic, all of these other determinants of health, it would give us a chance to see what information there is in each of those categories.
MS. CARR: In terms of the, now what’s the September --
MR. HUNGATE: I think September’s the right timeframe.
MS. CARR: The Committee, does the Committee envision that as summarizing the past work, stating recommendations that grow out of the past work, and looking towards some potential directions for the future. So that the recommendations are based on what was done up to 2002, up to 2003. Is that right?
DR. HOLMES: Exactly.
MS. CARR: And we are strong enough in those recommendations, we do have to have recommendations in this? It’s not just a status report?
MS. JACKSON: It is useful, it kind of shows your direction, it shows, it’s got more oomph. I don’t know another way to do that, when you make a statement and stick your neck out --
DR. STEINWACHS: [Comment off microphone.]
MS. CARR: I was trying to hear your desire to get us moving and I think we need to do that but not compromising, that’s not the right word, but it’s a closure.
MR. HUNGATE: I agree.
MS. KANAAN: Well it seems as though what’s been useful about this discussion and the one in February is that it allows us to lay out a broader conceptual framework within which this little piece that we’re going to report on findings and recommendations, where it fits in that broader context. And then we’ll move into the broader context.
MR. HUNGATE: I think talking about accountability and risk adjusted outcomes is an important piece, it’s a conclusion which is a big conclusion, it’s not a detailed conclusion as some of the others are. These are data gaps. If you say that part of what you got to get out of this data set is risk adjustment information for valid outcome measures, that’s going beyond what a lot of the testimony was saying, I think. But I think it’s something we could conclude --
MS. KANAAN: Are you saying you’d like to see that be a recommendation, one of the recommendations of the report, the September report?
MR. HUNGATE: If I look at the quality measures that are out there now, they’re not useful to me as an individual, they won’t make the system better for me as I see it. And so I’m dissatisfied with the quality measures that we now have, and I need a way to articulate that.
MS. CARR: Are you speaking now as a “consumer”?
MR. HUNGATE: I’m speaking as a consumer. I’ve been on the business side of health care for a long time, and was involved in the Washington Business Group on Health during the early HEDIS measure development, and HEDIS has had a tremendous amount of energy go into it but it is still not very useful to individuals. At the Group Insurance Commission in Massachusetts we tell people to pay attention to HEDIS, they don’t care, it’s just not beneficial to them. What would be beneficial is at the point of when you need care.
DR. JANES: Well that statement implies to me that we should certainly be very explicit as we work on this about, just as you do with a cost effectiveness study, about our perspective. Now we’re going to speak from the societal perspective, now we’re going to speak from the health care system, from the individual.
MR. HUNGATE: I have a strong bias in that direction. That’s where the flaw in the system is, is the individual’s under represented.
DR. JANES: I do think if we shift without being explicit about it then it’s going to be confusing to the --
MR. HUNGATE: And that’s why I keep dwelling on it, I’m sorry to do that because it is in a way dysfunctional in getting the hard recommendations, but I feel the need to do it.
DR. HOLMES: I was just going to say we had stated before that we wanted to make, and I think Susan did some in her summary, that we wanted to make the distinction between looking, the perspective of looking at quality of care of the system versus quality of care in terms of patient outcomes, and that even though we can say some things about patient, data related to patient outcomes in this first summary, that we can’t, that most of the testimony dealt with quality of care of the system that took place in phase one, and that consequently in phase two we’re going to take this different perspective and begin to focus more on data needs to measure quality at the individual patient outcome level.
DR. JANES: That brings us into the issue of health versus health care.
MS. CARR: But I think we should be careful, I think our focus still is measuring the quality of care and that quality of care can be indicated by outcomes, it could be indicated by process, it could be indicated by structure, and we have a lot of work around process and structure and that’s where the testimony did, but now we want to move to how do you see quality in a patient outcome, but it’s still quality of the system that we’re looking at.
DR. HOLMES: Right, it goes back to what Kathryn was saying, that it’s the quality of the patient --
MR. HUNGATE: Got to be careful there because we don’t really want to be judges of the health care system I don’t think.
MS. CARR: But that’s what we are.
DR. HOLMES: I think that’s our mandate.
MR. HUNGATE: Judge and judge and jury.
MS. KANAAN: This charge, and by the way this is a process issue that some people, maybe everybody knows, there’s nothing cast in stone about this charge. If you want to go back to the Committee and say we want the word health care, the word care taken out of our charge and we want it to talk about quality --
MS. CARR: And I’m not suggesting we’re the judge, but if we’re talking about quality somebody is making the judgement based on information, and our task is to look at the data and information and measurement systems that people need to make those judgments.
DR. YOUNG: The transparency issue that we’re providing, not the judgment issue but the transparency.
MR. HUNGATE: The problem that I am worried about is that I was here in Washington at the creation of -- originally and its evolution, and the creation of the pork projects. The pork projects did a lot of work, they did not result in outcome measures that are in use, they resulted a lot of guidelines, many of those guidelines were too large to get work. And so I conclude from that that we need to manage the chaos in the system by moving from process steps to outcome steps, and doing that pretty hard. And that means moving to a focus on health more than health care, and that means moving to a system where you don’t necessarily serve as judge, but create the conditions whereby the judgment will be accurate. And that’s a different task.
DR. YOUNG: Let me ask just a question. I think about this, I’m a consumer and I’m looking for a place for my kids to see a doctor, and it’s usually at least the first iteration I’m looking for a provider, it’s a doctor or a nurse practitioner, or whatever, I’m not, we talk about hospital specific stuff, but I usually don’t start out there. I usually start out with a provider and I can find a lot of hospital stuff out there, hospitals actually advertise, they actually push this data out --
MR. HUNGATE: How much of it do you believe?
DR. YOUNG: Well I don’t know if I believe it, I mean I don’t know if I believe the on time things at the airlines either, but it’s there, it’s a place for me to react from, and it’s available to me, at least from the hospitals easily and at a decision point. But whenever I’m making a decision about whether to engage a provider inside health care, I don’t think I have readily available, what I would consider a high degree of transparency and understandable data that consumers use. I’m glad Delta Airlines has an on-time record, if they told me exactly why with all the gobbledygook in there that did this, it would obviscate(?) the issues, is that kind of what you’re talking about, providing this kind of real time?
DR. EDINGER: Actually all we really have is something like, we don’t actually have anything on the CAP, if you had one you’d have that, maybe something like the Washington checkbook, the greatest doctors in Washington, D.C. voted by how many of their colleagues who voted them, but you really don’t have any real information, where you go to your neighbor, did you use Dr. so and so, what do you think of him, that’s --
DR. YOUNG: We did that in nursing homes, we put it in the newspaper, we put full page ads in the newspaper saying, and it was really surprising the providers got behind that. They wanted that information because, I don’t know whether the physician community is going to embrace this, but the nursing home community wanted to compete, and wanted to compete on outcomes.
MR. HUNGATE: The nursing profession is much more in that model and nursing homes I think are more often managed by nursing than by physicians, so it follows. Physicians are a tough knot and the direction that I think it’s important for us to go.
DR. YOUNG: Just to understand what you’re talking about is giving transparency to outcomes at the provider level --
MR. HUNGATE: I’m saying at the patient level --
DR. YOUNG: At the patient level.
MR. HUNGATE: I’m saying that if I need a coronary artery bypass graft, and I have certain characteristics, I ought to be able to get my individually figured perspective outcome, because statistically it could be done. Now I don’t necessarily need it for my precise surgeon, I need it more for my precise characteristics than I necessarily do for the providing system. If within the providing system there is feedback back and forth so there’s tension on who’s where on the spectrum --
DR. YOUNG: To a degree this is a good place for government to be, in that we will provide these levels of transparency, where maybe others would not be, we can break down those barriers, and that’s a good thing for us to do.
MR. HUNGATE: I think HHS has that opportunity before anybody else.
DR. YOUNG: The private side is not going to necessarily integrate those kinds of dialogues and those kind of reportings voluntarily and that government can come in and at least take the first step in that --
MR. HUNGATE: Government can take the patient side of this issue and I don’t know anybody else that can. Now that’s my argument.
DR. STEINWACHS: Two parts to what I think you’re saying, Bob, one has to do with what expected outcome is for a patient given their characteristics. I think for many things we still have quite a variance around that and so I guess one of the questions is is how do you help someone understand that expected outcome may be far away from the outcome that the individual experience, and that’s always problematic because some of us go into these things very highly risk averse so we’re not interested in expected, we’re interested in where that tail of that distribution is, that says could I be blind after cataract surgery.
The second part of it, I very strongly do agree with and that is that there are providers who do it better and do it worse, I don’t care what it is, and so we need to try and force the system, either through consumer choice, through payers, or through embarrassment of providers to move them in the right direction. And so I think that kind of public accountability is important.
The third piece which I think Kathy was driving at is very problematic it seems to me and that is that I go to one physician at the beginning of an episode, I go to someone else, I go to someone else, at the end of this episode of treatment you can line up and say there may have been 20 physicians who have had their hands on me, and I’ve got this outcome. And maybe the outcome isn’t too good and maybe there’s some reasons that we could point to and say some crummy things happened along that way. I think what we lack is a conceptual framework assessing who’s responsible, unless you say ah ha, the anesthesiologist overdosed him, or the surgeon cut the wrong eyeball, but when you get beyond that and you say what are the lapses in this continuity and coordination. One of the nice things, just to finish the thought, is of the 20 conditions, Kathy brought them up that the IOM report has, one of them is coordination. Well who’s responsible? The American system is mainly you and me, in my family and your family, who are the sponsor of that, and that’s not good, it’s like saying you have to build your own car and here are the instruction sets.
MR. HUNGATE: So I would pick up on that and say that CMS started to do some demonstration projects of episode of illness purchasing, where they took care of the coordination issues, they designated a point where the purchase was made. And somebody was responsible.
DR. YOUNG: At least someone was following the money, I mean managed care we thought was going to be responsible, they turned out to be funnel in some cases for just handing out the money.
MR. HUNGATE: But that I think has fallen by the wayside, it did not continue, so I think it was objected to by the silence.
DR. STEINWACHS: But I do like, and I think we’re agreeing, is that when you look at the handoff issues, I mean a lot of the safety issues in America in health care, a lot of the quality problems are in the hand offs, and it seems to me what we do in this framework and in the data has to highlight the fact that there is, with the exception of hand off problems we can identify we may be sure to hang this accountable, I think the system is not --
MR. HUNGATE: We may not be able to assign it but we ought to say this is the episode within which quality was or was not. We’re on the same wavelength.
DR. EDINGER: [Comment off microphone.]
DR. STEINWACHS: But that’s because you’re an informed consumer. You think you know what you’re doing.
MR. HUNGATE: That didn’t quite finish your answer I’m afraid, I don’t know that I gave you a full answer to your question.
MS. KANAAN: I forget what the question was.
MR. HUNGATE: It was about the involvement of the full Committee.
MS. KANAAN: That was really just throwing it in the stew, I didn’t need an answer.
MR. HUNGATE: Well, I think that it’s important and it bears more thinking.
MS. KANAAN: I think the part that we do need to pin down is what kind of background materials you want produced in the next six weeks.
MR. HUNGATE: Now the other observation I would make on the Committee process historically is there was way too little dialogue, and I don’t know how to correct that for sure, but I have the feeling that we ought to regard each other as people to communicate with on things around this without waiting for a meeting if that’s appropriate, that I think we ought to try to do some discussion back and forth through emails and try to get our thinking improved in that methodology, and then have periodic meetings of this sort, but not have big gaps in between.
MS. GREENBERG: Do you want to have a list serve? I mean we haven’t done that for the Committee but we’ve talked about it, we could set up a list serve through the Workgroup.
MS. KANAAN: Could I make another suggestion? When I send out drafts of the background materials to you that would be then provided to the Committee for this discussion we’re thinking of having in June, what about if each of you tried to identify specific questions that you want to pose to the Committee so that this isn’t just a report, if you really want to elicit something back from them, maybe as you read these materials, besides editing them, you could be suggesting so we could come up with a list of discussion questions.
MS. GREENBERG: When you’re talking about June, are you talking about the full Committee meeting?
MS. KANAAN: We postponed having a report of the recommendations until September, so this is going to be an interim report.
MR. HUNGATE: Ok, thank you all, a good meeting, I appreciate it.
[Whereupon, at 4:20 p.m., the meeting was adjourned.]