Sheraton City Center Hotel
1143 New
Hampshire Avenue, N.W.
Washington, DC
Barbara Starfield, M.D.
Simon P. Cohn, M.D.
Kathryn L. Coltin
Kathleen
A. Frawley, J.D.
Lisa I. Iezzoni, M.D.
John R. Lumpkin, M.D.
Vincent
Mor, Ph.D.
Clement J. McDonald, M.D.
George H. Van Amburg
Robert
Moore
Marjorie S. Greenberg
Robert M. Gellman, J.D.
Richard K.
Harding, M.D.
M. Elizabeth Ward
Introductions
Update from the HHS Data Council: David Garrison
Presentation from
Panel on Public Health Performance Measurement and Data: Jeff Koshell
Selection
of Future Meeting Dates and Agenda Topics
Reports on Subcommittee Work Plans
Recommendations on Health Data Security Standards
Recommendations on
National Payor Identification
Discussion of Unique Identifiers and Data
Content
DR. LUMPKIN: Although we did introductions yesterday, I would just like to have the committee introduce themselves, so people on the Internet will know who are associated with the voices, and then ask the individuals who may be coming to the microphone in the audience, when you do speak, to introduce yourself so that folks on the Internet will know who you are and where you're from.
Why don't we start off with our distinguished colleague from the state of California, without any public transportation anymore, with BART on strike? I was watching the news, and the rush hour had already started. That was at 6 o'clock this morning, our time.
DR. COHN: It is hard to make any great comment, other than that I feel badly. Anyway, my name is Simon Cohn. I'm with Kaiser Permanente Medical Care Program. Member of the committee.
DR. GELMAN: I'm Bob Gelman. I'm a privacy and information policy consultant in Washington.
DR. WARD: Elizabeth Ward, Washington State Department of Health.
DR. AMARO: Hortensia Amaro, professor of social and behavioral sciences at Boston University School of Public Health.
DR. SCHWARTZ: Harvey Schwartz from the Agency for Health Care Policy and Research.
DR. STARFIELD: Barbara Starfield from the Johns Hopkins University.
DR. GREENBERG: Marjorie Greenberg from the National Center for Health Statistics, CDC, and the executive secretary to the committee.
DR. LUMPKIN: John Lumpkin, director of the Illinois Department of Public Health.
DR. SCANLON: Jim Scanlon, U.S. Department of Health and Human Services and executive director of the committee.
DR. IEZZONI: Lisa Iezzoni, Beth Israel Deaconess Medical Center in Boston.
DR. COLTIN: Kathy Coltin, Harvard Pilgrim Health Care in Boston.
DR. VAN AMBURG: George van Amburg, Michigan Public Health Institute.
DR. HARDING: Richard Harding, child psychologist from South Carolina.
DR. FRAWLEY: Kathleen Frawley, the American Health Information Management Association.
DR. LUMPKIN: Great. We are going to continue with some of the updates we began yesterday from the Data Council, starting off with David Garrison.
DR. GARRISON: Thanks very much, John, members of the committee. I appreciate your letting us appear before you today as part of our process of communicating between your activities and our activities at the Data Council at HHS.
I want to begin by expressing special appreciation to all of you for the work you have been doing. This has been an exceptionally busy year, as you know, particularly to the run-up to some of the recommendations that John Fanning is responsible for, and who will talk about it in just a minute. We have been very attentive to the results of your work, your recommendations, and to incorporating as many of them as we could into the work that we have been doing. So we feel, and the Secretary feels especially good about this relationship, and I just want you to know how strongly she and I feel about the quality of our exchange.
I wanted to bring you up to date on a couple of activities since my last report. As you know, I'm the co-chair of the committee with Bruce Flagg from HCFA. Bruce is finishing up his service with us this week, unhappily -- I think he may be happy, maybe not; we're not happy to have him go, but anyway. There will be a new co-chair identified shortly to join me in this responsibility.
We are also, in addition to the normal kinds of things we do, the focal point for the administrative simplification requirements of HPPA, and that is actually one of our top priorities, and we are spending a lot of time on that.
You heard yesterday from Bill Braithwaite on our progress in dealing with those provisions of HPPA. As I mentioned, John will say a few things in a minute about the status of our work on the privacy recommendations.
I should say in preface of John's comments, the Secretary has appeared before the National Press Club and made some comments about her thinking on the recommendations she is about to make. She is scheduled to testify in two days before the Senate Labor Human Services Committee. That is the point at which she will release the recommendations. So you may find John a little less full in his description of the subject matter, being that we are two days in advance of the Secretary's appearance. You will have to excuse him if he is not as effusive as he normally is.
Now, I want to report on a couple of particular areas of activity. The first has to do with our data collection strategy work. First, as part of the strategy, the Data Council has initiated a department-wide advance plan for surveys and major data collections. A plan of all surveys planned and underway over the period from '96 to 2001 is being schooled, along with multi-year budget planning information.
In addition, during July the Council initiated a cross- cutting inter-agency review of data proposals and initiatives and agency budget proposals for fiscal 1999. This was a process which resulted in a presentation internally by the Council as part of our budget process, and it was an exceptionally helpful intervention.
Second, we also have a working group to carry out the survey integration plan and related activities under the auspices of the Data Council. You have been briefed on that plan.
We are making good progress on implementation. The plan is intended to bring all major health surveys within HHS into a coordinated strategy and integrated framework. It also addresses major data gaps identified during health reform and it increases the capacity of major surveys.
The major elements of the plan include a redesign and automation of the national health interview survey to serve as the sampling nucleus for many HHS population surveys. Secondly, the implementation of an ongoing longitudinal panel survey on insurance and expenditures, the medical expenditures panel survey, which we call MEPS, was also used in the national health interview survey as a sampling frame.
We have also begun the process of consolidating our survey of employers as well, so that for example, the national employer health information survey and the insurance follow-up component to the medical expenditures survey will be jointly fielded. We plan to implement joint field operations in common core questionnaires for HHS in the major surveys for health providers. The national health and nutrition examination survey is also being planned for 1998, and it too will be designed to use the NHIS as its sampling frame.
We are also rethinking of developing a conceptual framework for data on health care resources, capacity and the public health infrastructure, including HHS provider inventories and surveys in analytical databases. We would welcome this committee's advice in these rapidly changing areas.
Third, the HHS research planning initiative was begun last year, and is aimed at assuring that HHS has the capability to monitor the impacts of market and structural changes that are already underway, and affecting the organization, financing and delivery of health and human services in the U.S.
There are a number of major data issues in that area. During the first stage, we asked all our agencies to look at their capabilities and plans in monitoring the impacts of transformations. We sponsored a workshop at the National Academy of Sciences last December. We are now turning our attention to the second phase of the initiative, and we will examine ways to address some of the issues in monitoring the impacts of transformations. The Data Council is coordinating the activity for the development of the plan.
Now, fourth, the Data Council has also begun looking at various strategies for obtaining state level data, both as part of the survey integration plan and more generally. This is an area where there are a number of gaps and issues, and we are taking a broad look. It is an interest of specific and special interest to the Secretary.
We are looking at the possibility of developing modular designs for our national surveys that will facilitate state level estimates and provide a mechanism for states to buy into national survey activities to meet their own needs. We are also planning to augment some of our national surveys to provide data on states, especially in the substance abuse area, and work with several states and NCHS to explore the utility of a state telephone survey capability, in which the telephone survey can be combined with data from other sources to improve state level data.
In that regard, we are developing survey questionnaire modules related to health and welfare issues that states might be interested in using. Feasibility tests are planned in five states and in the local metro area this year. By choosing various questionnaire modules, surveys can be tailored to meet specific federal or state needs.
With the National Academy of Sciences and others, we are also supporting work in developing public health performance measures for use at national and state levels. You will hear about that later this morning from Jeff.
Fifth, we have also established a standing HHS working group on race and ethnicity data in the Data Council to review previous recommendations in that area, identify issues and make recommendations for department-wide approaches and strategies. The Council serves as the HHS focal point for involvement with OMB on these examinations and potential revisions of OMB Directive 15. The Council is proposing to the Secretary an HHS- wide policy on the inclusion of race and ethnicity information in HHS-sponsored data collections. This is still in internal clearance.
Now, the second topic I want to mention just briefly is that the Council provides a forum for coordination of issues and opportunities in health and human services related to the expanding national information infrastructure.
The Council has begun examining opportunities for health sector applications in the national information infrastructure, including telemedicine, enhanced consumer health information and improved methods of disseminating information electronically.
The Data Council is overseeing HHS activities relating to the directive from the Vice President to HHS to lead inter- agency efforts in four areas: health data standards, health information privacy, telemedicine and enhanced health information for consumers. We have established inter-agency working groups in these four areas, and we have recently sent a stage report to the Vice President on our progress.
There are a number of exciting developments, as you know, in the telemedicine area and in the computer based consumer health information area. We have developed for example Healthfinder, a new collaborative website that provides access to many consumer health information resources of HHS and other federal agencies in a user friendly manner.
These areas along with the new HHS responsibilities under HPPA, represent our data policy priorities at HHS, and as I mentioned at the beginning, we appreciate your work, and we continue to invite your advice and counsel on all of these issues.
I look forward to working with you in the future as always. I'd be pleased to answer any questions, and also am prepared to pass the baton to my colleague at my left, John Fanning, to talk to you about the privacy recommendations.
Thank you, Mr. Chairman.
DR. LUMPKIN: Okay. Any questions?
DR. STARFIELD: Could we have a copy of your remarks? They were very helpful.
DR. GARRISON: Sure, pleased to do it. Thank you, Dr. Starfield.
DR. LUMPKIN: Any other questions? I just have a brief comment in relationship to the state level data. I think one of the best models that has existed for state level data is the behavioral risk factor survey, which we in Illinois are utilizing the same model of purchasing, by using the same methodology to try to develop county level data, which is important for our planning.
One of my concerns has been that funding for that particular survey is down this year, and potentially will continue to go down. I think we at the states have concerns about the level of funding particularly in these types of joint ventures. I just wanted to raise that concern.
DR. GARRISON: Well, you can rest assured that the Data Council has been an advocate for full and adequate funding, let me put it that way, for a wide range of activities in the department. I think that one would say that in addition to trying to rationalize the activities of our various operating divisions across the department, the Council's activities of looking at data work across all of our functions has helped to serve in itself as a form of advocacy for our data work. So we are doing the best we can.
DR. LUMPKIN: Thank you. John?
DR. FANNING: Let me just speak very briefly. The department will unveil its recommendations to the Congress for health record confidentiality legislation on Thursday. We are especially thankful that this committee has been helpful in holding hearings and in making recommendations. The hearings and the easy availability of the transcript has been a great help in sorting out what various people want and what their needs and interests are. So we are very thankful for that.
I won't go into any detail about the content. I have discussed the outline of it before. The general idea is that we will cover providers and payors as the basic organizations to be covered by the rules. And there will be some rules that will follow disclosures that are allowed within the statement of the law.
But that leads me to up to the next point, which is that there is a great deal more to be done. There is a lot of health information around, and only some of it is held by providers and payors. I think we are now positioned with this committee, with the subcommittee tending to this issue, to be more expansive and more attentive to the other needs of the situation.
One subject for example that will need attention will be the protection of research information. I think our statement on Thursday will probably say we're going to look into that. I know NIH is beginning to look into that, and we will welcome the inquiries of this committee that would help feed into that discussion.
So again, thank you to the committee, and watch for the unveiling of our proposal.
DR. LUMPKIN: Will we be perhaps getting a copy of the Secretary's speech once she has given it?
DR. FANNING: The plan is to put up the text of the recommendations on the website. The Secretary will have the button next to her at the hearing, and press it at the beginning of her statement, and then the whole world will be able to call it up.
DR. LUMPKIN: Thank you very much. Any questions that he can answer? Well, thank you. I think we have certainly had a good working relationship during this period of time. We appreciate your assistance with our deliberations, and we are looking forward eagerly to the report on Thursday. Thank you very much for coming.
We will now move to a report of the National Research Council panel on public health performance, measurement and data.
DR. KOSHEL: I've got a few overheads which hopefully will summarize some of the findings of the committee.
I hope everyone has received a copy of our phase one report. This is a report of a committee whose formal name is Panel on Performance Measures on Data for Public Health, the performance partnership grants. Some of you may remember that a couple of years back, there was active consideration of a new form of block grant for some public health programs that CDC was funding and SAMHSA still is funding. That would be in exchange for giving states a little more flexibility in spending money require that they agree to performance contracts in effect with the Department of Health and Human Services.
So the idea behind all of this was that it would be useful to have an independent panel look at some of the possible measures that could be used to construct these performance agreements between state health departments. That was the origins of this particular panel.
You will see on the next slide, the members of the panel include Dr. Lumpkin, who is a very active member, as well as Ed Periman and John Baylor and a number of other very well- known and distinguished panel members.
The trick here of course from the NRC standpoint was to get not only people who are knowledgeable in the technical issues having to do with data and performance measurement, but also people who have experience in administering programs at the state level. In one case, we have a very distinguished member of the local health department, Paul Wiesner, who was also as some of you may remember at CDC for a number of years. So this is a fairly large panel by NAS standards, and it is also a somewhat more diverse panel that you have in a lot of IOM studies, for example.
Ed Perin when he does a little presentation on this panel often talks about the Dunn and Beatty model that was used as the framework by the panel in examining performance measures. Unlike the original conception, which I think many on the Congressional staff had and some within the department, where the idea was that outcome measures would be the basis of performance agreements, the panel fairly quickly concluded that outcome measures by themselves would really be inadequate for judging performance of individual state entities. Following the classical discussion about things that are needed to affect outcomes, the panel spent a fair amount of time looking at those things that support public health systems and capacity measures and process measures that states engage in to affect outcomes.
Even something as simple as these basic definitions and concepts took us awhile to get through, given the policy climate, where everyone was looking for outcomes and bottom line results. So we went through a fairly elaborate discussion, the panel did, that is, of administration indicators, capacity indicators, process indicators and outcome indicators, and tried to talk about the relationship between them, and the distinctions, subtle though they may be, between some process measures that we use as proxies for outcomes, as well as the outcomes themselves.
I take it there is nothing new in any of this for members of this particular committee, so why don't we just go forward. It did take us a long time to get through some of these things with Hill staff and others, so I won't bore you with that.
Again, with Hill staff and with some in the executive branch, we did have to talk about the connection between processes and outcomes, and how one has to make assumptions, and when one should relax those assumptions, and when one has to be very careful about drawing inferences.
Again, there is nothing here for this particular committee, I don't think, other than the fact to make the point over and over again in the report that when we are not sure about the specific effect of an individual process, we have to be very careful in drawing inferences. Some of these are the reasons of course why -- the reasons that are listed there of course are very basic reasons as to why one has to be very careful in drawing such inferences.
So let's go to conclusions of this report, some of which surprised a number of people who are looking at the work of this panel.
Number one -- and I wish there were more -- I have been asked to do a number of these discussions for Hill staff, and I wish some of them were here, because you do have to keep repeating these things. People seem to forget them pretty quickly.
Namely, one has to proceed with great caution in using outcomes in performance measurement. While it is important to use them, as members of our panel say, to keep your eye on the ultimate objective of what you are trying to achieve, you do have to in the panel's judgment continue to use those things which have traditionally been used by public agencies to demonstrate performance, namely, process and capacity measures, as adjuncts to those outcome measures.
So the panel has concluded, I suppose the single most important conclusion perhaps, that one should not look for quick or summary outcomes alone. The more measures that one has for getting a sense of what performance is being achieved, is to use as many measures that seem appropriate as possible, outcomes, process and capacity. The panel recommends that they be used in conjunction with one another as a series of measures.
Even though this report -- I guess it is an important point to make; this report does list 50 or 60 outcome measures, because we thought we had a fairly good handle on those outcome measures that could be supported by data at the current time, whether it was BRFSS or another source of information or some state data systems. The panel believed it had a pretty good idea of what within the first two or three years of performance agreements was feasible to list in terms of a fairly tight universe of outcome measures that could be supported at the current time with state level data.
But we tried to say just as quickly that the process and capacity measures that are listed here -- and we tried to list several in each of the substantive areas to illustrate what process and capacity measures could be considered. But we tried to make the point, those were only examples. There was no attempt to come up with a complete universe of relevant processing capacity measures.
I hope on the next overhead you will see an illustration we use in the report to indicate why that is the case. One of our panel members did a very nice job, I think, in identifying examples of process measures and capacity measures that one would think would be pretty self evident as being reasonable to choose in trying to reduce the incidence of smoking.
So we went through quickly in the report at least a half dozen of these program strategies, all of which have wide support among those who are knowledgeable in trying to address this question of reducing the incidence of smoking. I won't go over all of them, but there are six or seven that we list quickly in the report, just to give the reader a feel for why it is hard to be terribly proscriptive in coming up with a universe of process and capacity measures.
In some ways, it is even foolish, because what you do is, you put almost a tax on creativity. Until the definitive empirical evidence is in, that one process measure or one approach is better than another by a specific amount, all you are doing by recommending several out of a multitude is skewing the appropriate response that some people may have at the state and local levels to addressing an important problem.
The next slide simply illustrates a few more of these, I think self evident approaches to trying to do something about reducing the incidence of smoking and the health problems that result from the incidence of smoking.
I don't know whether readers really appreciate the importance of the panel's recommendation in this area. I hope so. We tried in the summary and in the introduction and the conclusion to underscore that point. But I think there is tremendous interest among people who vote on appropriations to have a magic bullet or a single outcome measure or even a handful of outcome measures. Okay, if we know that, then we know what we need to know, and let's get on with it. So we did our best to suggest it wasn't quite as simple as that.
So let's go over a few more of the conclusions and then you can ask Dr. Lumpkin some questions if you have any.
Of course, one of the conclusions of the panel that I think is very relevant of course to this committee is that there is really an urgent need to invest in data infrastructure to support performance assessments and program evaluations.
We tried in every instance where we listed a health outcome to identify the apparent relevant sources of data and to suggest what some of the limitations of those data are, and some of the limitations of the measure itself. If you noticed, in the back of the report, for each of the outcome measures the panel suggests, there is a complete page discussion which includes some mention of the data sources and the limitations and the sources and so on.
That was actually a suggestion made by a number of people who are in the policy community, who suggested that once you list something, you really run the risk of having someone just tear the page out and run with it and say, put this in the next piece of legislation. So we tried on each page, just hoping they would tear out the whole page, to at least list those other concerns that one might have in using that particular measure or that data source.
But clearly, the need for better data -- I'm sure you all know the limitations of our current data systems. We don't have a lot of -- John mentioned the BFRSS. Of course, you will see in this report, that is a principal data source for most of the measures, many of the measures, and clearly, if it is going to be eroded further, then there is a real question as to what you can gather from health outcomes.
The other thing that of course people who are interested in outcomes fail to appreciate is the -- and I'm sure people on this committee understand fully, is a dearth of research evidence about the relationship between program interventions and health outcomes, for a lot of good reasons, technical reasons, resource reasons. We simply don't have a real solid understanding of, if we do this, then something else occurs, and precisely or even approximately what effect one has on the other. So we try in the report that emphasize what is probably obvious to any researcher in the field, that one has to pay attention to these relationships, and one should in fact try to promote more research on the relationship between program interventions and outcomes.
I must say, that is particularly true for those of you who aren't as familiar with it in substance abuse and mental health.
The panel had a very large challenge, I guess is the way you would put it, to come up with measures that could be supported by research that showed that if you strike a particular intervention, you get a particular outcome. Then further, to measure what those outcomes are across states. It is very difficult to get a data source that is used by more than a handful of states for any one of the measures in our report in substance abuse and in mental health. In that sense, the public health community is much further ahead.
I suppose one standard the panel could have used is that a majority of states should be able to collect data on a particular measure, and for that measure to be incorporated in the report. Had that been the case, we probably would not have had a single measure in substance abuse or mental health.
So the rule of thumb was, if any state had a source of data, valid and reliable data on a particular measure, presumably the department would be interested in allowing that state to use that measure as part of its performance agreement for a given year or series of years, because they had a way of verifying what they were accomplishing. So that became a very minimal test, of course, of whether or not an outcome measure could be suggested in substance abuse or mental health for that reason.
Now, this is phase one. This is a phase one report of a two-phased project. We are currently in the second phase. The first phase was to identify those measures that could be part of performance agreements between states and the department in the next three to five years. So what is currently available in the way of empirical information about the relationship between processes and outcomes. That is the panel's best attempt to summarize that in this volume here.
The next volume is supposed to look at what data are needed to improve these measures and their limitations. Again, we tried to list some of the major limitations of both the measures themselves conceptually and the data sources that support those measures.
The next phase of the project is to identify what improvements are needed primarily in the data side, but also on the research side, that would allow the department and the states to have more accurate measures or more comprehensive measures down the road.
So that is the phase that John and the other members of the panel are actively involved in now. According to our current schedule, we are to produce the second report or the final report by June of next year.
John will answer any questions you have.
DR. GELMAN: Is it fair to conclude that this whole activity was driven by GIPRA? Or was it independent?
DR. KOSHEL: No, it is fair to conclude that GIPRA had a major effect on all of this, of course. But of course, in this particular instance, it was another piece of legislation that was driving a particular set of general block grant appropriations. That was a bill actually reported out of the Senate Human Resources Committee, that would have changed the nature of the SAMHSA block grants to performance grants.
But it is an interesting development. That particular piece of legislation went nowhere, other than getting reported out of committee on the Senate side. But now, all of the executive branch agencies are looking at coming up with strategic plans that are supposed to have outcome measures and other relevant measures as part of their submissions to the Congress, by next year, I guess.
So the members of the panel, by the way, -- another conclusion the panel reached, and those of you that have had a chance to look at the report will recognize it -- the panel said very strongly that for a lot of reasons, using performance measures to allocate resources was not a particularly smart idea. You can use performance agreements as crude flags, perhaps, to give you some sense of whether a state is proceeding in the right direction, in much the same way that CDC uses its weekly disease, morality and morbidity statistics to look at possible outbreak of diseases and then go investigate.
So the panel basically concluded the same thing, used these things perhaps to signal the need for technical assistance. In that sense, the panel I think concluded that quite apart from any legislation, this framework, the structure used here made sense for GIPRA or any attempt, really, to manage programs as a signal for technical assistance.
The federal government is moving forward. Even in the absence of legislation, HHS is actively working with not only SAMHSA, but NCDC, but with the maternal and child health program and really, all of the programs now because of the GIPRA requirements. So what was I think originally conceived of as fairly narrow in a sense, a narrow rationale for looking at this, I think is really much broader now, and probably doesn't need -- I'm not a legal authority, I'm not sure it needs legislative authority to proceed forward in this area.
DR. GELMAN: Let me ask you a different question. Got a performance measure for confidentiality compliance?
DR. KOSHEL: We don't.
DR. GELMAN: Have you thought about it?
DR. KOSHEL: No.
DR. GELMAN: Do you think it is a good idea, worth looking at on substance abuse and mental health? There are two areas where it is a particular concern.
DR. KOSHEL: John can talk to this better than I, but I think in the second phase of the report, the panel will make recommendations with regard to data systems. I would be happy to bring for their consideration any and all suggestions and issues that you raise.
DR. GELMAN: Consider it a suggestion. It is an area that no one has really spent any time looking at, are people complying with whatever rules are out there. It will be interesting to see if there is a way to do that.
DR. LUMPKIN: I think in response to that, one of the biggest challenges that we had on the committee was that of any sort of system -- and you see it most frequently when you develop information systems.
That is called scope creep. You get a job to do this much, and then people start throwing things on, and if you don't stay focused, you won't get the original job done or anything else.
So we had a fairly narrow focus, which was to look at a certain set of block grants that were being at that particular time looked at for the performance partnership grants. So in that regard, I think the importance of the committee's work is not that we looked at everything in public health, because we didn't, or we looked at everything in health, but in fact we looked at some models that could then be expanded by doing a similar type process to other components of federal programs.
But I think there was a limitation that also came into the system, in that we were looking at federal programs. Many of the programs that were implemented -- mental health is a good example -- less than five percent of the funding of mental health services in states come from federal sources. So you are looking at trying to measure performance of this little bit of money coming into this huge system.
Certainly, some of these measures are appropriate for a state to look at itself and how it expends its own dollars. But there are other things that a state that is performing its activities has to look at. That is, their accountability is not to the federal government, but back to the taxpayers and residents and citizens of their state.
So many models of measurement and performance may be different when you look at how I as a state health director would evaluate my performance versus how I would report on my performance to a funding source, and a major funding source, which is the federal government. That were some of the things we had to struggle with, which is why we felt it important to de-emphasize some of the outcome measures.
Clearly as a state health official, I am accountable to the people of Illinois for incidents of disease and all these sort of things, which are true outcome measures. But that doesn't mean that the federal component necessarily plays a role in how I do my job.
DR. IEZZONI: John, on this point though -- Barbara, sorry to interrupt -- our subcommittee on population specific issues has a point in its work plan that we are concerned with exactly what you just raised. As people begin to look at mental health measures and substance abuse measures, that the confidentiality issues become even more soaring than they are for other aspects of the health care delivery system.
I don't think that our subcommittee would necessarily advocate scope creep, John, but I think that it is important to at least raise the issue that this might be an area where data systems need to be particularly sensitive to confidentiality issues.
DR. STARFIELD: I think this is an enormously exciting report. I have already used it a lot. It has tremendous relevance to the work of this committee. It brings us I think back to the larger scope of our mandate. We have been very heavily involved in the last year in relative technicalities in response to HPPA, but we do have a larger mandate that carries us into the future, and I hope we can work with you.
Now, I think you are all not going to be able to avoid scope creep, because you can't really look at things as public health versus non-public health. We know that as we move on into new health care systems, that there is now increasing overlap between public health and clinical medicine. When you talk about capacity, process and outcome, and you are trying to monitor the impact of what you do as a public health agency, you are very dependent upon what the private sector does, particularly when that private sector is increasingly defining populations and taking on responsibility for populations.
So the work you are doing is really going to have to overlap with the work we are doing in the larger data needs. I don't think we can wait until June to hear your recommendations. I think we need to develop some kind of mechanism for us to share information. You are going to have to rely on data in the private sector to do what you've been doing in performance monitoring.
DR. KOSHEL: Yes. One of the things that the panel has been very concerned about throughout all of its discussions thus far is the changes brought by managed care, in terms of gathering data and making data available to the public sector, and whether or not the needs of managed care organizations would be similar to those of public agencies.
DR. AMARO: I also wanted to say thank you. It was I think a terrific piece of work that the committee is doing. I have also begun to use it.
I was wondering if there was going to be any initiative to support states to develop the data systems. I know in my state and other states, people might be very interested, but usually dollars are short. So is there going to be any initiative from the federal level to support the development of these data systems?
DR. KOSHEL: I can let Jim speak to that. But I do know for a fact that SAMHSA has requested -- I believe it was $100 million or $50 million?
DR. SCANLON: A hundred million.
DR. KOSHEL: A hundred million dollars for just such support in the substance abuse and mental health fields.
DR. AMARO: Requested it? So we don't know if --
DR. SCANLON: And then they bring a 1999 budget request. SAMHSA has requested about $100 million, 50 for mental health and 50 for substance abuse for state infrastructure development for data. But again, this depends on what Congress will ultimately do. And remember, other agencies are also looking at how they can beef up and increase the infrastructure on the availability of state data.
DR. KOSHEL: Now, of course, one of the arguments that I always remind my friends in the state substance abuse and mental health field is, it is all well and good to have some money for data development. But it wouldn't be a bad idea to have some common definitions across states.
I'm not sure that they are in the strongest position right at the moment to make a solid claim for how they are going to use $100 million effectively. If they were here, I would be willing to say that even if they were here.
I do think they have got a real challenge to get together, and mental health seems to be further along, for whatever reasons. They seem to have accepted the inevitability of being required to come up with better measures of what they are accomplishing. Substance abuse is having a harder time with that. They do argue -- some will argue in that world that you do not need to compare states, it is only within a state, and so on and so forth.
I was an NGA for a few years in the early '80s. If I have heard that argument once, I have heard it a hundred times. But it is becoming I think increasingly clear to most state administrators and local administrators that when you go before the city council or you go before the state legislation, they are going to ask pretty much the same kinds of questions. Demonstrate to us that you need the money we gave you last year, that you did well with that money.
So I think if the state agencies -- in not only public health, but substance abuse and mental health -- really meet the challenge by demonstrating some serious efforts to come to grips with improving data systems and concepts, as well as the existing information system, I think there is a possibility that some significant resources can be made available to them to improve things.
DR. IEZZONI: One really tiny question and then a slightly bigger question. Are you going to be looking at all at data systems in the territories?
DR. KOSHEL: I have met with representatives of territories, in the health agencies there. We have looked at what is available now. And of course, as you probably know, there isn't very much.
DR. IEZZONI: Right, right.
DR. KOSHEL: You even get to the difficult position of considering how qualitative information can be used for this purpose. It is not only in territories, as you know; it is also Indian tribes.
That is a real challenge. We have a hard enough time getting standard definitions of quantitative variables across states, and then it becomes even harder to think about how you get qualitative assessments across tribes or territories.
But they are actively engaged. To the department's credit, I think they have not been excluded. For those of you who may know about it, the process that was set up for this consisted of four regional meetings, where 400 or 500 people came, and people from the territories did come. They were invited to come. The department paid their expenses to come and to attend.
I think they recognize the need to move forward, in the same way that people in the substance abuse field recognize that this day is coming, if it isn't already here. But it is still a challenge, and that is a real challenge.
DR. IEZZONI: Well, I mentioned it because our subcommittee will actually be holding a day of hearings on data systems for the territories for public health in the second week in December. It might be good to get some input from you, or begin to interact around that particular issue.
DR. KOSHEL: Will that be here in Washington?
DR. IEZZONI: Yes, yes, it will be. The second question is, our subcommittee on population specific issues has taken on Medicaid managed care as our major focus over the next year. It must be a little frustrating for you to think about doing a report that is due next June, when the data systems and the programs that you are reporting on are changing at a lightening speed.
DR. KOSHEL: Yes.
DR. IEZZONI: So how are you dealing with this issue of the tremendous changes that are happening in Medicaid programs, which are very important for thinking about mental health and substance abuse treatment and so on?
DR. KOSHEL: Well, to the department's credit, they have graciously responded to two separate budget supplements and requests that we have made to them to continue this project on a longer time period than was originally conceived.
Having said that, in a somewhat more perfect world, I suppose, this project should continue, probably for as long as this committee continues to function. If you are looking at long term --
DR. IEZZONI: Forever.
DR. KOSHEL: Forever, right. Long term data needs are long term data needs. The field is moving very quickly.
The department really has been -- I think the word generous is appropriate here, in terms of supporting this particular panel. We are scheduled, as I said, to complete our work by the summer of next year. But clearly, some other entity, some group, I'm not sure what the vehicle will be, will have to track for the next several years at least to identify that.
DR. IEZZONI: Well, I am also concerned about us duplicating what you are doing, because we are going to be over the next year having a series of hearings and meetings and site visits around this issue. We are going to be meeting the last week in September to try to really formulate our work plan. It would be a shame if we duplicated work that the department is already funding.
DR. KOSHEL: Well, I can assure you that if there is any material available to the panel, we will gladly cite it, give you full credit for it, and do whatever we can.
DR. IEZZONI: Oh, and likewise, likewise. There is no problem with that.
DR. SCANLON: You might describe the process through which the remainder of this study is envisioned, the panel meetings and other --
DR. KOSHEL: Well, those of you who serve on IOM committees know all about it, but of course, we have two scheduled meetings. If I can figure out a way of freeing up a couple of dollars, I will try to get a third. Joan hasn't heard this before and Jim hasn't heard this before, but I will try to find a way of getting a third panel meeting squeezed into our remaining time period. We have one scheduled for next month, October 21 and 22. We have one in late January; we don't have the dates scheduled. I will try to squeeze in another one sometime probably in the May time period.
We currently have panel members actively drafting sections of the final report. As all of you know, that is a long process for getting anything through the Academy, and all the reviews that are required as well. Unfortunately, in the current time frame, we probably have to have something in fairly good shape, I would say, by February, March for release, because it takes all the rest of the period for reviews and comments and so on.
DR. STARFIELD: Is there a possibility of a joint meeting between the subcommittee on data needs and the population specific subcommittee, in at least a subset of your committee, sometime in the next three or four months?
DR. KOSHEL: I'd be quite happy to consider it and take it to the chairman and support it, yes, sure. You have one in September and one in --
DR. IEZZONI: We've got a series of them. I can't even keep them straight. I think that maybe we should work through Jim, since he is the project officer with you, and just make sure that when we come up with our work plan the last week in September for this specific project, Jim, that we make sure that we are not overlapping and that we are building on each other's work.
DR. STARFIELD: Is there any possibility that you could brief the committee on the general direction before -- DR. KOSHEL: I'd be happy to.
DR. IEZZONI: And what you might view is the gaps, what you don't have the time to look at, because that is what we are desperately trying to do at the end of September, is find our niche, where we could contribute something that nobody else is.
DR. KOSHEL: Good.
DR. STARFIELD: Yes, as the chair of the subcommittee on data needs, I am keen to include that subcommittee also in there. So we just have to coordinate it all.
DR. KOSHEL: I'd be happy to do whatever I can do to support it.
DR. COLTIN: Building on this notion of the increasing role of managed care in this area, I wondered whether you were intending in your next phase to think about the role and importance of public-private partnerships, not just in the intervention phase, but in the data and measurement phases for a lot of these performance measures.
I was pleased to see that there are a number of performance measures in here that you are advocating in the public health sector, that are quite similar if not identical to measures that are being used in the managed care area to evaluate performance, things like percent of women over 50 who had a mammogram in the last two years.
However, the measurement methodologies are different. In the managed care area, we are using record based measures, whereas here you are using patient report based measures. I recently had an opportunity to look at a report that was put out in Vermont by their quality group, and they actually compared the rates that they got from the BRFSS on some of these measures that are also in HEDIS with the rates that were reported by the health plans for the same populations. It shows that the BRFSS rates are higher for each of those measures than the rates that are being obtained from records data.
I wonder if there were any thoughts about trying to do some validation studies to look at the accuracy of patient reports for some of these measures, by comparing them with record based measures. It seems to me that some of these things could be done in collaboration with managed care organizations.
DR. KOSHEL: The panel has in fact considered some of those very issues. Several of the panel members have begun to draft some discussion of that for inclusion in the report.
But I am concerned that within the time period that we have, that our review of some of those issues may be not in enough depth. Also, when it comes to recommendations in the report, you will notice that in the first report, the recommendations were fairly general, the need to support data infrastructure and so on.
I think it is a real problem for the panel to get terribly specific, other than to illustrate broad points and maybe give one or two examples of why it is necessary to validate different reporting -- data that come from different reports and using different reporting methodologies.
I suspect that where the panel will come out is that it will give examples of some of these issues to try to illustrate the broad concern, but not provide -- unfortunately, it is going to be very difficult for this particular panel to provide a catalog of those endeavors that should be undertaken, I think primarily because of time, and maybe in some ways because of -- John can speak to this better than I.
This particular panel I think has tremendous breadth of competencies, not only academic, researchers and people who run the state programs, but even having said that, there are still major -- we do not have a strong contingent of private sector folks on the panel. So we have tried to get input from folks at our meetings. We have brought people in to provide papers and to provide technical briefings. But it is a very broad subject area that we are being asked to address.
I would be very, very happy to talk with you further. I would like to bring as much as I possibly can to the panel for its deliberations on this and we'll just see how much detail we can get into in the final report.
DR. VAN AMBURG: This follows a little bit on your point. The call for additional data systems has a rather familiar ring to it. But to me, the most important part of the data systems is the methodology development, and then on the back end, the analysis, presentation and distribution of the data. I haven't heard too much about that part of it. John mentioned funding for (word lost) is decreasing. You pointed out that the BRFSS is one of the major sources of a lot of measures in here. There are some methodology issues with the BRFSS that have to be addressed. We get response rates of 55, 60 percent in states on the BRFSS, and going down because the population is being over surveyed.
Is there any recommendation or thoughts about supporting methodology development and the analysis of data for states?
DR. KOSHEL: A quick answer is yes. I think the panel has an understandable hesitancy to get in the middle of very serious differences of opinion that different folks have, and which methodology is preferred.
If one for example takes a look at the data support for BRFSS or financial support for BRFSS, or the youth risk behavior surveillance system, one can get somewhat discouraged, especially when you talk about response rates for one. You look at the number of schools that participate in the youth risk behavior, you could also ask the question whether that is a viable long-term source of information. Therefore, you could start suggesting, why don't we try another approach.
It is pretty tricky stuff, to get in the middle of those kinds of discussions, other than to point out the problems, to point out alternatives that might be available to states, and to suggest what appear at this point in time to be relative advantages and disadvantages of alternative approaches without really saying, this is where the panel concludes that one shouldn't make the investments.
There will be that discussion in the report. How satisfactory it is, though, for people who are trying to make tradeoff decisions and resources allocation decisions, I'm not sure. It is like when you read about people who write creative fiction for a living, and they say, they are in the middle of the novel and they are not quite sure how it is going to work out. You kind of have a sense of that when you're study director of an Academy panel study. You're not quite sure where it will come out. You have seen it work before, you think it will work now, the right elements are there. So in all honesty, that is where I sit.
DR. LUMPKIN: Any other questions?
DR. WARD: Just picking up on Kathleen's, it is actually a question of our speaker. But I think it has brought up the long-term information responsibilities Barbara has talked about in this committee. I think the question you are raising is one that it takes a subcommittee to get through, and it couldn't be given to any other existing group.
It is why I get excited about this tedium of data standards, that we could finally get to the point where we could do performance measurement that is meaningful. It does mean we would not be doing it with all these disparate ways of counting things, and exactly what you are describing. But I think it could be a subcommittee's year's work to get all of these answers about how could we recommend that those partnerships exist, how would we fund it. How can you get over all of the confidentiality issues that the other subcommittee is saying could in fact prevent that to occur.
But I would just like to make sure we don't lose that track. It is something that I think we have to keep track of.
DR. STARFIELD: As you recall, we discussed this in great detail, and the fact that we have some big gaps in the committee, and we hope that a work group of the subcommittee on data needs, and can take that on. Hopefully that will be soon.
DR. SCANLON: Actually, Jeff, you can speak to this as well. The panel is looking at the potential of the administrative simplification standards -- I know you have been briefed on that -- and how it might contribute in the future and the data systems of the future performance measurement more broadly. I don't know how far along you are, but I assume you will have some recommendations in June on that as well.
DR. KOSHEL: The panel has been exposed to most if not all of these issues at some level. In all honesty, my concern is how much analysis of these issues the panel will be able to include in the final report. As you say, it is a real complicated field, and it has lots of -- and the panel tends to be -- there are people who expected this report to be -- well, some expected it to be much thinner. Originally we were supposed to have a letter report, and I argued against a letter report, saying that nobody would understand why you listed 50 measures if you didn't provide some rationale for why they were there. But others expected a typical IOM book.
This particular panel tends to be -- tries to get to the point. I appreciate that. But it does present a challenge though to people who are looking for specific recommendations on a number of these issues. So again, I'm not sure just what balance will be struck in the second report.
By the way, those of you who -- you may want more copies. I brought three, which I will leave. For those of you who need others, just give me a call. The department has been very generous and has actually purchased a bunch and mailed them out, and allowed the Academy to have some that we can send out. And others at the National Academy Press will be happy to charge for it.
DR. MASEL: I'm Jim Masel from Zydoc Technologies. I am chair elective host. I would appreciate if the committee would clarify the difference between a methodology that I believe you are developing for data analysis, collection, measuring, performance and outcomes, versus a recommendation for data that should be collected by practitioners.
I would like to point out that there is a certain cost associated with this data collection, which we would like to have for outcomes on a national basis, ideally for items such as breast cancer and substance abuse in mental health issues, but that it is not necessary for everyone to be collecting all of this data.
If you ask every single health care practitioner to collect ten fields on smoking and ten fields on alcohol use and so on, and 50 fields on breast cancer issues, you will end up with patients spending half an hour or more filling out questionnaires in offices, and physicians and health care workers spending enormous amounts of time on this.
There is a definite cost and time factor associated with this data collection. Hopefully the committee would standardize this data, so that it could be collected by organizations. If I understand this right, you are looking for a methodology that certain organizations could be sponsored for, such as the NIH now sponsors for research programs, to identify the best outcomes, so that these could be implemented by different practitioners, without necessarily measuring all that data, perhaps other than the outcomes.
DR. LUMPKIN: Rather than getting into a discussion on that, because that really is a full discussion, let me just give you a brief answer. I think that is why we are all here.
Your statement i incorrect. Clinicians don't have to take additional measurements. They measure that data already. The problem is, in the current system, that data is collected on paper and sits in their office, so it is not reported. The burden of reporting is in fact what we are trying to deal with. That is the reason why this committee will address the issue of the computerized patient report, and also with data standards.
In fact, that kind of reporting can be a byproduct of providing clinical care. As Chris Gebbe once said at one of the meetings, and I think perhaps it is most apt, the public health and evaluation systems and the clinical care systems all look at the same data; we just look at it differently, and administrative simplification, as well as computerized patient records, how to essentially eliminate that reporting burden from the individual clinician.
So I think we need to look forward to how we are going to be able to use this very rich data set that will be available, given with confidentiality and privacy concerns to answer many of these questions, that we currently have to use special surveys and have to use special data collection methodologies to achieve.
DR. MASEL: Is it your intention to promulgate the best outcomes and clinical care pathways to the practitioners by measuring outcomes? Is that the intent?
DR. LUMPKIN: That is not really within the charge of this particular committee. But certainly there are agencies that have that within their charge.
DR. MASEL: Right. Classically, the implementation of medical knowledge is a big issue right now with compliance with best-known treatments, giving certain medications after heart attacks or aspirin. There are numerous studies that indicate that these practices aren't being exercised. Would the committee attempt to address that?
DR. LUMPKIN: We don't really have time to go into that discussion. I think we are all with you, but certainly you can't even get to that point without some of the standards that we have been talking about at this committee. That is really the groundwork we are trying to set.
Jeff, thank you. It has been a long time. I am working, I'll have my assignment done by the end of the month, as required.
DR. KOSHEL: Thank you, John.
DR. LUMPKIN: Selection of future meeting dates.
(Discussion off the record, whereupon March 3 and 4, June 16 and 17, September 15 and 16, and November 12 and 13 were selected.)
DR. GREENBERG: Just so you know, the subcommittee on population specific issues have already selected their dates for the year, so you probably need to write this down. January 12 and 13, February 9 and 10, April 21 and 22, and that is a site visit to Boston, July 14 and 15 -- that's it so far.
DR. LUMPKIN: At this time we are going to go into the work plans, the subcommittee on privacy and confidentiality.
DR. GELMAN: The subcommittee met this morning, and we have a number of items. Life everybody else, we have more to do than we can possibly do. I'll just run through very quickly the issues.
A continuing issue is that of health privacy legislation. We are not quite sure what we are going to do on that, but we remain available if we can be useful. Other issues are the question of what is an identifiable record and what isn't. This is an interesting and difficult problem that comes up in a number of contexts, and needs a lot more exploration.
Another issue is that of disease registries, what are registries, when can records be disclosed to them and how can the records be used. That is a largely unexplored issue.
Another issue is that of certificates of confidentiality. HHS has a program to provide certificates of confidentiality to qualified researchers. In discussing this this morning, we broadened it out to include some other kinds of promises of confidentiality, for example in quality assurance or NCHS activities or what have you. We have to figure out how to restructure this issue, but the need to provide protection for classes of information, so that they are exempt from other uses, particularly subpoenas, is a common theme that needs exploration.
Another set of issues, and this one overlaps in part with the one we just talked about, that of provider confidentiality, when do providers or plans have a confidentiality interest in their data. Another issue is that of pharmaceutical marketing, use of information by PBMs, and information generated by prescription drug advertising, which is something relatively new.
Some other issues that came up this morning that are interesting are problems with employers and employees with use of information. That is one that has not gotten enough attention, and really hasn't been addressed clearly in the legislation. Databases for commercial use. That one may fit in with the pharmaceutical marketing kind of thing, because it is the same idea. Another issue is that of ownership of clinical data, and that may also actually fit in somehow with the provider confidentiality issue, I'm not sure.
What we need to do is figure out which of these to take up first. We will be limited for at least a little while by the lack of staff resources, which are all supporting the Secretary right now. Hopefully we will be able to take up some of these things, either later this year or early next year.
DR. LUMPKIN: Where are there priorities?
DR. GELMAN: The Secretary needs more help than we do right now.
DR. LUMPKIN: Any questions or additions? We are glad you don't have much on your plate.
DR. SCANLON: You might mention looking into the IRB issue with the National Bioethics Advisory Commission. The National Bioethics Advisory Commission is looking at how the IRB process works. Rather than heading off in a different direction, I think Bob was thinking that maybe at an upcoming meeting we would have that commission staff come in and brief us on where they were, either as a breakout session or maybe as a full committee in November or so on.
DR. FRAWLEY: The other thing, not so much for our subcommittee, the full committee is in the President's Commission on Quality, and they are going to be issuing their consumer bill of rights. I have been tracking some of the work of that subcommittee. It might be worthwhile for our November meeting. Their report is due in January.
They have six different groups in that commission. We need to really see what they are up to. I don't know in terms of their consumer bill of rights where they are coming out on confidentiality and access to health information.
DR. SCANLON: It is actually narrowing down considerably, the President's Commission. Should we schedule a briefing for the full committee in November?
DR. FRAWLEY: I think so. Their report is due in January. Sheila is on the commission, but still, she is not covering every group. They have six different working groups.
DR. GELMAN: There are a couple. There is that and there is the President's thing, and there is this genetics stuff that is going on next week. Any and all of those may be suitable for a short briefing for either subcommittee or more likely for the full committee. I think everybody is probably interested in this stuff.
DR. LUMPKIN: Thank you.
DR. STARFIELD: We have phase one of our work plan for the subcommittee on data needs and security. The most fleshed out one is the K2 work group. There is a schedule of hearings, although we haven't got specific dates yet.
The first one is on terminology. Whereas the original intent was to hold separate hearings on terminology, in fact it was decided to combine with the CPI meetings, the hearings on terminology, which is cosponsored -- is it not cosponsored?
DR. WARD: They are participating in the planning.
DR. STARFIELD: Participating in the planning of this meeting. That will be in November, so that will take the place of our own hearings on terminology. There also will be a hearing on claims attachments sometime in January or February, and sometime later on in the spring on procedures for changing standards, new transactions, new codes and new data elements.
Then towards the end of the spring, we will start having hearings on the computerized patient record, which will consume us for most of the next year in that K2 work group.
Now, the rest of the subcommittee business, on the broader charge with regard to data needs, we are going to start off by requesting that we have two people come to the full committee meeting in November. Those two people will be John Eisenberg and Ed Sondak, to talk about data needs for quality, accountability and population needs. Then after that, hopefully we will have the new members of the committee on board, some of whom can be expected to relate to the subcommittee business. We will set up phase two of our work plan.
DR. LUMPKIN: Any questions?
DR. COHN: I think the only question I had had to do with the prioritization of hearings on version control, maintenance, issues around the standards. I thought that we had --
DR. STARFIELD: Oh, yes, that is part of the changing standards, the procedure for changing the standards.
DR. COHN: Yes, but I thought that we at least discussed moving that sooner than spring, just because it seemed to be an issue of some import for there to be hearings on. We don't have to decide that now, but I think it is a comment that I thought everybody had nodded their heads about.
DR. LUMPKIN: I think they were just nodding. That is what happens when our meeting runs so late. We can discuss that in November. Any other -- ? Okay, the subcommittee on population specific issues. Do you have anything written out?
DR. IEZZONI: Yes. You should have received at your place our subcommittee charge and our work plan. This was reviewed and voted upon at our July 21 meeting, so this is our official work plan that is in place.
We are pleased to report that we already did one of the agenda items, which was the OMB, the response to the OMB proposal, which you all helped us with yesterday.
Basically, we also have a full plate, but we have set some priorities. Let me talk you through a little bit of this. We will be focusing primarily on Medicaid managed care over the next year. We will be meeting on September 29 and 30 to both hear from researchers and government people around the issue of information systems, to allow us to evaluate Medicaid managed care. We will on the second half day of that two day period try to come up with exactly what niche our subcommittee is going to fill and exactly what report we will plan to do.
We anticipate that we will actually go out in the field and do two site visits to states that are at different places in their Medicaid managed care. We thought of Massachusetts as one of them, since a number of the subcommittee members are from Massachusetts, and Massachusetts is far along in a number of the issues having to do with Medicaid managed care. One of the site visits will be in February, so we are thinking in terms of possibly of less snow being a criteria for choosing where we go. But we are still very much in the midst of trying to figure out exactly what we will be doing on that. But we will by the end of the September 30 half day meeting have figured that out.
DR. LUMPKIN: You are welcome to come to Illinois.
DR. IEZZONI: We thought about that, but -- we all like Chicago, but --.
The other thing that has been pretty well fleshed out is that at the December 12 and 13 meetings, we are going to be talking about data systems to support public health and other health care initiatives in the territories. Our federal government has responsibility for a variety of health care programs and initiatives in the territories, but the information systems to support our knowing what is going on there in terms of health and health care are somewhat limited. We want to hear about that more officially.
So on December 12, we will be hearing about that. On the 13th in a half day meeting, we will be deciding what we want to do, based on what we have heard. Again, Medicaid managed care will be the primary focus of the year, but we will probably write up a brief report with perhaps some recommendations to the Secretary based on what we have learned about the territories.
Again, item number three, race and ethnicity categories. We will continue to review what OMB's final decisions are having to do with race and ethnicity categories. But we did manage to fax our response in the nick of time yesterday with Hortensia's yeoman effort yesterday on their proposal.
Yesterday's meeting reviewed ways of looking at disability. Apparently the federal government has 45 different definitions of disability, and yesterday we heard about five of them. Yes, those five were quite different. Based on that, we are quite concerned about how the year 2000 census will be looking at disability. This will be the first census that will be conducted after the implementation of the Americans with Disabilities Act in the summer of 1990, so this will be a nice opportunity to be able to get some information about disability from the year 2000 census. But the draft questions that have been identified for asking about this are still -- let's put it this way, unfinalized. They are developmental. They are currently as we speak literally undergoing cognitive testing among ten people, which floored us slightly, although we understand why that is, the number ten.
Apparently, these questions will be presented in draft form for review and comment in the Federal Register. So what we will be planning to do as a subcommittee is keeping track of that. When the census questions do come out in the Federal Register, we will be preparing comments based on that.
We will also be doing some more monitoring type of activities through items number five, six and seven. We are concerned especially about item number six. As you might remember, at the June meeting you approved a letter that we sent on to Dr. Vladick about OASIS, which is the method that HCFA is thinking about for evaluating home health care quality. We have got to get some feedback on what is happening with that. We haven't heard anything about it.
We are very concerned about how post-acute care is being looked at by HCFA, especially with the reorganization of the organization box. So we would like to be kept abreast of that.
DR. FRAWLEY: Lisa, did you know that HCFA is starting a technical advisory group to develop a post-acute care standardized assessment tool?
DR. IEZZONI: We had heard something about that in the wings.
DR. FRAWLEY: A technical advisory group, and that has already met. They have a tool that we were collecting data and feedback in August. The principal investigator is somewhere connected with Beth Israel in Boston. I have the information back in my office.
DR. IEZZONI: Oh, really? How interesting.
DR. FRAWLEY: They plan to have a tool out next January.
DR. IEZZONI: I do think that this is a problem.
DR. FRAWLEY: Bob and I already had the conversation, because we came back from that meeting and said this is a concern. Sue Nanamaker was at the meeting, and I was concerned about the fact that a contract has already been awarded, that there is work in progress, that there have been focus groups held already, that we have the OASIS data set floating out there, and we still have the MDS Plus. Now we've got EPS moving. There is so much going on, and the concern is, how much can the industry absorb in terms of all these different changes.
Bob and I had a conversation about making sure that people understood the HPPA mandates, and also some of the reimbursement issues that have come out of the balanced budget act, and that we can't just be jumping into all kinds of new tools and data collection strategies.
DR. IEZZONI: I think that this is actually something that the full committee maybe should be briefed on. I think this has very definite implications for the data standards. It also will have implications for privacy. So maybe at the November meeting, we could have a brief briefing.
DR. FRAWLEY: The list is --
DR. IEZZONI: I know the list is growing, but I think -- let me just state on the record, for us right now, we are interested in this issue. And unfortunately, we can't have our little feelers out constantly. So to the extent that people on the committee actually get involved, we might want to see if we can have some formal way of like informing the subcommittee that now in writing has this as an interest. So we can just make sure that there is some good communication around this.
DR. LUMPKIN: So we would like to see this committee as being an oasis in the storm.
DR. IEZZONI: I won't comment on that, no. We cannot make any pretense toward solving anything, but we just want to know what the storm might look like.
Item number seven. There are two bullets on the second page from item number seven that Barbara, you know about, but we haven't formally talked about. That is,
SAMHSA and other federal agencies that deal in mental health issues and substance abuse are themselves developing data standards for those two areas.
I think that we need to have joint hearings on -- a joint information gathering on that. What I would propose is that we do this as kind of a half day before the full committee meetings. I just think the notion of flying back and forth for these one-day meetings is getting to be a little bit much.
So we have been waiting to hear from our case staff person on this, Ron Mandershide, when this is ready. I haven't heard anything from him yet about whether SAMHSA is ready yet to present the mental health and substance abuse standards, but as soon as they are, I think that we ought to jointly think about this.
Bob, you can look at the language above that and the paragraph above the first bullet. We are very concerned about privacy and confidentiality around substance abuse and mental health especially. So it might be something that you might want to also participate in if we do this.
The second thing is that we are concerned as we talk about code sets for the standardized information sets, that we don't forget about coding of functioning status and the way peoples' health affects their lives, in terms of their daily living.
We heard yesterday about the ICIDH-2, which is under development. I think we need to learn more about that, to know whether it is something that we would want to push for further. But I think that this also might be the topic of a half day meeting prior to one of the full committee meetings jointly with our two subcommittees to begin to look at coding sets for looking at functional status.
So that is our subcommittee.
DR. LUMPKIN: Can you perhaps help me with point number one?
DR. IEZZONI: Yes, we're still working on it.
DR. LUMPKIN: It was my understanding that substance abuse and mental health services were a part of HPPA. Are they operating under the assumption that there are going to be separate standards for those?
DR. IEZZONI: I don't think so, but I think that it is just that within these federal agencies, they have been working on specific ways of collecting information for their programs on this. I cannot say anything more about it, other than that, John. That is all I know.
But Ron Mandershide -- is Ron here from SAMHSA? Is anybody here from SAMHSA who can comment on this? Can you come to the microphone, please?
DR. BROWN: My name is David Brown. I am a consultant for the Center for Mental Health Services. I work directly with Ron Mandershide. There are several activities going on around confidentiality, and then the data standards, especially relevant to managed care.
These are still work in progress. It seems to me that sometime in the near future, the reports will be made available. I don't have an exact time.
DR. IEZZONI: Yes, that is what we had heard.
DR. LUMPKIN: But I think it is important that you do that, because we need to understand to what extent -- my understanding was, the charge to this committee goes across the HHS. To what extent there may be other entities that are looking at data standards that may not be consistent, I think we may have a major problem here.
DR. IEZZONI: That is why we want to bring this to a joint committee meeting.
DR. LUMPKIN: Why don't you put that in your work plan?
DR. IEZZONI: It is in the work plan. It is on page three. It is the first bullet.
DR. LUMPKIN: I know it.
DR. IEZZONI: Oh, you were teasing me?
DR. LUMPKIN: Yes, I was. I'm sorry. Barbara?
DR. STARFIELD: Two comments. This is a more generalizable issue about things that are important and still aren't being considered in the context of data standards. The core data elements is one of them, because we are really not even talking about the relationship of them to the data standards, as we were discussing them. So it is important we have to step back and look at the broader picture of what we are doing, even in the K2 area.
The second thing about the joint meetings. Lynette gave us three dates. You all have that. So why don't you keep that third date, at least those of you on the subcommittee on data needs, keep that third day --
DR. IEZZONI: Can we have those dates again then, because some of us only wrote down the two.
DR. STARFIELD: It was March 3, 4, 5, June 16, 17, 18, three days.
DR. LUMPKIN: Microphone.
DR. STARFIELD: Oh, I'm sorry. March 3, 4, 5, for which we selected 3, 4 for the full committee, but those three days. Then 16, 17 and 18, which 16 and 17 tentatively were --
DR. WARD: That is June?
DR. STARFIELD: June, excuse me, 16, 17 and 18. September 15, 16, 17, and November 12 and 13. I think the problem with November 11 was that the Data Council meets on the 11th.
DR. COHN: Maybe I'm not tracking on this. Are we mandating now that these meetings are going to be two and a half days or three full days now?
DR. STARFIELD: No, it is just one of them, but you are going to pick one of them for the joint meeting with that group.
DR. COHN: Because in my own mind, I am hearing that there are some issues that need to be discussed across -- in my own mind, I am uncertain whether it is a half day session, whether this is part of a full two days worth of hearings that are separate. Can I have some clarification on that?
DR. IEZZONI: I think one of the problems is that we don't yet know from SAMHSA, for example, for the mental health and substance abuse, what this is going to involve. I can't imagine it will be more than a half a day, but we don't know when it is going to be. Ron Mandershide in June told us it would be sometime this fall, but we haven't heard and we don't know when that will be.
Around the ICIDH, I don't feel any pressure to do that one immediately. I think that that one we could be thinking about for next spring or early summer. They are currently in beta tests around that, and it might be revised frankly by the time we get to it, fairly significantly. So I think that these are -- just hold these dates, and when we come up with something specific we'll let you know the minute we know.
DR. LUMPKIN: Can I suggest that we be a little bit more proactive than that? Can we perhaps ask for a two or three or four page document from SAMHSA that would be a briefing document, so that we can perhaps in November more accurately determine how much we need to delve into this?
DR. IEZZONI: We certainly can -- Marjorie, is that something that we can get from SAMHSA, a briefing document about what they are doing?
DR. GREENBERG: We can certainly ask. Is that reasonable, David?
DR. BROWN: Yes, that is reasonable.
DR. COHN: Can I ask a question about that also? I guess I'm a little concerned about how this is being described, and maybe I am misunderstanding. But my presumption is that what we want to find out is how SAMHSA wants to use standard data transactions for substance abuse treatment and mental health services, as opposed to review federal agency proposals for standard data transactions that are somehow different than otherwise mandated standards. We are dealing with encounters, we are dealing with claims, we are dealing with request for services already. Am I missing something here? Do mental health and substance abuse services require different standards?
DR. MAYES: This is Bob Mayes. Just to let you know, this is an issue that has come up recently in our discussions, in fact, at our infrastructure meeting last week. We have been focusing on outreach plans and public relations outside the federal government. It has become quite clear that we need to make a formal effort to be sure that the various agencies are one, aware of HPPA, because many of them are not, and there are in fact plans by some -- I know within HCFA even, different agencies are beginning to think they are going to be able to get different things by using the data, and they are not realizing we are talking about solidifying this.
So we are putting together a plan that will actually go out as some sort of OMB notice or something like that to the federal agencies, saying you are going to have to comply with this, here is the highlights with what it is that HPPA standards are actually about, and see if we can't begin to get one, recognition and two, participation by a much broader group within the federal agencies. Even though we have representation, we tend to have representation only from one small group within the larger agency.
DR. COHN: It is obviously an NCVHS issue, but it is really an HHS Data Council issue. I am looking at Dr. Braithwaite as I say this. Clearly it is not our responsibility to coordinate HHS, to have them follow federal legislation and federal mandate, but it is the responsibility of HHS Data Council to make sure that all the agencies are following it.
I am already telling the people within my organization that they just can't go our and redefine the standards. Everyone out in industry would like to basically redefine, yes, standards are fine, but we want it this way. As I am telling my organization, I think obviously everyone needs to be spreading that message.
DR. MAYS: Probably the most focused issue right now is -- and we touched upon this yesterday when we were talking about what goes in a claim and how we should clean out the data content and things like that.
That is the problem. There are many organizations that are either capturing information that they use currently within the claims structure, and would be very upset if we reduced that content structure without getting some other mechanism to collect that information.
So I think that the committee does actually have both responsibility and interest in our discussions about how we structure claim and encounter versus claim attachment, and that whole discussion, into making recommendations. Certainly there is concern, particularly on the public health side, that these standards are either going to freeze them in or freeze them out of certain of these transactions that they have historically been using, or they are counting on that we are going to add more stuff that they would like to get in and have never been able to in the past.
So it would be appropriate, given all the work in the past that this committee has done, to begin to make recommendations on whether or not the best approach is to fold it under claims or treat it as a special public health claims attachment or what have you, whatever the approach.
DR. SCANLON: SAMHSA has been working with the states for a long time actually, and they have worked on data sets covering other than claims and encounters. They have developed data sets over the years for state reporting for public programs relating to facilities and organization and financing. Each state reports this aggregate information.
But to the extent that they are actually looking at individual client level data, I think we have several times tried to get them into this. So I hope they are aware of that now.
One of these issues is coming up before the Data Council tomorrow, the issue of public health, quality assurance and oversight input into these standards' content process.
DR. LUMPKIN: I think this raises an issue which is perhaps -- as you suggested, while there is a SAMHSA and mental health component of that, I think that what it raises to me is concern -- and I think about my colleagues on a Medicaid level in our Medicaid agency, our mental health agency, our substance abuse agency, which when I mention HPPA, don't have a clue what I am talking about.
At some point, we have been so focused on what the standards are to be because the time frames are so tight, we may need to come back and review how we let other governmental agencies that potentially at the state level may be establishing standards. Certainly Medicaid agencies are involved with a small portion of the medical market, how we make sure they are on board with what is being done.
DR. COHN: It sounds like the first item here is a very high priority item to get information on. I know you appreciated that.
Obviously, the code set piece -- I guess my own thought was that that should somehow be folded into -- we had all been talking about a couple of days more of hearings as we delve into code sets again. I am wondering if that should be part of that, because I know that the subcommittee that you chair has a lot of interest in the overall issues of code sets and classifications, too. That may be something we should consider.
DR. IEZZONI: That would be a fine way to do it, too, if you felt that it should be folded into a day or two days of hearings on code sets. I just would not like it to be the second poor cousin, tacked on to the end, when everybody is looking at their watch and wondering when their airplane is going to leave.
DR. COHN: Agreed.
DR. STARFIELD: That is the problem with the scheduling, is that all these third days are at the end.
DR. IEZZONI: Right, I know, that is the problem with the scheduling. I'm thinking about it.
DR. GREENBERG: We could revisit the days that we selected. Actually, when you talk about these two subcommittees, you have about 90 percent of the members on the committee, 95, on one or the other subcommittee.
DR. LUMPKIN: But I think we need to also explore the utility of tacking something on at the end or in the beginning, and whether or not we are going to get the most utility out of the meeting itself if we start putting too much in the beginning or end.
DR. GREENBERG: Or whether it should be part of the actual meeting, because you are talking about essentially the full committee, when you talk about these two subcommittees.
DR. IEZZONI: That is a very good point. We could just have an afternoon of the full committee meeting. I would like to have actually Bob Gelman be aware at least of the substance abuse and mental health portion of the discussion.
DR. GREENBERG: He is already an honorary member of the standards subcommittee.
DR. FRAWLEY: I think that in some situations, some of these topics we could accomplish if we could get a briefing document ahead of time. They had hearings a couple of years ago in ICDH. I know that we offered comments on that classification.
DR. IEZZONI: That was a while ago.
DR. FRAWLEY: Was it three years ago?
DR. IEZZONI: No, it was longer than that ago, according to Dr. Placak yesterday. It was quite a few years. It was more than five years ago.
DR. FRAWLEY: No, it wasn't.
DR. IEZZONI: That is what he said.
DR. FRAWLEY: No, it wasn't, because I offered testimony and it would have been 1994. The only reason I am raising that is because I am sitting here, thinking when we had the hearings on ICDH three years ago, there were very few people who were familiar with the classification, and it was not really in use in this country.
There were a number of international people who came to the conference. The concern that I have is, three years later, I don't think we're going to hear anything new or different, because the utility of that classification -- again, I'm talking about widespread acceptance and use in this country, unless something has happened in the last three years that I am not aware of, I think a briefing document could be useful. Then the subcommittee could decide do we need to jump further into this issue.
DR. IEZZONI: Actually, I was very careful in the way that I phrased this, to not just limit it to ICIDH. I don't think this is a yes-no ICIDH issue. I think it is more, are we going to be capturing things beyond the acute medical model of disease.
DR. FRAWLEY: If you look at the standardized assessment tool that HCFA has got right now, that they are testing and getting comments on, it is not at all using ICIDH. So that is what I worry about, that we have got all those different little projects going on, and there doesn't seem to be any integration occurring anywhere.
DR. LUMPKIN: But I think that certainly this discussion indicates that the broader scope that the committee is suggesting is very appropriate to look at those kind of coding things. Given the large membership that is represented by the two involved committees, it does seem like this might be more appropriately done within the context of a full committee meeting. Those who aren't on those two committees probably would be interested, anyway.
DR. STARFIELD: Are we thinking the same thing about the mental health and substance abuse?
DR. IEZZONI: I think that can be shorter, though. That could be a several hour kind of thing.
DR. FRAWLEY: The only suggestion I have is that it is really helpful to have a briefing document ahead of time, as opposed to coming to a full committee meeting, having no documents and then people coming to present. It is much easier to have a document that you can review ahead of time, and then be prepared for the discussion or ask questions. But to have people show up and make a presentation without anything ahead of time to me is a waste of very valuable committee time.
So I just don't think we should be putting things on the agenda until we have commitment from the agency that there will be a document prepared ahead of time for the committee and a representative who is knowledgeable. That is the other problem we get, is that the person that comes is probably the least knowledgeable person within that agency on that particular project.
It has happened already, we know that. We know that with OASIS, so we have to be careful of that.
DR. LUMPKIN: On that particular issue, we may want to get a briefing document, have a brief discussion at the November meeting, and then that will allow us to better assess how much we need time we need to spend on it. I think you are correct; there are so many things on our plate, this is another thing we need to assess. Having a briefing paper and short discussion in November I think would be the best way to do that, if that is agreeable.
DR. GREENBERG: Okay, we'll follow up with SAMHSA. It might be more complicated on the functional assessment, but we will consult with Lisa.
I did want to ask, I think at the last meeting in talking about classifications, there had been a request for a presentation on the clinical modification of ICD-10. That is going to go out for public comment in November or December. That doesn't exactly fit with the -- it is obviously part of the family of ICIDH, but with the particular functional assessment issue, is less relevant to this, to ICD-10. So we need to know if and when you would like that.
DR. LUMPKIN: And I just love putting things on Don's plate. He'll be sorry he left me in charge. Do we have anything else on your work plan?
DR. IEZZONI: No, that was enough.
DR. LUMPKIN: Are there items that other members would like to put on their plate?
DR. COHN: I just have a question. I am very impressed with the fact that you have everything scheduled out for the year, which is obviously much more than other subcommittees I am on have. But is there a way for us in advance to get a very good idea of what issues are going to be handled in exactly which meetings?
DR. IEZZONI: I think that the September 29 and 30 meeting is going to be key, because that is where we really are going to be trying to figure out what we're going to do on the Medicaid managed care. We already know that the December 12 and 13 meeting is going to be on the territories. So that will be the topic of those two days. Or the 9th and 10th, whatever the dates are of the December meeting.
The rest of the dates that were listed for us, Simon, are all having to do with the Medicaid managed care, and two of those will be site visits.
DR. COHN: Okay, great.
DR. GREENBERG: I think the agenda for the September 29 and 30 meeting is already on the --
DR. IEZZONI: Carolyn Rimes is working on exactly who we will have be presenting. It will be most researchers who have been hired by contractors of the federal government to look at Medicaid managed care information systems and federal data folks involved in that.
DR. LUMPKIN: Thank you. Jane, anything else in the work plans? Then we will move forward. Data security standards. You have in front of you a document from the work group and the subcommittee, which is -- we attempted to incorporate the comments from our earlier discussion yesterday. It is entitled draft. I'll give everybody a chance to take a quick peek at that. It is to be signed by Don.
PARTICIPANT: Do you want a motion?
DR. LUMPKIN: Please. It has been moved. Is there a second?
PARTICIPANT: Second.
DR. LUMPKIN: Moved and seconded. Discussion?
DR. COLTIN: We had I thought agreed upon two additions or amendments to this letter last night in the subcommittee, and I don't think they have been incorporated into this version. Is that correct? There was one under individual authentication of users. It did get in?
DR. FRAWLEY: And then we added the audit trails under monitoring of access.
DR. COLTIN: Okay, so it is under access controls, is that where it is?
DR. FRAWLEY: Monitoring of access.
DR. LUMPKIN: Monitoring of access.
DR. FRAWLEY: Your question went under individual authentication abusers. We added the phrase, each organization should have policies and procedures in place to enforce the appropriate use and maintenance of access methods. That was the addition.
DR. COLTIN: I don't see that.
DR. LUMPKIN: It is on the top of the second page.
DR. FRAWLEY: And under monitoring of access, you asked us to address audit trails, and we changed it to say, organization should survey audit trails and mechanisms, and then the last change was that organizations that license or accredit health care organizations should consider incorporating these requirements into their standards. Those were the three changes.
DR. LUMPKIN: I thought we were going to say, especially HCFA.
DR. FRAWLEY: No.
DR. COLTIN: Thank you.
DR. LUMPKIN: Other questions? We have a motion for us to send this letter to Secretary Shalala on security, which is an expansion of what we discussed yesterday. All those in favor, signify by saying Aye.
(Chorus of Ayes.)
DR. LUMPKIN: Opposed? Abstentions? Carries. We'll send the letter. Recommendation on payor I.D.
DR. STARFIELD: I do have a summary of that, if I can just find it here. We had a report from the Data Council. We concur with their plans to send that document they prepared out for comments.
The main proposal there is for a nine-digit number and the subcommittee proposed that they consider the possibility of potential change to an alpha numeric at some time in the future. But basically, the action we took was to approve it and to concur with the plan to send it out for comments.
DR. LUMPKIN: So the motion is that the full committee accept the development of a nine-digit payor I.D., initially numeric, with the provisions to be alpha numeric in the future, with the proviso that consideration for internal routing of sub-organizational units be considered. That is the motion. Is there a second? It has been moved and seconded. Is there discussion on payor I.D.?
DR. GREENBERG: The internal routing was to be separate from the number itself.
DR. STARFIELD: Yes, and it wouldn't necessarily be a national standard for that. It would be to allow separate organizations --
DR. LUMPKIN: It would be to take into account, so that they could incorporate it into that nine-digit number. They could allow some other mechanism that would be an additional number. There are any number of options. We just believe as a committee that there should be some consideration of the fact that sub-organizational units should be identified.
DR. MOORE: But I thought that wasn't part of the identifier itself. What you're saying is that you could make it part of the identifier itself?
DR. LUMPKIN: No, no. We are not prescribing how. We are just saying that it should be considered. If somebody gets this great idea about incorporating within it, I don't think we would be opposed. We are just saying that it should be considered.
DR. HARDING: Mr. Chairman, I don't want to obstruct, but is this written down anywhere? We are going to vote on a motion that we haven't seen yet in written form? This seems to be an important issue. I'd like to be sure that we know what we are voting on exactly.
DR. LUMPKIN: Okay. We'll get something written up, if it is agreeable with this group. We have now heard the motion. We will consider it after lunch. We will take a vote on that motion in written form.
DR. STARFIELD: You understand that what we are proposing is a little different than a motion by this committee. We are concurring with the Data Council's plan to send it out for comments, okay? So it is not a fait accompli.
DR. LUMPKIN: So we'll do it over lunch. The next item then is the unique health identifier for individuals.
DR. STARFIELD: Let me just give you a brief introduction. You have something at your place on this item, don't you?
DR. LUMPKIN: Well, I was going to say that it has come to my attention that our consultant on privacy and security, Mr. Gelman, has developed a letter that is very consistent with our couple of paragraphs, yes.
DR. FRAWLEY: There are a couple of problems. The other members who were not in the subcommittee meeting have no idea what the recommendations from the team were, or our consultants' report. They never saw it. I don't think this is something we can just jump into before lunch.
DR. STARFIELD: We're not going to jump into anything.
DR. FRAWLEY: No, they haven't seen it.
DR. STARFIELD: We haven't even seen Bob's letter, is that right? I think you can present the thoughts that are incorporated in the letter.
DR. GREENBERG: May I confirm this, that members who are not members of the subcommittee on health data needs, standards and security did not get any of the information about the unique health identifier, either the team's report or the consultants' report? I'm asking, because they were supposed to get it.
DR. FRAWLEY: We did get it. What we didn't get was Bob's letter.
DR. STARFIELD: At least one NCVHS member told me that he or she did not receive it. Okay, you did all receive it. But you weren't of course at the briefing or the discussion yesterday. Okay, I just wanted to find out about that.
DR. LUMPKIN: What we got ahead of the meeting was essentially the draft before the draft that was handed out yesterday, but the content was not any different. It was just reformatted, as far as I can tell.
DR. COHN: The only thing I was commenting on was that I thought everyone should take a look at Bob's letter, only because I think it is a very well-written further discussion of some of the principles that we came to at the subcommittee meeting yesterday.
DR. STARFIELD: Who is Bob's letter addressed to?
DR. GELMAN: I just prepared this in the event that it would be useful, not to impress anybody.
DR. GREENBERG: You also didn't receive either the contractor report or the one by the implementation team?
PARTICIPANT: This I got off the e-mail.
DR. LUMPKIN: This is where we are at. It is by my watch, which is fast, 11:45, which means it is 11:40. We are scheduled to go to 12:30, at which point we have lunch at 1:30. I would like to proceed. If we have some concern, let's perhaps take a few minutes and read the letter, and then if we have questions or any members would like to ask about the proposal. Our recommendation is -- Bob, is your letter an expansion of what the committee voted on?
DR. GELMAN: By coincidence, pretty close, yes.
DR. LUMPKIN: The internal committee, the implementation committee came forward with a recommendation of one identifier. The committee last night in their discussions felt that they were uncomfortable with coming forward with a recommendation for a single identifier, for two reasons.
DR. STARFIELD: For a specific one, right?
DR. LUMPKIN: Yes, for a single specific identifier, for two reasons. The first is that without a confidentiality and privacy framework, the presence of a unique health identifier increases the risk of individual health data becoming accessible. So therefore, there should not be a change to a unique identifier without there being a structure.
Based upon our prior communications to the Secretary, we did not feel that the HPPA provisions, which directs the Secretary to write regulations, will provide adequate protection. So the committee felt that there needed to be legislation in place before implementation of the unique health identifier.
The second concern of the committee was that they felt that the jury may still be out on which one is the single health identifier. We wanted to have an opportunity to get public comment upon it, so that we would identify what we felt were the highest ranking models in the analysis that we received, which was an enhanced social security number, the ASTM UHIT, and a master patient index as an example of a multiple identifier system for identifying patients without having a unique number assigned. We would let the public know that we were considering those. We would look at the public comment that we received back, and discuss the issue and see if we are ready to make a decision at our November meeting.
Is that a fair summary?
DR. STARFIELD: Fair, excellent.
DR. LUMPKIN: Based upon that, let's take a look at the letter, and after we read the letter there are questions or concerns, then we can have some discussion. We will discuss until we get to 12:30, at which point we will adjourn for lunch. We will write down the tabled item and then we'll bring that for discussion after 1:30.
DR. MAYES: Point of clarification. On the master patient index model, would that mean that in terms of what goes on for instance on a claim to identify a patient, would there be no standard for that? Anybody could use whatever they are currently using?
Something has to go down on a block at some point. The master patient index works very well if you are trying to locate information, but if you are trying to actually identify a single instance of somebody, I'm not sure how that would work.
DR. LUMPKIN: Well, I think that the question that you are raising is one that is a valid question and one which we would have to consider. I don't think we particularly want to get into that right now. But it certainly filled a number of criteria, the ASTM criteria. It ranked relatively high.
We heard some testimony on that in San Francisco about a similar system in California, so before we make that consideration, we did want to get some public comment on that methodology.
DR. MAYES: Right, but since we have to standardize those transactions, then we would have to put in, in the regulation for the standard, because that is one of the fields. We would not then be requiring a specific I.D. for that?
DR. LUMPKIN: Those issues were discussed, and were raised, and we want to hear more input on that apparent weakness.
DR. STARFIELD: There is one other issue. This committee is on record as far back in history as we can recall as favoring a unique health identifier. The question is that there should be no decision on one until --
DR. MAYES: Right, it is just a timing issue. Either we decide we are going to put off all of the transaction reg, or we state it in the transaction reg that at this time, that field would not be standardized, and go forward there.
DR. FRAWLEY: Can we ask Mary Emerson from the team to come back, that I just have a few questions about the team's ranking?
DR. LUMPKIN: Sure. I'm just looking at -- from the team's report the matrix of the evaluations, where the team assigned a numeric based on 30 criteria. What I was confused about, because I didn't have the chance to tally this up when you presented the list yesterday afternoon, according to what we were looking at last night, the ASTM criteria did not come out the highest ranking on your numeric evaluation. Yet, that was the recommendation that you put forward.
So I was just curious, if you could just for the full committee walk us back through that matrix, and why you chose the ASTM approach, as opposed to some of the others that ranked more highly.
DR. EMERSON: I'll be glad to. The way that -- the proposal as we evaluated it, what we had done was to contact the proponent of the proposal for each of these, and if that proponent had a document or a method that they proposed for implementing that particular candidate identifier, we used that scenario for the implementation.
For the ASTM method, the person that we contacted was Barry Heebe. His proposal for the infrastructure was a very modest one, where the identifiers would be assigned by a small staff, 30 people or so, and the software supporting this and the other infrastructure supporting it would be very small, he felt.
That was the scenario under which we evaluated the ASTM proposal. However, as we -- and that is what you see here in the rankings. A number of the categories that deal with the infrastructure that was needed for the proposal did not score highly there, because we felt that that was not a realistic or robust enough infrastructure to support the identifier.
That is why, as we began to evaluate this further and really felt that the Social Security Administration had that infrastructure in place, we married those two ideas, the infrastructure and the supports that the Social Security Administration could bring to this because of their experience with the technical merits, we thought, that there would be in having a unique identifier that was used only for health care.
So we don't actually have here an evaluation of what our proposal is, in this matrix.
DR. LUMPKIN: I would also suggest that the ASTM criteria was a rough screen. What they never talked about is that you cannot weight each one of these criteria the same. In fact, those numerical summaries assume that each one was of equal weight. So when you look at it, there are some that clearly stand out.
I think our document is perhaps in error, because I think we were talking about the hybrid VA model of identification, which ranked much higher than -- which I think is the master patient index we were talking about, not the MRI, which was the Medical Records Institute.
DR. EMERSON: The hybrid VA model, actually some people have asked us, what do you mean by hybrid, and we probably should have explained that. But it is the hybrid of the ASTM identifier along with a master patient index.
DR. FRAWLEY: The reason why I ask is because, under your -- on this matrix, the highest ranking under 135 is hybrid VA, followed by the CPRI proposal, which is 131 points, followed by ASTM, which is 128 points, followed by personal immutable properties, which is 123.
I guess my question was, looking at the numerics there, what I was trying to get to is the proposal you put before us yesterday to adopt the ASTM approach. What I was trying to get to is, based on this, how did you get to the next?
DR. EMERSON: I'll just make two comments on that. First of all, the proposal as we proposed it isn't ranked here. So if we were to rank that proposal, some of those infrastructure scores would go up. It might then, if you totalled everything, come out higher or equal to hybrid VA or CPRI.
But I think another point John just made and also is made in the paper, that the criteria have not been weighted or ranked themselves, and you are not really sure that a five for accessible means the same thing as a five for mappable.
So I'm not sure there is too much value in just adding up the scores.
DR. STARFIELD: Just a clarification again. We also have this contractor's report, which in its previous version had a rating tail. This one doesn't. If I recall it correctly, the approximate rankings were the same as you came out with. Although I think the modified social security number came out higher than ASTM.
DR. EMERSON: As long as I am here, I would like to make one clarification on a question that was asked yesterday about the cost figure that was in the paper. I'm not sure I explained it adequately.
The figure referenced in the paper, three to six million dollars, was the estimate that the Social Security Administration had made for their own process to re-verify the data and re-issue social security cards. It was not an estimate of the cost of implementing this proposal.
DR. FRAWLEY: Thank you, Mary.
DR. LUMPKIN: Has everyone had a chance to read the letter? Except me. Further comments?
DR. COLTIN: Having read it, I think the one area of inconsistency between the recommendation that the subcommittee came out with last night and Bob's letter is in his bullet four, where the subcommittee's recommendation begins with the statement that the committee continues to support the concept of a unique -- it says patient identifier, but I think we meant health identifier for individuals. That was the term we were going to use. Bob's fourth bullet is very equivocal about that, saying it is not clear that that is a good idea.
DR. GELMAN: And basically, my point of view -- and it may not be shared by everybody -- is, I'm not prepared to concede anything in this area, whether we need an identifier, what criteria should apply in selecting an identifier, or anything else. I think everything is open. If the committee is on record as supporting something, I don't care about that, either. I wasn't on the committee then. And even if I were, I would revisit the issue.
One of the purposes of this proposal is, if nothing else, -- basically I tried to list all the reasons that are floating around out there on this, and to offer the Secretary as much cover as possible for not selecting an identifier, which appears to be what HPPA says, although I'm not sure that is the only interpretation of the law.
DR. COHN: Perhaps I should clarify. As I look at the brief two paragraph note, I should first of all comment that as this was written, it noted the concept of a unique patient identifier. In the second paragraph, it mentions basically a unique health identifier for individuals and then in blocks, or another methodology that accomplishes the same purpose.
So I think we were trying to be admittedly a little bit perhaps vague or open in terms of this. I think we all believe very strongly that we need to be able to link and uniquely identify the individual, which I think is the key concept. The question is, how do you do that.
I think that that was what Bob was commenting here, that this would need to link, but exactly how you do that and exactly what methodology, is a little less clear at this moment.
DR. FRAWLEY: The problem that I have is just having been at the National Press Club when the Secretary gave her speech. She talked about the fact that health information is used for health purposes only, and then in limited situations for other purposes. She was very clear on her message about the current use of information for non- health care purposes and how it wound up in the payment stream, and that employers had access, and on and on and on.
So after listening to her speech and understanding her principles and what her vision is, I go back to -- and I had this discussion with Bill last night, having worked with Bill when he was on the Senate Finance Committee, we are not going to be able to link information.
It is very clear what the Secretary is proposing in terms of patient authorization and access and her principles. She was very clear on her five principles and how this would all -- her framework. So then I began to say, you're not going to be able to do a lot of linking without the specific authorization of the patients, so then it behooves the question, why do we need a unique health identifier?
If I go to the doctor one time and use my employer sponsored benefits to pay that claim, and the next time I choose to pay out of pocket, I don't want a number of that second encounter lining back to that first one, because I chose to pay and I don't want the information linked unless I specifically authorize it.
So it then starts to beg the question, why are we spending so much time on a unique health identifier when the Secretary's privacy principles are pretty clear, that you're not linking; you are going to be linking at the lowest level non-identifiable -- she was very clear on that in her speech at the Press Club. You can download the text from the website. Then I am like, why are we spending all this energy, and then we hear proposals that it is going to cost three to four billion for the Social Security to verify and issue new cards.
It is kind of like we have the cart before the horse. I am beginning to say, why are we spending so much energy on a unique health identifier when the privacy recommendations that are going before the Congress on Thursday are pretty clear about what you can and cannot do with data and linking and so forth.
So I just get concerned, because I still don't think I have the answers to the question about why we are spending so much time on an identifier, when once the claim is paid, technically you are not supposed to be linking it to anything or doing anything more with that data. As a patient, I don't give you authorization for anything else.
DR. LUMPKIN: Let me perhaps deal with procedure. With the agreement of the committee, we had a subcommittee that met last night, and they came forward with a proposal that they voted on. That represents the document that you have before you.
My reading is that there is I think a significant difference between the letter in that proposal and how the issue is presented. So what I will do is accept the statement by the committee as being the item that is before us, that we can then modify that statement if we so choose by amendment, and then we can address the letter if we adopt or don't adopt the statement as to whether or not the letter reflects the position of the committee that we voted on.
So I will take that -- since there are at least two members of the committee on that, I will assume that this document that you have before you, with the word unique health identifier clarified instead of unique patient identifier in the first term, as being the motion that is before us.
DR. COLTIN: I think you have to add the words "for individuals".
DR. LUMPKIN: Okay.
DR. COLTIN: Health identifiers for individuals.
DR. LUMPKIN: For individuals, which is the HPPA language.
DR. GREENBERG: Do you want to read this?
DR. LUMPKIN: Okay, I will read it, so everybody knows what we are talking about. While the NCVHS continues to support the concept of the unique health identifier for individuals, we believe it would be unwise and premature to proceed to identify and implement such an identifier in the absence of legislation to assure the confidentiality of individually identifier health information, and to preserve an individual's right to privacy.
Once the public is assured of this protection, we recommend that a unique health identifier for individuals, or another methodology that accomplishes the same purposes, be selected on the basis of the ASTM criteria, coupled with a cost-benefit evaluation, and public comment on the leading candidates: an enhanced social security number, the ASTM UHID, and a master patient index, which could also be called a hybrid VA model.
Discussion and a motion?
DR. COLTIN: I thought from what Mary said that the hybrid VA model was actually a combination of the ASTM and the MPI, the ASTM UHID and the MPI.
DR. LUMPKIN: If it is agreeable, if we will take that --
DR. COLTIN: Or a combination of these.
DR. LUMPKIN: I think that what we can do is call that the hybrid VA model, and if we can get language to substitute in that, the ASTM UHID is a modification of that; it is not exactly that one. The enhanced SSN is really the CPRI model.
DR. COLTIN: Or some combination.
DR. LUMPKIN: Or some combination. What we are doing is picking what appears to be the front runners for discussion and debate. Then we will then, as we have done with other standards, having had some discussion and debate and opened it up to public input, then try to make a more informed decision on the single candidate.
DR. EMERSON: I think the point has actually just been made, but I want to make sure that the committee understands that the VA model does definitely include a unique identifier, that is, the ASTM identifier, plus its master patient index. You can't just say that the VA model is the master patient index.
DR. LUMPKIN: I think our paper from Solomon, does that not describe these models? So that will be on our web page within the next week. So when we refer to that, this language will refer to particular models described in our commissioned document.
DR. COHN: I just want to make a general comment. I think that there are two fundamental issues that we need to discuss. Somehow, to discuss them together creates great confusion, at least in my mind.
This document and its two sentences was meant to address two issues. One is, is it too early for us to make a comment about things because of issues of privacy and confidentiality. I think that is question number one.
Now, beyond that there is a second question of, if indeed that is the case, do we want to wade into anything? Do we want to make a comment about cost justification, ASTM criteria, current leading candidates and all of this stuff?
I think those are the two pieces of this. I almost think for the sake of some simplicity, rather than trying to discuss them like this, it makes sense to try to disaggregate those issues and see what sort of consensus there is around the table.
DR. LUMPKIN: I'll take that as a motion to divide the question.
DR. FRAWLEY: I think that is the problem that I have. We are jumping at this in different ways. The first thing we know is that there are other privacy concerns. Secondly, we are concerned because the team came forward with a proposal and said, here it is. My specific concern was, we don't have enough cost-benefit information. The industry has not had an opportunity to provide the committee with some input.
Our concern was that we don't know that the
ASTM approach is necessarily the best approach. Most organizations have a master patient index, including social security number that they use to identify their plan participants or their patients. So what would be the impact to industry to all of a sudden say, okay, guys, we are switching over here now?
So my concern was, before we make a decision, we needed some more information. But I think they are very separate issues.
DR. LUMPKIN: So we have now divided the question, unless there is opposition. The motion before us, I will read it.
DR. GELMAN: Before we do that, I don't want to make this too complicated in a parliamentary way, but I am prepared at an appropriate time to offer an amendment, which is to strike the entire subcommittee recommendation and substitute the text that I distributed. I don't care about this, we can have a discussion and figure out what to do later on. That is an option.
DR. LUMPKIN: Well, I will entertain that motion before we divide the question because otherwise, it won't fit.
DR. GELMAN: I think we can have a discussion and not worry about a motion quite yet.
DR. LUMPKIN: Well, but what I heard other members of the committee saying is that by discussing these two issues together, it is --
DR. GELMAN: Let's discuss them separately then. Whatever you want to do, but I don't think it matters much.
DR. LUMPKIN: If it is agreeable, we are going to discuss these issues separately, and then we will go back into a more formal discussion in which we will entertain the full motion.
So we are just going to have a general discussion on the issue of, should we have a unique identifier sans legislation on confidentiality.
DR. HARDING: Just a point of information. What were we asked to do?
DR. LUMPKIN: HPPA was to delineate a unique health identifier for individuals to the Secretary.
DR. HARDING: To the Secretary, the best unique identifier.
DR. LUMPKIN: That is correct.
DR. HARDING: So it wasn't whether we should or not.
DR. LUMPKIN: That is correct.
DR. HARDING: They asked us --
DR. LUMPKIN: We are taking issue with Congress.
DR. HARDING: There also was not a timetable.
DR. LUMPKIN: Yes, there is.
DR. HARDING: There is a timetable for that recommendation?
DR. LUMPKIN: So we are charged by Congress to make that recommendation. The discussion that we are having is whether or not to respond to the Secretary that what Congress said was well and good, but barring a legislative framework, we don't believe we should have a unique identifier.
Discussion on that issue?
DR. WARD: I would say that this language addresses that.
DR. LUMPKIN: Which page?
DR. WARD: The first paragraph of the committee's proposal.
DR. LUMPKIN: Can you maybe pull the microphone toward you?
DR. WARD: Addressing that issue of timing, the proposed language the committee has submitted that is in front of us addresses that well.
DR. LUMPKIN: Further discussion on the concept?
DR. GELMAN: Yes. I certainly agree with the sentiment. But I don't think that is the only reason not to make a decision about a unique identifier. There are lots of other reasons. I think we are dealing here in an incredibly changeable area. There are lots of activities going on elsewhere. It is quite possible in five years, everybody could walk around with a plastic card with your public key/private key infrastructure stuff on it. This is a problem.
One of the things that I think is missing from all of the discussions here is that everything has been too narrowly focused on the health care system. There are identification problems everywhere, everywhere. The Internet of course is making this even more central. There are going to be solutions to this, and we don't know what they are. There may well be better forms of identification, better forms of authenticating who people are, verifying everything that are going to come on board in other directions. Electronic commerce activities may lead us all in that direction.
It may well be that the health care system will use the same kinds of identification that will be used for other government services, or credit cards, for who knows what. No one can tell you exactly how this is going, whether it is going to be a digital signature or a PKI or something else. There is just a lot of development here. I think that is another reason for saying that this is a bad time to proceed.
It may be if you wait five years and this all develops, that the infrastructure for a better identification -- and I don't mean to suggest that there aren't privacy problems with anything that happens; those are there regardless. But the infrastructure will develop in another way and the health care system can get much more of a free ride out of because somebody else developed this. I don't know.
But to try and make a decision today -- if Congress passes a privacy bill tomorrow, the bill may not be good enough to deal with these questions. Even if it is a wonderful bill, I am still prepared to say we are not at a point where we need to pick a health identifier. This is a bad time to do that. We are going to have to get through for a period of time until the technology develops, until other people make decisions and see what happens.
So I think there is more than one reason here.
DR. STARFIELD: I find your arguments very persuasive, but I don't agree with the conclusion. We can always build into a recommendation that this should be reviewed at such a time when there are other technologies.
To make a decision on the basis of, something may happen in the future when we really don't know what that would be or how applicable it would be to our needs I think is just missing the boat.
So while I agree with what you say, I think there are other ways around to do it.
DR. LUMPKIN: I think I would like to also comment here. Right now, when someone shows up in an individual practitioner's office, they either recognize that person or they go through a process of enrollment. In that process of enrollment, they associate certain information with that individual. But generally, the process of identification is an anthropomorphic analysis of their facial structures and the way they walk and all these sort of intangibles that allow us to recognize people that we know.
Then when that person comes in and enrolls at that hospital or that office, they assign a number of some sort. They can then associate the records when that person comes back with that particular number.
The difficulty is, when you are working as I have done in an emergency department in Hospital A, and a person comes in and says, I've had my surgery, that's why I've got this big humongous scar at Hospital B, and all you have is their name; you don't have a hospital number. Frequently it is next to impossible to get that medical information that will tell you what happened to them, even though it may be significantly pertinent to their care.
A unique health identifier will allow that to happen, because if you give that number at Hospital A, in that emergency department, that will facilitate the process of associating these two records for clinical purposes. I think there is a significant clinical rationale for having these numbers.
Now, in the future, the process of associating that individual with that number will change. Right now, we take personal history information and then we try to associate them with that number if they don't have that number with them. There may be a system in place in five or ten years where they walk up and you put a scan of their iris, and that iris will say, this is this individual, but it will associate it to some number. The data that is associated with any sort of physical characteristic is a huge chunk of data. It will never serve as an identifier. It will never serve as a way to link two records together.
So I think that there will be technological advances that will allow us to associate individuals with their number. But I believe that we will always have it as the basis, and now is the time to start to develop the infrastructure of relating one piece of a medical record with another, based upon some reasonably sized number that correlates those two pieces of data.
DR. GELMAN: Well, I agree with your analysis, but not with your conclusion.
DR. LUMPKIN: I've heard that before.
DR. GELMAN: Of course you're going to link records. Of course linking records clinically is a useful thing. No one is going to dispute that. But whether you need a number to do that remains to be seen.
I would like to offer one principle here, which I think for me is more important in making a decision in this area. That is, first, do no harm. Creating a new numbering system for health purposes will not be used for health purposes. There is no way that anyone can create a number and use it only for health purposes.
If you make a list of all of the institutions that will have access to the health I.D. number, and you can start with the providers and the payors and the employers, just to begin with, and all of the thousands and tens of thousands of institutions that there are, and then everybody else that is going to have it as well, government agencies, the IRS, credit bureaus, you name it, this is just people who function to keep the health care system working, the notion that you're going to have a number and that it is going to be used for health, as far as I'm concerned, is an impossibility.
If you create a new number, the world will beat a path to your door. This is what happened with the social security number. This is why the social security number is in the state it is in today, because as soon as the number was there and someone said, we can use it to catch welfare cheats, we can use it to control driver's licenses, we can use it to control immigration, as soon as those possibilities were thought up, the law was changed so that could be done.
That is exactly what will happen with the health number. That is one of the reasons I say I'm not prepared to concede we need a number yet. Some of these things may develop otherwise for their own purposes, but a health number cannot be controlled. It is a bad idea to allow something of this sort to be created if there are alternatives, and I don't think the alternatives have been adequately explored.
DR. SCHWARTZ: In the second paragraph, in parentheses, there is a referencing for another methodology that accomplishes the same purposes. So it is not clear to me -- my interpretation of the two paragraphs would not be that the committee is recommending a number, that a number is a possibility, but that something else is also a possibility. That is my interpretation. Perhaps that clarifies something.
DR. LUMPKIN: Simon?
DR. COHN: I actually just wanted to ask Bob whether or not -- I guess I am unclear of exactly his position, so I wanted to understand it a little better. My sense is that you agree with the first sentence in this document. I'm just not sure whether the issue is that you don't think it goes far enough, or do you even disagree with that.
DR. GELMAN: No, I agree with that. I would just go further. My position, just to be clear and simple is, I'm against everything. That is my position.
DR. FRAWLEY: Just to follow up with your example, maybe we need to keep focused on the language in HIPAA. We are talking about the nine transactions, excluding claims attachments. Your example in terms of linking clinical information in the ER is not quite the scenario we are talking about. We are talking about a unique health identifier for use in those nine specified transactions. My concern is that once those transactions have taken place, and knowing that the Secretary has already formally gone on record about data linkage and using non- identifiable information, I go back to the question which we don't answer: why do we have to have the unique health identifier at this time?
She is very clear on her privacy principles, what she expects. She laid them out. We have those nine transactions. We know exactly what has to happen in the delivery system in terms of those transactions. The question that I have is, after that encounter ends, the claim has been paid, all the benefits have been coordinated, I'm not entering the health care delivery system now; what do I need the unique health identifier for. That is why I keep raising the issue.
I think we have to look at the language, look at what the intent is, not talk about the computerized patient record, how nice it is going to be down the road, but the person who is in the medical records department, getting those frantic calls to find a record.
So I understand your situation, but I think we have to look at here and now, specific language in the legislation. I think we have to make sure we keep focused on that in our discussion, because we are going all over the place about emerging technology, the clinician and the ER.
The bottom line is, there are nine transactions, and that is what the health identifier is supposed to be used for.
DR. MAYES: That is not our interpretation. Our interpretation is that we are to develop an infrastructure and systems for four unique health identifiers which will be used by the nine transactions, but as with the security in the law are not restricted strictly to those nine transactions. It has to satisfy the needs for the nine transactions, but we don't read it as being restricted to just nine transactions.
DR. MOORE: That was my same point, that the intent of the law was to have a unique identifier that goes across all provider settings and all payors, so that there is one -- the individual is known the same in all those settings, rather than having separate identities.
And it does facilitate what John is saying, because in that same legislation it talks about the computerized patient record. It would appear -- I'm just making a leap here, that it appears to move in that direction eventually. I think that was some of the discussion within this committee back last year, when we first went over this legislation. I didn't think that intent had changed until today.
DR. LUMPKIN: I know there was an NCVHS before HIPAA, and there will be one afterwards. We took a position on the unique health identifier which had nothing to do with nine transaction sets. We took a position on health identifiers because there was a utility to that.
I don't believe that everything is black and white. One component of the medical tenet is, first, do no harm, but another proponent -- and this is a lesson I learned, which we called Rosen's Rule Number Two, never be afraid of using a life saving procedure for fear of the complications.
If in fact we have a procedure which is a unique health identifier, it has got bad features. But let's also identify that it has good features. Our challenge is to weigh the good and the bad and to say what is the characteristic of that, which way do we make a decision. But we also have to recognize that there is significant clinical utility and functional utility to having the ability to uniquely identify and make available to clinicians medical information in a timely fashion. Perhaps because I have practiced emergency medicine, I am acutely aware of the implications to individual people today, whether it be by giving medications that cause known reactions because we did not have the background information that says that they have allergies, or not knowing what prior treatments were happening. The lack of information is leading to unfortunate medical occurrences and death, even as we speak, somewhere in a clinical setting in America. A unique identifier can make a difference.
Now, we have to weigh the harm that an identifier can do and make our choice. But I just want to make sure that we understand that there is a balance here.
DR. COHN: I guess I'm getting a little confused. I don't believe anyone around the table believes that there is a problem with the issue of linkage. Though since we are really talking about this first issue about whether we think today is the time to identify specifically that linkage or whether we are saying it is premature, based on all these other things.
I'm not sure where you stand on that anymore. I thought that you were fundamentally in support of saying, yes, we need to move in that direction, but there are barriers called privacy and confidentiality that make it unwise and imprudent to move on that today, but that fundamentally we believe in the concept of linkage. We do need that clinical information you are describing, people need to be able to do it.
Now, is that what I am hearing from you?
DR. LUMPKIN: We kind of slipped into the second issue.
DR. COHN: I'm trying to keep them separate.
DR. LUMPKIN: Yes. It is probably my job to do that, but I didn't.
DR. COHN: This is a very tough area to discuss.
DR. WARD: I'm glad that Don left the easy decisions for you.
DR. LUMPKIN: I have ways of getting even. Further discussion? Before Bob makes his motion, what I would like to do is take just a straw poll, which is not a real vote, but just so we can see how many people feel in this committee that essentially the first statement that it would not be a good idea to have a unique health identifier without a legislative structure to assure confidentiality of individual identifiable health information and preserve an individual's right to privacy? How many feel that that should be a pre-condition to a unique health identifier?
DR. FRAWLEY: To the implementation of it, not the development; to the implementation.
DR. LUMPKIN: Right. That was my sense, that we are pretty much all on page one. How many people feel that if that protection is in place, that a unique health identifier is a good idea for individuals?
PARTICIPANT: Or a methodology that accomplishes the same purpose.
DR. LUMPKIN: Or a methodology that accomplishes the same purpose.
DR. GELMAN: No, those are two different issues. Those are two different issues. If you are talking about linking records, I don't know that anyone is against linking records. Your argument for linking records is just fine. The question is, how do you do it?
DR. LUMPKIN: The statement that we have before us is that we recommend that a unique health identifier or another methodology that accomplishes the same purposes.
DR. GELMAN: If you are asking, do people support linking records, either with an identifier or without, I think everyone will probably say yes, and that won't accomplish anything. You have not precisely stated a question.
DR. LUMPKIN: Well, actually, what I'm saying is that we have a statement before us. If you are uncomfortable with -- this is an A and B statement. It says a unique identifier or another methodology. If you are uncomfortable with that statement, then -- it is not a vote, just indicate no. Just so we can get an idea of how many are comfortable with that linkage of A or B, or do we have a problem that we need to perhaps refine and then come back.
DR. GELMAN: But let me make the point that it also says that it be selected on the basis of the ASTM criteria coupled with whatever. I'm willing to buy the first part of that. Do we have some way of linking records, unique identifier or something else? Sure. I'll vote no, because I don't agree with the rest of the sentence.
PARTICIPANT: Let's start with the first part.
DR. LUMPKIN: No, actually, I think Bob has clearly identified --
DR. FRAWLEY: His position.
DR. LUMPKIN: -- his position. Is there anyone who has doubt what are the two positions that we have here? PARTICIPANT: Say Bob's again?
DR. LUMPKIN: Which is that he is in favor of linkage, but he is not in favor of a linkage process that involves a unique identifier. Is that correct? Is that a fair statement?
DR. GELMAN: I'm not prepared to support that at this time, yes.
DR. STARFIELD: I guess part of the problem is that we don't know what the same purposes are. We don't know how to interpret the purposes.
DR. LUMPKIN: Well, no. If I can state this as clearly as I can, which is that -- one is, we are all in favor of linkages. I don't think that is an issue.
DR. FRAWLEY: For clinical care.
DR. LUMPKIN: But clinical purposes, right. We're all in favor of that.
DR. STARFIELD: For public health or just clinical?
DR. LUMPKIN: We'll debate later whether or not public health is clinical purposes. We are in favor of, for good reasons, for good clinical purposes, not financial --
PARTICIPANT: So we are not interested in linkages for fraud, waste and abuse?
DR. LUMPKIN: We're not going to talk about that right now. That confuses the issue. So there is the good of doing the linkage. One position is, we think that is a good idea, as long as it doesn't involve personal identifier, is that correct? Would that be a fair statement? Then the other part is, --
DR. GELMAN: Linkage is a good thing. I am prepared to concede that linkage is a good thing, period.
DR. COHN: Then basically you are in favor of this first part of the sentence.
DR. LUMPKIN: Okay. Is there anybody who is opposed -- we are in a discussion period. Then we are going to come back and we're going to actually vote on text. But I think it is going to be important for us to delineate and understand the points of difference in this group.
So the issue is, linkage -- no one disagrees, am I right, that linkage is a good idea for medical purposes? Okay. For health purposes. So the point of contention is whether or not we would say that that linkage should be accomplished with a unique health identifier or other methodologies selected on the basis of the ASTM criteria.
DR. COHN: Coupled with cost benefit and public input.
DR. LUMPKIN: No, I think the ASTM criteria is the crux of it. Help me out here, Bob.
DR. GELMAN: We heard this morning that the ASTM criteria are not all that clear, because there is no weighting between the different factors. So you don't even know what it is you are getting when you start talking about the ASTM criteria, is that right?
DR. EMERSON: That is correct, but I would like to make one other point also. The ASTM criteria are criteria from a universal health care identifier. When you start trying to look at other things that might accomplish the same thing, those criteria do not really match up very well as criteria to judge those other types of things.
DR. LUMPKIN: And we have three examples of what we are talking about. Let's get very specific. We are talking about an enhanced social security number, a different unique number, but which has some enumeration process, where that number is associated with an individual, or a methodology in which you have the unique number and a master patient index to help you associate an individual to it. Those are the three options that we are talking about, whether or not we think evaluating those three options for adoption in the context of confidentiality and privacy legislation is a good thing. Those who think that it is a good thing, raise your hand just so we can get a feel. This is not a permanent view; you can change it at any time. It is a pre-decision.
DR. GELMAN: And we are discussing that entire paragraph, or entire sentence?
DR. LUMPKIN: Pretty much.
DR. COHN: I guess I would say I would support it, though I would really like us to talk a little more about the ASTM criteria.
DR. LUMPKIN: We are just trying to deal with the concept of this sentence.
DR. GELMAN: I'm not necessarily prepared to say that the search should be limited to the three items that you have described, although I don't have any objection to evaluating them.
DR. COHN: I actually thought that we were going to divide this up into two concepts. One is, once we get the privacy and confidentiality, are people willing to go forward. Then the other piece is, what are the criteria and what are our suggested top runners.
DR. LUMPKIN: Okay.
DR. COHN: I thought that was going to be conceptually a little bit easier.
DR. LUMPKIN: First, if we get it, are we willing to move forward with the unique health identifier of some sort.
DR. COHN: Or another methodology.
DR. LUMPKIN: Or another methodology for purposes of linking health stuff. Raise your hand. We are not reporting it, so we just have a feel for where we are at. Then, if we are in favor of it, most people at this point seem to be in favor of it, should it be based like it says in the second part of the sentence.
DR. MAYES: Just as a point of information to remember, the ASTM is an ASNE accredited SDO, and these criteria were a balloted standard from an ANSE credit. So that reflects some of the other constraints that we had in the legislation.
DR. LUMPKIN: That's correct.
DR. MAYES: And the process is an open consensus process.
DR. GELMAN: And there were lots of people that weren't represented in that process.
DR. MAYES: But it was an open consensus process.
DR. GELMAN: I understand that, but they still weren't represented.
DR. LUMPKIN: But I'll just point out that in this whole discussion, we don't seem to be bound by the charge of the legislation, so why start feeling constrained now?
Okay. Having said that we think moving forward once the legislation is passed is a good idea, do we think that moving forward in the way that is outlined in the last part of the sentence is a good idea? Raise your hand if you think it is.
DR. COHN: Which means using those three criteria or those three criteria and these three frontrunners? I would divide them again.
DR. LUMPKIN: Okay, we are going to divide the priorly divided and divided question. Using those three criteria, do we think it -- if we are going to go ahead, do we think it is good to go ahead, using those three criteria? Yes?
DR. FRAWLEY: What are the three criteria?
DR. LUMPKIN: ASTM, cost-benefit, public comment. Do we think that is the methodology we should use? Raise your hand. This is just getting to where we are trying to get. Opposed? I understand that some of you may have dropped out because you don't agree with the contention, but this is a straw poll, so it doesn't matter.
Finally, should we tell the public that we think there are three frontrunners, as listed there?
DR. GELMAN: Three frontrunners, and no more.
DR. LUMPKIN: Is anybody opposed to saying that -- on the leading candidates, willing to accept comment on additional candidates? Or other candidates? And we can provide the list of candidates that are in the list of our commissioned paper, and say that of those, these three are the leading candidates, but we are going to accept comment on everything.
DR. GELMAN: And other things yet to be identified.
DR. LUMPKIN: And other things yet to be identified, yes. Anybody opposed?
DR. GELMAN: Actually, I do have a question. I wrote this down. I guess I am still myself confused, whether the VA hybrid is a leading candidate. I've just got to say that. I don't know how we want to represent it. That would be my own personal comment. It is fine for us to put it in there, but I certainly didn't come away from the HHS recommendations feeling that there was great clarity.
DR. STARFIELD: And it wasn't part of our contractor's report, either. I don't think it was addressed in our contractor's report.
DR. LUMPKIN: My suggestion would be, if we take the -- we will just pull numbers out of the air and associate them with the various candidates that we have. Let's just suppose we will pull like the number 123 out in different locations, because there are no public documents that we can refer to, but we may want to pull some numbers out.
I think it is just fair to let people know what is ranking up high and what is low. That doesn't mean that something that is low couldn't be the accepted candidate. I think that is where we are coming at.
DR. FRAWLEY: I think what we were trying to get to when we were talking in the subcommittee last night is, there might be some industry experience already with one of these candidates, and we just haven't had the benefit of the input. Somebody could come forward and say, I have been using this, here is our experience to date. But we don't even have that before us. It is kind of like going through and trying to pick something, and not quite sure what you are picking.
DR. LUMPKIN: So something to the effect, cost- benefit evaluation and public comment on candidates, indicating a ranking of the currently existing candidates in our commissioned paper. A tentative ranking, very tentative.
DR. COHN: So what are you trying to put here, John?
DR. LUMPKIN: What I am trying to get to is, do we reach some level of comfort if we say in there, instead of saying we have got these three candidates, and that is all we really want to hear about, to say that we have got this whole laundry list of candidates, but let's at least let you know, by using the ASTM criteria, there are some that rank higher than others.
DR. COHN: I just wanted to bring it up as an issue that we should just talk about. The wording that we come up with, which is three and other candidates, might be fine verbiage, though you are beginning to say that we also need to somehow in something like this state that we will go out and do some of these things also.
DR. LUMPKIN: My guess is that we reach some level of comfort on what it is that we need to do. Let's think about it and see if we can try to put this down on paper, along with our other recommendation we need to vote on.
(The meeting adjourned for lunch, to reconvene at 1:45 p.m.)
DR. LUMPKIN: The first item on the agenda was a motion to select different music for the interim. I'm from the Midwest; I don't quite understand why these issues should come up, but perhaps our distinguished colleague from Massachusetts -- or one of our distinguished colleagues, one of our many distinguished colleagues -- do you have a proposal?
DR. IEZZONI: I would like to move actually on the letter having to do with the payor I.D.
DR. LUMPKIN: It has been moved and seconded that we send the letter to Secretary Shalala that says what it says right here. Since you guys don't have a copy, I guess I'll read it.
On behalf of the National Committee on Vital and Health Statistics, I am pleased to forward to you our recommendations relating to another of the health data standards being proposed for adoption in concordance with the administrative simplification provisions of the Health Insurance Portability and Accountability Act of 1996.
The NCVHS is very pleased to provide support and advice and consultation to you in this effort. The NCVHS has been briefed on the proposal for the national standard for identifiers for health plans or payor I.D., all in caps, and we offer our strong support.
The proposal includes a nine-digit numeric identifier that would be assigned to al health plans. The identifier includes a check digit and contains no embedded intelligence. We recommend that HHS proceed to establish the proposal for public comment without delay.
In the interests of operational efficiency and simplification, we suggest that the department also leave open the option of moving to an alpha numeric identifier in the future. While public comments are likely on the technical details on the number and the optimal approach to numeration, we have found broad support for the proposal in general, and urge you to proceed.
The committee did identify one concern that we bring to your attention, the all in caps payor I.D. as proposed replaces the plan I.D. and sub-I.D. used in current transactions. The sub-I.D. is currently used for electronic routing by clearinghouses, and they have expressed concern that this function will be lost. We recommend that this functionality be considered before the final draft is issued.
We appreciate your leadership in health data standards, electronic data interchange and privacy, and we are privileged to work with you on these issues.
Signed, Sincerely, Don Dettmer, chairman, on behalf of NCVHS.
It has been moved and seconded. Is there discussion on the letter?
DR. COLTIN: I just have a point of clarification. I believe that the sub-I.D. is used for electronic routing by plans as well as clearinghouses. Not all claims go through clearinghouses. I think some of the larger plans also do use it more directly. So I'm not sure that this should be a clearinghouse issue solely. So just take out the "by clearinghouse", and "concern has been expressed that".
DR. FRAWLEY: Can you read that sentence again, please?
DR. LUMPKIN: Yes. It would say, the sub-I.D. is currently used for electronic routing and concern has been expressed that this function will be lost.
DR. GREENBERG: And the recommendation is that the functionality be considered before the draft is published for comment?
DR. LUMPKIN: No, final draft is issued.
DR. GREENBERG: If it is final, would it be a draft?
DR. LUMPKIN: No, what we are saying is that -- we are not saying hold up the notice of proposed -- NPRM, but before the final rule is issued.
DR. COLTIN: I would actually like this to be a little bit stronger and say, be addressed before the final rule is issued. Considered is a little weak. I would like to at least know, if they are not going to do it, why.
DR. LUMPKIN: Anybody opposed to doing that, addressed? Because if it is addressed, then we can route it, right? Any further discussion? All those in favor, say aye.
(Chorus of Ayes.)
DR. LUMPKIN: Opposed? Abstentions? Carries. That brings us back to the slightly shorter document that you have in front of you.
That brings us back to the slightly shorter document that you have in front of you. Let me read this one. While the NCVHS continues to support the concept of a unique health identifier for individuals, we believe it would be unwise and premature to identify and implement such an identifier in the absence of legislation to insure the confidentiality of individual identifiable health information, and to preserve an individual's rights to privacy.
I think we would put this all in one sentence, so it wouldn't be a separate sentence, so just that one change. Once the public is assured of this protection, we recommend alternative methods of identifying patients and linking health information of individuals be evaluated on the basis of the ASTM criteria, coupled with the cost-benefit evaluation and public comment, period.
That is the reinstated motion that is before us. It has been moved and seconded. Discussion on the motion?
DR. FRAWLEY: One thing we talked about before the break was for health purposes.
DR. LUMPKIN: For health purposes?
DR. FRAWLEY: Yes, we had a phrase.
DR. STARFIELD: The trouble is, you have to define what you mean by a health purpose. We talked about this before. I don't know that it adds anything. I don't have any objection to it.
DR. FRAWLEY: It is just that we wanted to make sure it wasn't for commercial purposes.
DR. LUMPKIN: We are going to get real structured here. There is a proposal to amend this identifying linking health information of individuals; for health purposes, be inserted after individuals and before evaluated. Is there a second to that?
DR. STARFIELD: Second.
DR. LUMPKIN: It has been moved and seconded. Is there any opposition to that? We can just insert that, then.
DR. GELMAN: I just want to say, I oppose the first sentence that says NCVHS continues to support the concept of a health identifier. I don't agree with that, for reasons that we have talked about before.
The second paragraph, there is almost a disconnect between the two. To say that we support an identifier in the absence of legislation, which I certainly agree with the sentiment there, but we may not have legislation for 20 years. We have no idea when legislation is going to pass.
Then to say that once legislation passes, that we know on what basis we are going to evaluate a health identifier doesn't seem to me to make any sense. In 20 years we may have better criteria, in ten years, in three years, whatever. But to decide the standard for decision in advance and what may be substantially in advance seems to me to make no sense. I am certainly not prepared personally to say that the ASTM criteria, which I think are certainly of some relevance here, are the only criteria to be used to evaluate things.
So I have to vote no on this thing.
DR. LUMPKIN: Further discussion?
DR. STARFIELD: I guess I'm not sure whether this motion, statement, allows us to proceed to evaluate the criteria, even in the absence of the legislation. I don't know why we have to be held up from developmental efforts, as long as it is not implemented.
I think we can just cross out that first phrase in the second paragraph. We recommend alternative methods be evaluated but not implemented until --
DR. LUMPKIN: There is a motion to delete the opening phrase, once the public is assured of its protection, so that the line will read, we recommend alternative methods. Is there a second?
DR. COLTIN: I would like to propose an amendment to that motion, because I think it goes hand in hand with it.
DR. LUMPKIN: If it has not been seconded, --
DR. COLTIN: It is a friendly amendment.
DR. LUMPKIN: I'm going to be easy. Okay, go ahead.
DR. COLTIN: In the second line in the first paragraph, take out the words identify and, because it is the same concept.
DR. LUMPKIN: I'm sorry, say that again?
DR. COLTIN: Premature to proceed to, and now it says identify and implement. I would say, if you are going to make that second change, you should also take out the words identify and, and just say proceed to implement.
DR. STARFIELD: Where are we?
DR. GELMAN: End of the second line and beginning of the third line.
DR. LUMPKIN: So to be consistent then, the amendment would be to delete the word identify and in the first sentence, and in the second sentence to delete the opening clause.
DR. COLTIN: Right.
DR. LUMPKIN: Is there a second for that? It has been moved and seconded. Is everyone clear on what the amendment is? Is there discussion on the amendment?
DR. COHN: I'm trying to think what comment I want to make about it, because we are beginning to change the whole concept of what we are talking about here. I think we are trying to say that it is fine to go ahead and identify things now, even though we want these other things to be in place before they are implemented. Obviously, a very reasonable person could say, gee, isn't that already in the law? Why are you even writing this letter? Why aren't you just writing a letter to the Secretary saying we are in the process of identifying?
DR. GREENBERG: The law permits regulation as a substitute for legislation.
DR. LUMPKIN: The law mandates us to recommend an identifier.
DR. COHN: Right, and so why are we -- I think what we had been saying is that we thought it was premature to identify and implement prior to having legislation in place. Now we are suddenly saying that -- I think it is fine to evaluate.
DR. COLTIN: Which is what the second paragraph -- DR. COHN: Yes, but I'm not sure that we need to knock out the identifying for selecting.
DR. COLTIN: There is an ambiguity about the word identify. It may be broader than select and may include evaluate. So I think what we are saying it is premature to do is to actually select one, as opposed to evaluate.
DR. COHN: Yes, and I agree with that.
DR. COLTIN: Then you change identify to select?
DR. LUMPKIN: To select? So it would be premature to proceed to select and implement?
DR. COHN: Fine.
DR. LUMPKIN: Are we comfortable with that?
DR. COHN: What is the change in the second sentence?
DR. LUMPKIN: The second sentence, we would delete the initial clause, so it would begin with the word, we recommend. That is the second part. The health purposes is already added in there. We did that earlier.
DR. COHN: I guess the question is at that point, we basically begin to seem a little silly, this whole sentence, because what we should say at this point is that we want you to know that we are in the process of evaluating alternate methods, because that is what we are supposed to be doing, but that we haven't reached a conclusion yet.
DR. LUMPKIN: I think actually, that clause is not crucial to the statement. There is nothing in this that would prohibit us from identifying candidate identifiers, but we urge that they not be selected or implemented until the law passes.
DR. COHN: Right.
DR. LUMPKIN: So that would allow us to do work on this issue without selecting.
DR. COHN: So the point of this sentence needs to be that we want to assure that we are in the early processes, we want to tell you that we are in the early processes of evaluating methods or whatever? Is that what we want to communicate to the Secretary?
DR. COLTIN: I think there are two points that we were trying to get at with the second paragraph, in addition to the notion of what we have just talked about, which is that work could proceed on this one, without the legislation being in place the evaluation portion.
One is that we used the term alternative methods of identifying to be more all-encompassing than simply saying a unique identifier. We are saying look at methods to accomplish these purposes, which may or may not include the unique identifier. Then we are adding in this idea of a cost benefit evaluation and public comment, which we didn't feel were adequately addressed in the work thus far.
DR. COHN: Okay, I guess I agree with the conception. I am just trying to think of this -- rather than a recommendation, I think this becomes more of a comment on how we are planning to proceed, and we are already in the business of doing this. This is what we are describing what we intend our work plan to be?
DR. STARFIELD: We don't have to recommend it.
DR. LUMPKIN: We have an amendment, which would change the word identify to select.
DR. COHN: I thought we had already passed that one.
DR. LUMPKIN: And in that same amendment would delete the clause, once the public is assured of its protection.
DR. WARD: What we are trying to get to is, what is the intent of that communication. The intent of the first paragraph is very clear. What have we accomplished with the second paragraph, I think is what Simon is struggling with and I am struggling with. What is the value that that sentence has been sent forward?
DR. GREENBERG: One might be that this is to say what the committee is planning to do. As Simon says, maybe you don't need to say that here, although it might be informative. But I don't know that the only evaluation of this alternative, and of all sorts of alternatives, should be done by the committee. I think we have already agreed, a lot of this is going on in the private sectors. I think it is more than saying this is what the committee is planning. That is how I read it.
DR. WARD: So the intent is that we are encouraging continued investigation of what will be some good methods that will be available to us when this legislation passes. That is what we are trying to say in that sentence.
DR. MOORE: I hesitate to recommend anything, but can we change the word patient to individual? Because we are looking at people who are not patients yet.
DR. LUMPKIN: Sure. I will rule that out of order, but we'll do it anyway.
DR. HARDING: As usual in these situations, I am ambivalent. The first sentence, I happen to not -- I think in the big picture, I don't support the idea, but I understand that we are probably going to vote to support the idea of the unique identifier.
My only thought would be that we are -- instead of saying we continue to support, to just answer the question as requested by Congress, we are processing, we are doing these things as was requested, to come up with that. It doesn't matter if we support it or not. It is not our decision. It is that we are asked to give the best we can to detect that or to pick that out.
That is how I would start that paragraph. I don't have the wording and so forth, but I just wondered -- if that is not an issue to anybody else, then --
DR. GELMAN: Why don't we dispose of what we are dealing with, and then move on to another thing?
DR. LUMPKIN: Okay. The motion is to delete the phrase once the public is assured of its protection, so that it will read -- the second sentence will read, we recommend alternative methods of identifying individuals and linking health information. Is there further discussion on that motion to amend? All those in favor, raise your hand.
DR. COLTIN: Is this pre-decisional?
DR. LUMPKIN: No, this is decisional. I count eight. All those opposed? I count two. That phrase is now deleted. Are there additional amendments? Bob.
DR. GELMAN: I move we strike the first clause in the first sentence.
DR. LUMPKIN: There is a motion to strike the first clause in the first sentence, which reads, while the NCVHS continues to support the concept of a unique health identifier for individuals. Is there a second? It has been moved and seconded. Is there discussion? I think we have had certainly a fair bit of discussion. All those in favor of the amendment, raise your hand, which is to delete that initial clause. All those opposed? I count three. The amendment fails. Further discussion?
DR. COHN: I was just going to ask, because I didn't get a sense of why people wanted to have that one knocked out. I guess I was trying to figure it out while the vote was going on, so I abstained.
DR. LUMPKIN: I suspect we have already voted.
DR. COHN: That's fine, if we have already had the discussion. That's fine.
DR. LUMPKIN: Are there other discussions or amendments? Are we ready to vote on the main motion?
DR. MOORE: Could you read it in its final form before we vote?
DR. LUMPKIN: Yes, I will. While the NCVHS continues to support the concept of a unique health identifier for individuals, we believe it would be unwise and premature to proceed to select and implement such an identifier in the absence of legislation to insure the confidentiality of individually identifiable health information and to preserve an individual's right to privacy. We recommend alternative methods of identifying individuals and linking health information on individuals for health purposes be evaluated on the basis of the ASTM criteria, coupled with a cost-benefit evaluation and public comment.
DR. GELMAN: What is this? Assuming that this is approved, what are we doing with it?
DR. LUMPKIN: It would be my understanding that this would then be communicated to the Secretary as the position of the committee. I suspect we would do a letter that would say, on behalf of the National Committee on Vital and Health Statistics, I am pleased to forward to you our recommendations relating to the health data standards being proposed for adoption in accordance with HIPAA. NCVHS is very pleased to provide support, advice and consultation -- that is our usual boilerplate in the front. Then we would say this statement, and we would end with, we appreciate your national leadership in health data standards, electronic data and privacy. It is a privilege to work with you on these issues.
DR. GELMAN: I want the opportunity, if the letter goes from the committee, to submit dissenting views. I want to be able to submit dissenting views to the Secretary on behalf of those who don't agree.
DR. FRAWLEY: Point of information.
DR. LUMPKIN: Yes.
DR. FRAWLEY: If we vote on this, are we ever going to consider this proposal?
DR. LUMPKIN: What is --
DR. FRAWLEY: That Bob had presented earlier. I'm just confused. Just a point of information on what we are doing.
DR. LUMPKIN: I think if we vote on that, then given the boilerplate that you have before you from the previous one, this language will be substituted in there, and that would be the letter that would be sent. I don't know, not having been on the committee for any length, if there is a minority opinion, how that would be handled.
DR. GREENBERG: Bob had suggested earlier, if you voted on the proposal as you just read it, inserted it into such a letter and passed it, would it still be in order for Bob to propose that his letter be substituted.
DR. GELMAN: I had not planned at this point to do so, mostly because I don't think it will have support.
DR. GREENBERG: Okay. You had said earlier that that was what you intended to do.
DR. LUMPKIN: But my personal opinion, unless other members of the certainly would disagree, I think that certainly we feel that if there is an issue, then certainly this is the issue, where members of the committee feel strongly. I would not have a problem with individual members signing on a minority letter to be sent to the Secretary.
DR. GREENBERG: I would agree. My only hesitation at this particular time is that the Chair of the committee - - I don't think that this has ever been done before. That doesn't mean -- but there is absolutely no rule that it can't be done. I personally don't have any problem with it. But since it is precedent setting, I guess I would be a little more comfortable if we could discuss it.
DR. LUMPKIN: My suggestion would be that this be raised with Don. If he has got a problem with that, then we come up with a mechanism for us as a committee to discuss that.
DR. FRAWLEY: That wasn't what I was asking. What I am asking is, there are points in Bob's letter or proposal that many of us on the committee like. My point was, why can't in the letter that we send to the Secretary incorporate some of the things in Bob's proposal that we like to enhance our position? I'm not particularly interested in a dissenting opinion. I think most of us feel comfortable with what we have come to. But the point is that there are a lot of good points that Bob has raised in his proposal.
My question was, do we have to defeat the motion in order to discuss Bob's proposal, so that we can then take Bob's proposal and roll it back into the letter to the Secretary. That is all I want to know, as a point of information.
DR. LUMPKIN: My suggestion is that since we are voting on the boilerplate in the beginning and the nice stuff at the end with this in the middle, if you believe that there are chunks of Bob's letter that ought to be in there, that you would make an amendment to include those at this time, before we take a final vote.
DR. FRAWLEY: I so move.
DR. LUMPKIN: Can you tell us which ones?
DR. FRAWLEY: We never discussed it. That is what I'm trying to figure out, what do we discuss this document.
DR. LUMPKIN: But I would ask, rather than trying to put in the entire letter and then have us argue on what to take out, that you would say what you would like to include from his letter.
DR. FRAWLEY: Well, actually what I wanted to do was move for consideration of the proposal. We have never discussed it. It was handed out to us and we started to wordsmith this motion. So all I'm asking is, when are we going to be able to discuss this two-page document without being out of order.
DR. LUMPKIN: We have a committee report. If you would like to bring up this letter, which was submitted as a recommendation as a separate item on the agenda, we can do that after we make this vote. But if you would like to include language from this letter into the current motion, once we vote the motion, we will have to bring it back, which means that we have to reconsider, we have to take a vote to reconsider it. So now would be the time to take language out.
DR. COHN: John, I think what I am hearing from Kathleen is almost that she would like an opportunity to discuss this, so that we can identify what should come out. I think the only way to do that would be to move substitution of this as the new one, which I think could be done at this point.
So I presume, Kathleen, if you move substitution, then this whole thing can be discussed, and then we can go back and figure out what pieces go in.
DR. FRAWLEY: Fine, then that is what I move.
DR. LUMPKIN: No, you are amending by substitution, which would be to delete everything that we currently have. That is by substitution.
Let me give you a couple of options that we can do this parliamentary. Let's pause. Let's take up this later. Are there things in this letter that people like? Then shall we start with the first paragraph and walk through and see if there are things that we like? Okay. I knew there would be a way that we could find to do it.
Can we skip down to the starred sections, perhaps? Okay. First section, first paragraph. The selection of a unique health identifier for individuals will become the focus of tremendous public attention and interest, far beyond that afforded to other, more important health privacy decisions. No choice should be made without considering more public notice hearings and comment.
Anybody have a problem with that?
DR. STARFIELD: It is just a restatement of what we have said, I think, perhaps a little more elaborated. I think we already said that.
DR. LUMPKIN: Next paragraph. Until a new federal law adequately protects the privacy of identifiable health information, it is not possible to make a sufficiently informed choice about an identification procedure or number, number or procedure. The degree of formal legal protection for personal health information will have a major influence on both the decision and public acceptance of that decision. Passage of a comprehensive health privacy law may make the choice of an identifier easier and less threatening to privacy.
DR. STARFIELD: It supports what we say, I think.
DR. LUMPKIN: So we put that in. The next paragraph. Changes in technology are creating new identification options, including digital signatures, biometric identifiers and public key encryption. The identification alternatives that have been proposed to date are mostly based on known technology. Until new alternatives have been developed and tested in the marketplace, selection of the patient identifier is premature. Any selection made in haste today could become obsolete in a few years. The result would be a significant increase in health care costs in the long run. If a patient identifier is necessary, it must be certain that the identifier remain useful for the foreseeable future.
DR. MAYES: All the pieces are in the evaluation. They weren't selected, but all three of those technologies stated as being new technologies are included in the evaluation today.
DR. VAN AMBURG: The problem with this paragraph is that technology is always changing. If you continue to wait, you'll wait forever on this. The facts are right, that there is technology that needs to be evaluated, but we can't wait until forever to do this.
DR. LUMPKIN: Is there anyone who feels strongly about keeping this in? The case for a single health identifier for individuals remains unclear. It has been shown that health records can be linked with a high degree of accuracy, using a combination of non-unique identifiers. In addition, advances in computer technology and networks are creating other methods of linking health records without the need for a common number. These are not necessarily positive developments for privacy protection, but they suggest that a single identifier may not be as important as previously thought.
I see some heads shaking. Anybody want to include this?
DR. GELMAN: Yes, I do.
DR. LUMPKIN: I wasn't going to pick on you, Bob. Anybody want to include this other than Bob?
DR. COHN: This somehow relates to that second paragraph that we had. Maybe it isn't a bullet point. Maybe it needs to be somehow woven into this place where we recommend alternative methods of identifying whatever we said, and linking health information be evaluated. This could be used in some ways to flush that out, some wordsmithing. It probably needs to be removed as a major bullet.
DR. FRAWLEY: I agree. I think there are some thoughts in here that are helpful, but I wouldn't necessarily incorporate the entire bullet.
DR. STARFIELD: It seems to me it could be incorporated into a caveat that anything that is recommended can change in the future as new technologies are developed. In other words, we might two years from now come up with a proposal for a unique identifier, but recognize that this isn't case in concrete forever.
DR. LUMPKIN: Let me ask this question. I think g certainly, it is implied in our recommendation that we are not saying that it should be a number. Do we feel the need to make this statement?
DR. COHN: I guess I would propose that we don't put that in as a bullet, but maybe after this process, after we get the bullets that we agreed to, then we can do wordsmithing with the other concepts along with this, and present a complete proposal. So I'm saying, let's forget it as a bullet for the moment, but recognize that most of us think it has some good ideas in it.
DR. LUMPKIN: Okay. The next bullet. A unique health identifier for individuals cannot be properly protected from misuse under current law. The committee reaches this conclusion, notwithstanding enactment of criminal penalties for wrongful disclosure. As part of HIPAA, additional legislation may be required to authorize the use of some alternatives or to provide adequate restrictions for other alternatives.
DR. STARFIELD: It provides justification for what we have said.
DR. LUMPKIN: Anybody dislike this?
DR. FRAWLEY: I don't object to it.
DR. LUMPKIN: At this time -- the next bullet, at this time, the choice of an identifier cannot be made solely based on technical or economic criteria alone. The decision has a significant political component. The public's elected representatives should participate in and direct or ratify any choice through the regular legislation process. The choice should not be made administratively.
Are you going to be in order this time? Go ahead, Bob.
DR. MOORE: The regulation that would go up from the Secretary would go through OMB and to Congress, so Congress would have to approve of the regulation in order for it to be adopted. So it would not be one that would be just solely administrative.
DR. GELMAN: That is pretty much, propose a regulation and wait to see if they pass a law to overturn it. That is not the same thing.
DR. MOORE: No, but they have 60 days to non- concur with that reg, if it is above a certain -- and this would fall in that category. So if they are silent, then I am assuming they concur.
DR. GELMAN: Silent doesn't put them on the record.
DR. LUMPKIN: But the question is, given that the regulatory process allows legislative oversight, but that is not what this sentence says. This sentence says Congress should pass it as a law. Is there anyone other than Bob who agrees with that? I saw a lot of heads shaking. Then we will take this out.
The next sentence I think we have already voted on, the therefore sentence. So I think that covers all the points.
I think that what I would like to suggest for a procedure is that we take the first bullet point, the second bullet point, the fifth bullet point and directly incorporate those into the letter. We would take the fourth bullet point, which is on the different technologies and whether or not it needs to be a number, and try to smith it into that second sentence. We add on our standard bullet point, the stuff in front and back. If we vote this, we will then produce a letter. That letter will then go back to the committee for a veto. In other words, if you take the letter you will have seven days or ten days to say no, this isn't exactly what we agreed upon.
DR. GREENBERG: We should also give Don an opportunity to see it.
DR. LUMPKIN: Give Don an opportunity to see it, and if that period of time happens, the letter will be sent. But what we would be voting on would be a letter that would be an amalgam of three things. I am not asking you to vote on the letter that would be signed. What I want you to do is a process by which you get another chance to see it before it goes out. But it would be an amalgam of the boilerplate in the front and the end, our statement in the middle, and the three and a half bullet points from Bob's letter.
Why don't we take a -- how long will it take to get it printed up? A 15-minute break? Because this is the last item on our agenda.
DR. STARFIELD: No, no, we actually have another one left over, the guidelines.
DR. GREENBERG: Yes. I revised them.
(Brief recess.)
DR. GREENBERG: I changed the rules to guidelines. Actually, I had originally had ground rules, so we are pretty much back where we were, to guidelines. I pretty much took out the word pre-decisional since it raised some questions, but it now reads, NCVHS members receiving draft documents from the department. It doesn't say pre- decisional draft documents -- and I added, that are not for public distribution, we will keep them in strict confidence and not discuss them. That stays the same. I then said access to such draft departmental documents, taking out the word pre-decisional, will be coordinated. That stays the same.
DR. HARDING: Could you put something in there, that these documents will be clearly marked?
DR. GREENBERG: Well, I was going to put something like that in there. I guess just from a logistics point of view, there could be a cover note to it, saying this is pre- decisional. Sometimes we get documents that we don't have electronically. I think as a procedure, I intend as executive secretary to try to make sure now that all documents that are of this nature will be properly marked. I am committed to that. But exactly how it was going to be done I hated to put in the rules.
DR. LUMPKIN: I think the difficulty with that is, those of us who participate on listservs as part of participating in internal committees, I think we have to assume that those documents are all not for public discussion.
DR. GREENBERG: Right.
DR. HARDING: Or just say the opposite, say this one is all right or something.
DR. GREENBERG: I would say, when in doubt, assume it is not for public -- and it hasn't been said it is for public distribution, assume it is not.
DR. LUMPKIN: I think it would be appropriate for us to ask first. What we would like to encourage as a committee is for us to continue to participate in the way that we have under HIPAA with other internal departmental committees. It adds significantly if each document that may go on the listserv has to be identified if it is on an internal listserv.
DR. GREENBERG: Also on this second one, I think I took out the word pre-decisional. That was pretty minor. That was it, I guess. I think that was it.
DR. LUMPKIN: That was it?
DR. GREENBERG: Yes.
DR. LUMPKIN: Okay, so we'll wait for Bob to come back.
DR. GREENBERG: Just let me make one last check here. Yes, that was it.
DR. LUMPKIN: Any discussion on the guidelines? I think Bob if he were here would point out the futility of having guidelines. So should we take a vote? All those in favor, raise your hand. Opposed? It carries, with one vote against.
DR. GREENBERG: Then just to remind you, if you had any comments on the informational documents that Elizabeth circulated, if you could get those to me say within a week, and then we'll finalize those. We'll get them out to you. The two-pager on the HIPAA is really very similar to the fact sheet on the website, except it is geared toward the national committee's role, whereas that comes later in the overall fact sheet.
The other one, on the status of NCVHS recommendations, I would suggest once that is finalized that it should go on the website, because it nicely summarizes -- actually, we have to add the recommendations that were passed here for that matter, so we can't finalize it without that.
DR. LUMPKIN: At this point, while we are waiting for them to finish the document, we usually at the end of the meeting invite those in the audience, if they have any last statements. So I would like to open that up at this point.
PARTICIPANT: In the context of guidelines for keeping decisional information closed in, you mentioned, John, earlier how the states are simply unfamiliar with the implications of this coming down. What sort of planning is going to be considered for really disseminating this information?
Clearly, the normal modes of dissemination aren't working, except for the little colleges that communicate on specific issues, but the larger implications for the people affected by it, health care provider organizations around the country. They as well as the state entities simply aren't receiving information in a way that they can assimilate it and recognize what their interests are.
So I just pose that as a question, because I haven't heard that much discussion in all these meetings, in terms of a concrete methodology. Here in the United States, we assume there is news media, there are professional associations, et cetera, et cetera. But I think as you mentioned isn't working, at least at this stage.
DR. WARD: Speaking for that group, what do we call ourselves? The public information work group, that is exactly what we have been trying to address. That is why we had to produce and okay those two documents, because those are the documents that will not only be up on the website, but will be starting to be -- we are collecting long lists of organizational names and we are working with the media people at HHS to get those continually sent out, in addition to controlled contacts with the media. We are trying to get more and more people on convention conference activities, but that is what we are trying to do.
That doesn't mean that necessarily everyone is responding to that. But we do have the limit in terms of the number of -- we don't have media staff for this committee, so we are having to add that to the burden of the media staff in the department. But they are at this point willing to work with that, and to add to what it is they have already got planned to continue to pressure the communication.
But that was the whole point of these two documents that you reviewed today, was that those will be what those summary -- can be one and two-page documents, continually sent out to those organizations.
DR. MAYES: If I might add, on the federal side we too have a public education awareness group. We have had in place an outreach plan, as we move from the phase of standard identification to standard implementation. We are in the process of revising that plan now.
There has been actually a very significant amount of outreach. I would say we probably meet with two to three groups every week around the country, different staff, and give presentations. But we are also working with the press departments of several of our agencies as well.
DR. WARD: And those groups actually work together.
DR. MAYES: Yes, and we are liaisoning with the committee's group.
DR. MOORE: There is another issue here, and it is not just that they get the knowledge. It is that there are groups and individuals as well as organizations, including HCFA in some parts, who do not accept the fact that this legislation applies to them. They are aware of it, they know what it means, but they don't believe that it includes them.
DR. LUMPKIN: I have friends who can talk to them in Chicago. They are very persuasive.
For those of you on the Internet, this is described as a pregnant pause.
DR. COHN: The organization is going to follow up on the educational issue, since we have got a second to talk about it. Certainly one is impressed that the information doesn't seem to be getting out very well. One has to wonder why. And certainly, as I think about it, part of the issue is that nobody has any idea what effect it is going to have on them. So everyone's assumption is, it is not going to affect me one way or another.
I was actually asking Dr. Braithwaite a question, does this apply to workers' comp, since that is a state activity. Most of the people involved in occupational health don't have any idea of this, don't think that anything federal has anything to do with them ever, since it is only a state involved activity. They haven't been communicated with.
Now, part of the problem gets to be, we really don't have any scenarios of how this plays out or what effect it might have on various groups so we can target and identify them that are likely to be affected, as well as communicate to them what the effect might be.
Now, I don't know if that is going to be part of the HHS plan or part of our plan or what. But as long as we are talking about this high level of standards, you're right, nobody is going to be very interested.
DR. MAYES: In our presentations, we try to tailor the presentation to the audience. I personally, when I do presentations around the country, try to pull out as many examples of how I think it is going to affect -- I have spoken to the national immunization program and talked about setting up immunization registries.
Two weeks ago I was in Chicago talking to a clearinghouse program, where I had providers and payors. We went into a great deal of detail. There were two days of discussion as to what their problems are now, which will be impacted by this legislation.
So we're not just going out and saying there are these standards coming down, you should look at it. I think that we have been trying to tailor our presentations to particularly those groups that at first glance say, this is back office stuff, this doesn't really apply to us.
There is an interesting group actually that is going to come to the forefront with some recent proposed changes in Medicare. We now have certain provider groups that are going to become plans. Guess what? They are suddenly going to go from being a group that has options as to whether or not to comply with the group that is mandated to comply with all of the standards, including their having to communicate electronically.
So there are some real interesting groups that we still need to target more specifically. But we have tried that. We continue to try and hone our message. In fact, one of the very things that the work group that is putting together the outreach plan is trying to do is just that, is not try to come up with just a single press release that we send out to everybody, but rather identify significant sectors of the industry that we can tailor the response to and go towards what you are saying, Simon; here is an area that you may not have thought, here is a scenario that might impact you beyond what you might have read.
DR. LUMPKIN: But I think we can look at the experience with Y2K issues. There was certainly a key community which understood that these had broad ranging implications, but it has taken the general business community years to recognize that this is going to have an effect on the bottom line.
So I suspect there is going to be a delayed oh my God response. Until they get somebody they know and trust to say this is going to affect everything you do, most people are just going to say it is another thing happening in Washington.
We now have a copy of the letter. Before Bob gets a copy of it, I want to take a vote.
On behalf of the National Committee on Vital and Health Statistics, I am pleased -- this is to Secretary Shalala, of course -- to forward to you our recommendation relating to another of the health data standards being proposed for adoption, in accordance with HIPAA.
The NCVHS is very pleased to provide support. While the NCVHS continues to support the concept of a unique health identifier for individuals, we believe it would be unwise and premature to proceed to select and implement such an identifier in the absence of legislation to insure the confidentiality of individually identifiable health information, and to preserve an individual's right to privacy.
The selection of a unique health identifier for individuals will become the focus of a tremendous public attention and interest, far beyond that afforded to other, more important health privacy decisions. No choice should be made without considerably more public notice hearings and comment. Until a new federal law adequately protects the privacy of identifiable health information, it is not possible to make a sufficiently informed choice about an identification number or procedure.
The degree of formal legal protection for personal health information will have a major influence on both the decision and public acceptance of that decision. Passage of a comprehensive health privacy law may make the choice of identifier easier and less threatening to privacy.
A unique health identifier for individuals cannot be properly protected from misuse under current law. The committee reaches this conclusion, notwithstanding enactment of criminal penalties for wrongful disclosure as part of HIPAA. Additional legislation may be required to authorize the use of some alternatives or to provide adequate restrictions for other alternatives.
It has been shown that health records can be linked with a high degree of accuracy, using a combination of non-unique identifiers. In addition, advances in computer technology and networks are creating other methods in linking health records without the need for a common number. These are not necessarily common elements for privacy protection. But they suggest that a single identifier may not be as important as previously thought.
We recommend alternative methods of identifying individuals and linking health information of individuals for health purposes be evaluated on the basis of the ASTM criteria coupled with a cost-benefit evaluation and public comment.
We appreciate your national leadership in health data standards and electronic data exchange and privacy. We are privileged to work with you on these issues.
Sincerely, Don Dettmer.
DR. HARDING: Are we reopened for comments?
DR. LUMPKIN: We are. This is now the motion in front of us.
DR. HARDING: This is the motion? I have one friendly and one change. In the third paragraph, the word far beyond that afforded to other, quote, more important, I would ask that more important be taken out, because I'm not sure what that --
DR. LUMPKIN: Okay. Anybody feel strongly about that? It's done.
DR. HARDING: It is just making a judgment. The other is coming back to the first phrase in the second paragraph. What I would suggest is that it be changed to a sentence in that first phrase, and it would be this, and it is different. The NCVHS is responding to the request of Congress to recommend a unique health identifier for individuals, period. Then capital W, we believe, and then it would continue on. So it is changing it from the NCVHS continues to support, to, the NCVHS is responding to the request of Congress to recommend. It is not saying whether we agree or disagree or whether we support it or not. Just, we are doing our duty as a good soldier to respond to the request that we were given. Therefore, that to me is a concern that we continue to support. I'm not sure I do. I am a good soldier, and I will respond to the request of Congress to make a recommendation.
DR. LUMPKIN: It has been moved to amend that opening phrase with a sentence which would read, --
DR. HARDING: The NCVHS is responding to the request of Congress to recommend a unique health identifier for individuals, period, and then capital W.
DR. LUMPKIN: And then, we believe. Is there a second? It has been moved and seconded. Is there a discussion?
DR. STARFIELD: I think there is no other way. You have said there are possibilities, I guess, but I don't know of any other way, and we certainly haven't discussed any other way. So until there is another way, there is a purpose that needs to be served, and I think that is why we support the unique identifier, although we are certainly willing to change that at the appropriate time.
DR. LUMPKIN: Further discussion?
DR. VAN AMBURG: I don't support the recommendation, but we do have a problem, because we are saying up here in the second paragraph, we continue to support the concept of a unique health identifier, and then down here at the bottom, we are saying there are other methods, and we really recommend you explore these methods. So that is an inconsistency that has to be resolved somehow in this letter.
DR. STARFIELD: Am I correct that we haven't identified any of these other methods?
DR. GELMAN: I think we can link recorda without an identifier, without a unique identifier. We do it today. It is a matter of degree.
DR. VAN AMBURG: That's right. It is a probability match, and they are not necessarily good matches all the time.
DR. GELMAN: And they are not necessarily going to be good matches with the unique patient identifier, either.
DR. VAN AMBURG: Well, I think you have a better chance.
DR. STARFIELD: The probability method has not been used for the kind of data -- some of the kinds of data we are considering. They have been used for population based data, not for clinical data.
DR. COHN: I was just going to comment that I actually support your amendment. I think it doesn't add anything, having that phrase at the beginning, especially with this additional paragraph that we have down there now. So I think it sounds fine.
DR. LUMPKIN: Other comments? I would like to toss in a --
DR. COLTIN: I would like to second that notion of the administrative purposes. I can't imagine a payor who would be willing to use a probabalistic mechanism. It is 90 percent likely that this is the person covered under your benefits, so pay this bill.
DR. LUMPKIN: The additional point I would make, first of all, this committee has gone on record doing that, and so we would be in a sense saying that that is not our position.
Second, it is my belief that while there are other methods of linking records, if you decide to link records based upon name, address, mother's maiden name, those three things now become our identifier. Instead of having a number this long, you have got an identifier that long.
I think that in our interest of protecting privacy, we now get to the point where the only person who doesn't have access to this health information are the people who need it most to provide services. Yet those who are disingenuous or for whatever financial reasons can use these, you give me three or four characteristics, I can find any record in the world type of individuals, will always have access to that.
But it is like when we put the childproof covers on medications, and the kids could open it but the seniors couldn't. Sometimes I wonder if we are doing too much to protect. So I think I would be opposed to that change.
DR. HARDING: If I could comment back to the Chair, I'm not saying that we should reverse. It is a neutral statement. I am saying that we are responding to Congress' request. I'm not saying we are for it or against it, but that we are responding. That was the purpose, is to leave that out of it, whether we are for it or against it. We are just responding to the order that we were given by a Congressional bill.
The purpose was to neutralize that sentence to not state that we are against it or for it, but that we are responding -- we have been given an order and we are in the process of making a response.
DR. LUMPKIN: I don't know if this is helping at all, but as I look at this, I am wondering -- certainly a true statement is that the NCVHS has historically supported the concept of a unique health identifier. I am wondering if that is the modification that we need to make to it. We either get rid of it, or we say has historically supported the concept and, comma, we believe. That certainly neutralizes it, but it also indicates our historical support of that, without making any current comments.
DR. SCHWARTZ: I would just like to raise an issue in response to that for clarification. My understanding of the common data element report has -- in the report, the first element is a unique identifier. My understanding is that there is text in there that says something to the effect that the committee suggests consideration of the social security number or some other identifier generated from other variables, something to that extent.
So this is a current report of the committee. It would seem to me that the committee is continuing to recommend or to approve of the concept of a unique identifier. That is my understanding, for whatever it is worth.
DR. LUMPKIN: We have before us the language that was suggested, that would say that in response to Congressional demand, we are --
DR. HARDING: The NCVHS is responding to the request of Congress, or maybe it was a demand, I don't know, but a request of Congress to recommend a unique health identifier for individuals, period, and then capital W, we believe.
DR. LUMPKIN: Further discussion on that amendment?
DR. STARFIELD: I think it is antithetical to the role of an advisory committee to respond as a good soldier. We are doing things because we believe in them and not because we are directed to do it. I just think it is a bad precedent to say in terms of what we do in the future.
DR. LUMPKIN: Are we ready to vote?
DR. COHN: If you could read it one more time.
DR. HARDING: The NCVHS is responding to the request of Congress to recommend a unique health identifier for individuals, period, then capital W, we believe.
DR. LUMPKIN: Those in favor, raise your hand. Those opposed. Abstentions. Fails. Are we ready to vote on the letter? I guess not.
DR. VAN AMBURG: It seems the only way you can fix that right now is to eliminate at least most of that paragraph that begins with, it has been shown.
DR. LUMPKIN: If I can respond, I don't think that is inconsistent. It just says that there are other ways of linking records, but we recommend you use the ASTM criteria.
DR. STARFIELD: No, but there aren't any other ways that we know of. Not clinical records, no. There just aren't any.
DR. COHN: Not for administrative purposes, but there are other ways of linking records.
DR. SCHWARTZ: I'm going to move that that paragraph be eliminated.
DR. LUMPKIN: Okay, there is a motion that, -- yes, up until, we recommend. So it would be the first three sentences of that paragraph. The motion is that those would be stricken, removed from the document. Is there a second? It has been moved and seconded. Is there discussion on that?
DR. COHN: I guess I just want to understand a little better, because I'm not sure that I see there is a -- it doesn't look to me like there is a conflict, so I guess I need to understand better from you the conflict between the concept of a unique identifier and the fact that there is a capability of linkages.
DR. VAN AMBURG: Well, there is a capability of linking records for research purposes and there are probability matches. This is primarily an administrative use. As Kathryn pointed out, that isn't necessarily going to fly for administrative purposes.
It appears that we are saying in this paragraph up here, we support the unique identifier, and down here we are saying, we're not really sure anymore, you ought to explore all these other alternatives because there are other ways of linking records. I don't think that is a consistent position.
DR. STARFIELD: I wonder if there is a compromise position that we might say, that advances in computer technology may create other methods. These may not be necessarily positive developments, but they should be considered as they arise.
DR. VAN AMBURG: We have that in that last sentence we left in anyway, Barbara. It says we recommend alternative methods for identifying individuals.
DR. STARFIELD: Fair enough. I guess I am thinking about past the point at which we might recommend an identifier past that point, that there is a possibility of reconsidering at some later date. But maybe we don't need to say that.
DR. LUMPKIN: If I could suggest some alternative that might accomplish the same -- if you look at those first three sentences and only change the third sentence, and change it to say these are threatening developments for privacy protection, period. Then we note that there are all these computer things. The fact that people can link records with a single number enhances the risk to privacy. Therefore, we recommend the following. But I think it takes out the inconsistency -- never mind. We have in front of us a motion to delete the first three sentences from the paragraph.
Seeing no further discussion, can we take a vote? All those in favor of the motion to amend by deleting the first three sentences, raise your hand. All those opposed? Abstentions? The Chair votes positive. The motion carries.
DR. COHN: What was the vote?
DR. LUMPKIN: The vote was six in favor, four opposed, one abstention.
DR. GREENBERG: Can I make a friendly suggestion? This paragraph as it remains that says we recommend, et cetera, on the basis of the criteria developed by the whatever that ASTM stands for, the American Society of Testing Materials.
DR. LUMPKIN: Can we give a specific citation for that.
DR. GREENBERG: Yes, because people that will be reading this won't know what it is.
DR. LUMPKIN: Right, we'll give a specific citation for that. That is an editorial.
DR. SCHWARTZ: Can I suggest another friendly consideration? That is, in the first line on the second page, instead of saying, and linking health information of individuals, say perhaps, linking individually identifiable health information.
DR. LUMPKIN: Are you on which paragraph?
DR. SCHWARTZ: First line of the second page of the letter, which begins, alternative methods of identifying individuals. I am suggesting it should continue to read, and linking individually identifiable health information for health purposes, as opposed to linking health information of individuals.
DR. LUMPKIN: We'll consider that editorial. Wouldn't something that links -- for example,you could take diagnoses and link individual identifiable health information by diagnosis?
DR. SCHWARTZ: You could be linking health information of individuals by actually pooling data and linking aggregated data. I don't really -- that opens up that possibility. I don't think that is really what this committee is suggesting. But the other is more definition with respect to the level of the data. It is individual level. That is my opinion.
DR. LUMPKIN: I see some confusion. Does anyone want to make that as a motion?
DR. STARFIELD: Make what as a motion? Yes, I move that we change it.
DR. LUMPKIN: Can you read how it would read again? It has been moved and seconded. Alternative methods of identifying individuals and linking --
DR. SCHWARTZ: Individually identifiable health information for health purposes, be evaluated on the basis of the ASTM criteria coupled with the cost-benefit evaluation and public comment.
DR. LUMPKIN: Is there discussion on that motion to amend that language?
DR. COLTIN: Yes. I don't find that that clarifies it for me, because I can imagine two data sets, both of which contain individual identifiable information, but which are linked on variables other than the identifiers. Whereas, what this is trying to get across is that they are being linked on the basis of the identifier, that all of the information belonging to an individual is being linked.
That is what confuses me about the change. I think it makes muddier rather than clearer.
DR. LUMPKIN: Okay. Any other -- it has been moved and seconded. Those in favor of the suggested language, raise your hand. Opposed? Fails. Abstention, I'm sorry? Two abstentions. Fails.
DR. COHN: I guess I actually had two questions. One is, in the very first paragraph and very last paragraph, the introduction seemed very odd in light of what we are saying, as well as the conclusion. It says, I am pleased to forward to you our recommendations related to another of the health data standards. The NCVHS is very pleased to provide support, advice and consultation to you. I think that we probably want to say -- first of all, this is not just another health data standard.
DR. LUMPKIN: Rather than wordsmith this, is it acceptable that we make this introductory have the right tone, considering the nature of the characteristic?
DR. COHN: Okay, that sounds great.
DR. GREENBERG: I realize that Richard was recommending it as a substitute, but you could almost incorporate his concept that this is one of the standards that this included in HIPAA, and this is the way the committee is addressing it.
DR. LUMPKIN: If it is agreeable, that is fine. We have made a number of changes. We'll send this out to everybody before it goes out, so they have a chance to see that.
DR. COHN: Can I ask one other question?
DR. LUMPKIN: Please.
DR. COHN: This is just a question for the committee. In here, there is nothing that we are saying we are doing. There is no action that has anything to do with the committee in this. It says, we recommend against this, we recommend -- are we doing anything? Are we pledging to be involved with any part of these recommendations? If so, it isn't here. Are we recommending alternative methods of identifying individuals, and if so, what part are we going to play in this? Or are we just throwing it over the wall to implementation teams and saying, go ahead and do it?
DR. MAYES: The implementation team has made their recommendation. They presented it to you, and we will follow through. We have to present it to the Data Council and that sort of thing.
I guess my question was an extension of that, Simon. You had indicated that for the hearings, et cetera, et cetera, that should be held. We are glad to take those and to listen to those and take that into consideration and possibly modify it. But is there any schedule for this activity? We still have the requirement to go forward by the legislation and propose this. But we would like to continue to work on developing this idea, but I'm just curious if there was going to be any formal schedule for hearings, et cetera.
DR. LUMPKIN: Perhaps we can include in that last sentence something along the lines of, the committee intends to continue to receive public comment on this issue and revisit this at our next scheduled meeting in November.
DR. WARD: We make the point with more public notice hearings and documentation should occur. I am assuming we are going to participate in that, because we feel that strongly.
DR. LUMPKIN: Well, we had better say it.
DR. STARFIELD: I just have a question. You said something before about vetoed. Did you mean that somebody could veto this when you send it out for comment? You didn't mean veto, did you?
DR. LUMPKIN: By vetoing, someone could say, this isn't what we said. Then if in fact someone says, this isn't what we said, then we need to figure out either a conference call or some way to adjudicate that. We have to change the first paragraph. That is really what we are looking for, for people to say yes, that's okay. If everybody is fine, we're just going to say, we are forwarding our recommendations on the unique health identifier, period, boom, that's it. That is the first paragraph, under HIPAA.
DR. STARFIELD: Then Simon gave some final language.
DR. LUMPKIN: Yes, and the committee intends to continue to receive public comment. It may not be testimony. And we will revisit this issue at our November meeting. Do we all understand what we are voting on?
DR. SCANLON: One clarification. In the process of -- can we go back and read that, what you are saying here, without waiting until a week from now?
DR. LUMPKIN: No, actually the committee says we are going ahead. There is no week. We have reduced the first paragraph to saying we are sending our recommendations on the unique health identifier for individuals under HIPAA, and then we go into the first paragraph, and then we add the thing about, we are going to continue to receive public comment, that will be added to the last sentence.
DR. GELMAN: If we are all done, I just want to make a quick statement before we vote.
I just want to state for the records my principal objections to this letter. First, I object to the decision at this time to support a unique health identifier for the reasons I stated before. Secondly, I object to the endorsement of the ASTM criteria at this point. I think that is a premature judgment. I think the criteria are not clear. Third, I continue to believe that this is a political issue that requires a vote from elected representatives before a decision is actually made.
DR. LUMPKIN: Thank you. At this time, we will take a vote. All those in favor, signify by raising your hand. All those opposed? Abstentions? Passes with two opposed. I think we're done. The agenda before us has been completed.
(Whereupon, the meeting was adjourned.)