[This Transcript is Unedited]
Quality Hotel Courthouse Plaza
1200 N. Courthouse Road
Arlington, Virginia 22201
DR. LUMPKIN: The business of the committee is before us in the agenda. The meeting will commence. Why don't we start off by doing introductions? As before, yesterday, we are going live over the Internet, so people can recognize our names.
My name is John Lumpkin. I am Director of the Illinois Department of Public Health and chairman of the committee.
DR. SCANLON: I'm Jim Scanlon from HHS. I'm the staff director for the full committee.
DR. HARDING: Richard Harding, University of South Carolina.
DR. IEZZONI: Lisa Iezzoni, Beth Israel Deaconess Medical Center in Boston.
DR. STARFIELD: Barbara Starfield, Johns Hopkins University.
DR. NEWACHECK: Paul Newacheck, University of California, San Francisco.
DR. ZUBELDIA: Kepa Zubeldia, Envoy Corporation.
DR. COHN: Simon Cohn, Kaiser Permanente.
DR. BLAIR: Jeff Blair, Medical Records Institute and member of the committee.
MS. WARD: Elizabeth Ward, Foundation for Health Care Quality, Seattle, Washington.
DR. FITZMAURICE: Michael Fitzmaurice, Agency for Health Care Research and Quality and liaison to the National Committee on Vital and Health Statistics.
DR. ROTHSTEIN: Mark Rothstein, University of Houston.
DR. GELLMAN: Bob Gellman, privacy and information policy consultant.
DR. FRAWLEY: Kathleen Frawley, St. Mary's Hospital, Passaic, New Jersey.
DR. GREENBERG: Marjorie Greenberg, National Committee on Vital and Health Statistics, CDC, and Executive Secretary to the committee.
DR. FRIEDMAN: Dan Friedman, Massachusetts Department of Public Health and member of the committee.
MS. JACKSON: T. W. Jackson, staff for the National Committee on Vital and Health Statistics.
MS. WILLIAMSON: Michele Williamson, staff, NCHS.
MR. SANDEL: Steve Sandel, Centers for Disease Control and Prevention, staff to the committee.
DR. HORLICK: Gail Horlick, Centers for Disease Control and Prevention, staff to the committee.
MR. FANNING: John Fanning from the Department of Health and Human Services.
MR. ISMELDA: Don Ismelda, American Health Information Management Association.
DR. ELLINGER: Stan Ellinger, ADC Biocare Research and Quality and staff to the Quality Subcommittee.
DR. KINSMITH: David Kinsmith, JCPR for a new project and Veterans Administration, staff to the subcommittee.
MR. FURIL: Richard Furil, staff to the Subcommittee for Standards and Security.
DR. LUMPKIN: Welcome. I did want to report on one item. Kathleen Fyffe called me early this morning and said that she was called into the office, even though she intended to be here. I guess an issue that came up despite our best efforts not to be involved in the Presidential campaign, evidently the Gore-Lieberman campaign yesterday issued a press release and stuff on the website on medical privacy, and had an open forum, and I guess Kathleen got involved in figuring out what their response is going to be to that. So she was called away, and wanted me to convey that.
The first item on our -- I would like to comment however that the position of the committee seemed to be compatible, although we don't like to get involved in those kind of things.
The first item on the agenda is a briefing on the Institute of Medicine report, Institutional Review Boards, Health Services Research, Data Privacy Protection. Dr. Lee Zwanziger is here from the Institute of Medicine. For any concerns and issues which Lee can't handle and which we get particularly upset, we'll be able to direct it to Lisa. I understand she was also part of the committee that put forward this report. I just didn't want her to get away without being duly recognized.
DR. FITZMAURICE: I might mention, John, that the study was funded and supported by the Agency for Health Care Research and Quality and the Office of the Assistant Secretary for Planning and Evaluations Office.
DR. LUMPKIN: Thank you. Lee?
DR. ZWANZIGER: Thank you, Dr. Lumpkin, and thank you, members of the National Committee on Vital and Health Statistics. I appreciate this opportunity to address the committee.
As you heard, I'm Lee Zwanziger. I am from the Institute of Medicine. The Institute of Medicine recently released a report by our Committee on the Role of Institutional Review Boards and Health Research, Data Privacy Protection. This committee was chaired by Dr. Bernard Lowe, and the report that was issued is called Protecting Data Privacy in Health Services Research. The full text of the report is available on the National Academy Press Website. I am going to leave flyers that give the whole URL, so any members of the public who are present can easily get access to that report. My contact information is also on the flyer. I hope that anyone would feel free to contact me if you either had trouble getting the report or had any other questions.
One of my first things that I love to do here is give some acknowledgements. As staff for this committee, I can say that this committee was outstanding to work with. In particular, I am delighted to be here today with Dr. Lisa Iezzoni, whose insight into health services research lent so much to the committee. We also hopefully will have access to the legal and policy expertise of Jan Goldman by telephone a little bit later, depending on her schedule back in her home office.
Of course, like any report, this report could not have been done without the support of our sponsors. As Dr. Fitzmaurice just mentioned, that was Agency for Health Care Research and Quality and the Office of the Assistant Secretary for Planning and Evaluation. Since the sponsors' contacts, Dr. Fitzmaurice and Mr. Fanning, are here today, I hope that you gentlemen will both feel free to join in the briefing and the discussion whenever you have things to add or to emphasize.
While I am thanking people, I would also like to thank this committee. Various members of the committee and staff of the committee have been very helpful in the development and carrying out of this project, both directly and by the discussions you have had and your attendance at other workshops, where I was fortunate to hear your remarks.
The history and the policy context of this report is certainly as well or better known to all of you than to me. There have been various privacy bills and rules about when it shall be considered acceptable to release health information. Frequently if not always, those rules and policies have assumed that such release will be acceptable if an IRB had looked at the protocol in question and deemed the release adequately protective of the human subjects involved.
The question is, are IRBs in a position to do that, and are they currently supported at a level where they can carry out that role? After discussing this, this was initially seen as a question, if a possible source of concern, by our sponsors, who then raised the question with the Institute of Medicine. After assessing the questions and perhaps questioning the assumptions behind some of these rules about what IRBs may be currently doing, the IOM and our sponsors agreed on a charge for the committee of gathering information about current practices as well as principles, but important to notice practices, of IRBs to protect confidentiality and to wherever possible identify the ones that seem to be superior in efficacy, to do the same for organizations that don't have IRBs, because they are doing not research, they are doing things that are similar that do not come under the common rule, because they are not funded under federal auspices, or are not presented to agencies such as the FDA that require everything to be so covered, even if it isn't federally funded, and finally, if appropriate, to recommend a set of best practices for protecting confidentiality.
This has been a relatively short project. We have had to start running and move quickly to get done by our due date. We had our initial meeting by telecon; we did a lot by telecon actually, with a published number for a public call-in, which many members of the public did take advantage of and lent helpful suggestions at that initial stage. That was back in January of 2000. We also posted a call by listserve to IRB professionals and other interested parties to submit practices and suggestions that they were using at their institutions. Some of you may have actually been on one of my listserves and gotten it, I don't know.
We held a public workshop with representatives of academic and other research institute IRBs, health services researchers and pharmacy benefit managers. We had representatives of integrated health systems and commercial IRBs to discuss practices that they were using, wish they could use, problems that they were having, and discussed that with the committee and with other members of the public who wished to participate. That was in March of 2000.
We commissioned two background papers which are available as appendices to the full report, one emphasizing special issues of confidentiality where the subjects of the data are minors. That was by Ross Thompson, a developmental psychologist, and the other on international standards of privacy. That would be both standards of privacy that are applied internationally as in the EU, and an international comparison of different standards of privacy. That was done by a lawyer from Montreal, Barthe Goffiers.
After our workshop, we held a couple of private closed meetings for deliberations about what the committee's products should say. We issued -- our first product that we issued was a summary of the workshop. That is yet another appendix of the final report. Our final report was issued in August.
Last week, the 11th, the Chair moderated a followup symposium on what should be next steps after this report that was followed by a lively discussion. I thought that was very helpful actually to give people some feedback. We can't change the report, but we didn't really think this report was going to finish the problem, so it was nice to see where people thought we might go from here.
Just some key points in the committee's deliberations, at least from the point of view of the staff. Privacy and confidentiality are serious concerns, certainly to health services researchers, certainly to persons in the industry and the academic departments concerned, but also to the general public. They may or may not be well or technically understood, but they are certainly matters of concern to citizens in many walks of life.
We already have examples of technical and procedural practices that can significantly enhance privacy protections, but also allow good research to proceed. We do indeed want the fruits of that good research to enhance our health care system.
These enhancements don't come for free. They will require investments. Every kind of IRB I heard from in the course of this report are stretched to the absolute thinnest breaking point of what they can do on how little sleep, and with how little staff and how little resources. So we do feel that to the extent that we really are committed to privacy and confidentiality and certainly we seemed to hear from many members of the public that many of us are very committed to that, we do need to support our institutions that will help us enhance privacy and confidentiality.
Things of this report include that -- we did try and look for specific concrete practices, things that actually make the work get done, get done well and get done in a timely manner, to fill the gap between principles, which are of course very important to start from, and the day to day work of getting things done.
Examples of practices are some heuristic lists of what kind of data are identifiable, how can data sets be safely combined to protect privacy while the results of the combination can be used.
We heard from many sources, as I mentioned, private sources, private industry as well as academics, that it is important to have a genuine and thorough institutional commitment to confidentiality, both something that is instantiated in policies and procedures throughout the institution, also something that is an expression of an honestly accepted institutional culture that respects privacy and confidentiality.
That would include support of IRB practices, policies that favor confidentiality, and frankly policies that publish improprieties, such as unauthorized accessing of medical records, much less unauthorized publicizing of medical records. This is certainly not something we think is a normal feature of research, but something that occasionally happens, and if it occasionally happens like once, then it freaks people out for participating in health services research in instances far beyond the original unfortunate incident.
We as I said do need resources to be invested in the IRB process. We found that as I'm sure many of you know, many IRBs work on a totally voluntary basis without having additional time granted to them from other duties, and of course that has a cost in how well the IRB is able to carry out its duties. Certainly I think people are rising to the occasion with as much commitment as they can muster, but they do in fact have to be able to support their labs and carry out their regular duties.
Other kinds of resource commitments that we would suggest that IRBs would benefit receiving, and their institutions and research would also benefit from such receipt, is investment in a technical infrastructure such as computer professionals and computational facilities. We are talking about generally merging very, very large databases, scrubbing large amounts of data, and the health services researcher him or herself may not be in a position of expertise or time to do that, but it benefits certainly from IT professionals who would be able to help if such were available.
Finally, IRBs may need access to other sorts of expertise, either on the committee or through accesses on a consultancy basis, in certain kinds of special topics. This came up with studies where the subjects of the data are minors, so that the IRB may benefit from having a consultant who is particularly sensitive, in the sense of being an expert in dealing with the psychosocial risks of harm or wrong from possible inadvertent or improper disclosure that would affect children, how that might affect children differently at different stages in their life, and how that might affect other family members who are not involved.
Many of these kind of risks can certainly be mitigated by building in some care at the front end, and we just need to make sure that IRBs have access to the expertise to help them do that.
Finally, health services research of course overlaps with many other activities that are similar in many respects to research, except they may have a different goal. They may be -- instead of general knowledge for the sake of general knowledge, they may be activities undertaken to improve the quality and delivery of health services for a particular institution or business. The subjects whose day is being used in these services also, although they are not research subjects, they certainly still, the committee felt, deserve protection from possible inadvertent or improper disclosure.
So the committee felt, and argued strongly, that any sort of research or research-like activity would benefit, the PI would benefit, and most importantly, perhaps the subjects would benefit from having some sort of an outside look at the protocol to make sure that any possible protections that could be reasonably implemented were in fact implemented. This might be an IRB. In other cases, it might be something sort of like an IRB, but not exactly, would be under IRB regulations, but in any case, would relieve the PI from what might otherwise be an unfortunate position of deciding for him or herself whether the research or the other project was to be exempt from further view. It is always, we felt, helpful to have an uninvolved eye to take a look at the protocol for that purpose.
If I can answer any other questions, I'd be very happy to, and I'm sure that Dr. Iezzoni and Dr. Fitzmaurice and Mr. Fanning would also be very helpful here. I don't know if Ms. Goldman has joined us. We haven't heard anything, probably not. But I could also carry your questions to her and get back to the committee's staff on that.
Thank you.
DR. LUMPKIN: Thank you. Lisa, anything to add?
DR. IEZZONI: Nothing at this point.
DR. LUMPKIN: Okay. Bob?
DR. COLTIN: I'm just wondering if in the course of your study, you looked at potential issues of IRB liability for decisions that they made or didn't make, or for overseeing researchers, or not doing it, or anything like that.
DR. ZWANZIGER: I think that is a fabulous question. No, we did not look at IRB liability. We did occasionally touch in passing on how the IRB should not be left hanging to dry, particularly for what might be seen for a shield for an institution that hasn't supported it in the first place. Several IRBs protested that problem, of being castigated for having let something through that they may not have seen or may not have even had time to look at.
So the answer is no, but we wish we could have.
DR. COLTIN: There was a story in the Post yesterday about the Jesse Gelsinger case, the kid at Penn who died as a result of genetic therapy. The interesting thing on that was, they sued the ethicist in the case. So the answer is, somebody is getting closer.
DR. LUMPKIN: In the business area, I mean, in tech law field, is that an ethical thing to do?
DR. COLTIN: Suing somebody is always ethical.
DR. FRIEDMAN: First of all, I think the report is very helpful, and particularly helpful in encouraging IRBs to develop some health services specific research, and to differentiate between health services and clinical research in the applications and the forms.
I did have a question about the first recommendation, or what is listed here as recommendation 3.1, which says that organizations should work with their IRBs to develop specific guidance and examples on how to interpret key terms in the federal regulations pertinent to the use in health services research of data previously collected for other purposes. Such terms include generalizable knowledge, identifiable information, minimal risk and privacy and confidentiality. It goes on to say that organizations and their IRBs should make such guidance and examples available to all investigators submitting proposals for review.
My question is, what was the thought of the committee on why this should come from individual IRBs, rather than for example from OPRR or some other central source?
DR. ZWANZIGER: I'll take a stab, and then other parties can add.
One of the things that we ran into time and time again, and it is certainly not specific to this area of consideration, is that the central guidance, central rules, central education, central examples, are all hugely, hugely helpful. I think they are always, but certainly they are usually. There comes a point when you can't predict every single thing that is going to come up. At some point, there is going to have to be some local judgment as to how do we apply such and so regulation or such and so definition in this particular case.
So since we cannot do it all the way down to the individual practice level from the central source, I believe the committee is trying to emphasize here that the local source has to begin to start thinking ahead, as well as of course thinking judiciously on an ad hoc basis, but think ahead a little bit about what shall we give our PIs as guidance about what is generalizable knowledge, for instance.
So it is really not meant to preclude guidance from the central human protection office, but to emphasize that it always has to be supplemented on the local level.
DR. IEZZONI: I think it was as straightforward as that. You can't anticipate everything. But it is a useful exercise actually for the IRB to go through, to think through what these terms mean. It was as straightforward as Leslie has suggested.
MS. WARD: I was wondering if you think there will be an effort to follow up to see if in a couple of years, the recommendations have been put into place.
DR. ZWANZIGER: I'm in an organization that works on contracts, and so that is going to depend on whether there was a desire to do that.
DR. LUMPKIN: You were looking to your right when you should have looked to your left.
DR. FITZMAURICE: I think our purpose for initiating this project with ASPE and having the IOM do it is that, while a lot is known about human subject protection in clinical trials, not an awful lot is known about how IRBs protect the privacy of health services research data.
That has become more and more important in the past two or three years as Congress periodically debates four or five different privacy bills. When it comes to research, they say, as long as the research has gone through an IRB review, we're going to bless it, we're going to say, that's all right. The Department's Notice of Proposed Rulemaking on privacy, again, the same blessing, as long as it was reviewed by an IRB or a privacy review board.
We became concerned that we didn't know enough about how IRBs were protecting information. We didn't know what the best practices were, and there wasn't enough funding to go out and do a comprehensive survey, even if you could identify all the IRBs in the country. So we had to go to the most august body that we knew, to say, give us some advice on information we can distribute to IRBs, to the health services researchers, on how best to protect this information and protect human subjects that the information is about.
They would come back with a report that says, you have to be aware of the resources that are needed, the expertise that is needed, and the understanding that is needed in order to protect this information.
We think this is a great first start to help push the state of the art in this protection and health services research, but it is only a start. I would focus not on, did they follow the rules in two years, but what the next steps are, what is it we need to do to improve the protection of this information. It would help if we knew what the privacy regulations said in other areas than research, but that is where I see us going, is to build on this, rather than to wait for a time and say, did anybody look at it and did anybody pay attention, and are they following some of the suggestions.
John may want to speak to this as well.
MR. FANNING: Obviously it is a subject of continuing interest and concern, and we are not prepared to say now, we're going to do this or look into that particular thing.
However, note that in our notice of proposed rulemaking for our confidentiality regulation, we did say that we would re-examine the rules for protection of research subjects with a special eye to the confidentiality requirements there.
The second point to be noted is that the former Office of Protection from Research Risks that administered that regulation and looked out for this issue, has been elevated to an Office of the Secretary unit, now called the Office of Human Research Protections. I have spoken with Dr. Kosky, who has been appointed the head of it, and he said the confidentiality of information is one of his priorities.
DR. LUMPKIN: Jim?
DR. SCANLON: I'm interested not so much in the traditional IRBs where you looked at the practices and the policies; I was interested in what you may have found when you looked at research or research-like activities that were sponsored by others or carried out in other organizations, health plans and so on, that are not subject to federal regulations, and what sorts of procedures if any did you find, not necessarily in terms of an IRB, a full IRB, but at least some sort of review policy, a confidentiality review board, some sort of a third party look at the disclosure policies underway.
I saw you have at least one health plan, or two. They may be the best. What was your sense of the situation?
DR. ZWANZIGER: Our sense of the situation was that the variation is wide, very wide indeed, from responses on the order of very genuine expressions of, we really care a lot about our customers, we certainly don't want to hurt them at all, but we haven't really thought about this. Two, organizations that already had extensive -- as you have seen -- written policies and boards in place that reviewed all kinds of things, all kinds of studies or study-like things that were done.
So I think it varies quite a bit, as you would expect, as things are evolving gradually in different organizations without necessarily a central push. So I would say that we were also very impressed, or at least, the staff here was impressed, I think the committee was impressed, with the level of commitment and in some cases the willingness to share information about what the organization in question had found to be helpful, and how they set up review of research or research-like projects, and in other cases, how open the organization might be to talking about their much more informal processes.
Something that does not make it to the workshop was conversations I had with organizations that said, yes, we've been thinking about it a lot, yes, they would talk to me indefinitely over the phone, but no, they were not going to talk in public because they were afraid that if anybody heard what they were trying to do, then they would get dinged in public about not having done it yet, or they would be afraid that if anybody heard what they had done and were successful with and were proud of, they would get dinged for not having done it sooner or not having done more.
So they were in fact another example of how we can sometimes shoot ourselves in the foot by not encouraging people who were trying to come along instead of hollering at them for not doing it sooner.
DR. SCANLON: Were there any industry -- in terms of industry self governance? Again, I'm not talking about academic research. Were there any of the associations that have tried to make some headway in this area in terms of recommendations and policies that institutions could use? Or did you not really find any? Joint commission, AAHP or places like that?
DR. ZWANZIGER: There are a number of different industry groups that have issued different kinds of recommendations. We tried to refer to and briefly summarize reference to several of them.
I was a little bit concerned about doing that in too much detail, because in most cases my impression is that the principles are developed with a view to a particular problem that somebody is having at that time, or some particular thing they are trying to accomplish for their industry or improvement within their industry, or their profession. So they may give emphasis in a way that they would not choose to emphasize if they knew it was going to show up in context of what we are doing right here.
That doesn't mean they don't think it is important; it just means that it may be taking something out of historical context. So I think it is important to be sensitive to that when one compares or researches the proposals of different groups, that they are internally directed in how they are formulated.
DR. LUMPKIN: Thank you very much for coming and giving us an update on this. We will be interested to see how widely the recommendations are accepted.
DR. ZWANZIGER: Thank you for having me, and as I said, I'll leave information if anyone has followup questions. Don't hesitate to call me. I'll be glad to do anything I can to help.
DR. LUMPKIN: Thank you. The suggestion was made yesterday about trying to rearrange the layout, because it really was as conducive as we would like. But it involved bringing somebody in at the time we tried to do that, and the logistics of it were just a little bit too much to overcome. But we will make efforts at the next meeting to make sure that -- well, it will be at the Humphrey Building, so we'll be back to our old digs, but any future meetings we have at this site, we'll try to do more of a U-shaped arrangement than the T.
DR. IEZZONI: Are we trying to negotiate whether we have future meetings at this site?
DR. LUMPKIN: I'm saying if we have future meetings at another site, an off site, yes.
DR. STARFIELD: I wasn't really sure whether I was supposed to discuss my commentary or the World Health Report. But I'll start off with a couple of remarks about the commentary and summarize it for the electronically wired audience, and then go to the World Health Report very briefly, at least to the extent that I understand it, and then come back to the commentary, mainly from the point of view of the implications that it has for data needs for the country, that I think this committee could play a big role in.
First of all, just to summarize the commentary, I should say one thing. This commentary was not written in repose to the World Health Report. It actually was stimulated by our very own Jeff Blair, who wrote me an e-mail about six months ago and said he had heard that maybe the health statistics in this country weren't the best in the world, and he wondered whether I could provide him with any source of information so he could explore that. I thought, if our very knowledgeable and sophisticated Jeff Blair didn't know, perhaps I should write something about it. That was the genesis of the article.
I thought I wrote it as a sounding board for the New England Journal. They sent me a letter back and said they hadn't sent it out for review, but all the editors read it, and they found it very interesting, but they didn't think their readers would be interested in it, and therefore it wouldn't suit their needs. I said, sure, I quite understood the rationale for that. I thought maybe it had something to do with the very large amount of drug advertising in the Journal.
Let me summarize the article. What I did in the commentary was, I reviewed the data on where the United States stands relative to 12 other Western industrialized countries. The only country actually in the comparison that isn't West was Japan, but all comparable industrialized countries.
I found that all the comparable health indicators in the United States did pretty poorly on most of them. Let me just give it to you for the electronically wired audience. The United States is 13th or last for low birth weight percentages, 13th or last for neonatal mortality and infant mortality overall, 11th for post-neonatal mortality, 13th for potential years of life lost, and that excludes external causes; that means it excludes all the injuries from accidents and homicides, 11th for life expectancy at year one for females, 12th for males, 10th for life expectancy at 15 years for males, 12 for males, 10th for life expectancy at 40 years for females, ninth for males, 7th for life expectancy at 65 years or age for females, 7th for males, third for life expectancy at age 80 for females, third for males, and 10th for age-adjusted mortality.
So what you can see from there is that the disadvantage that the population of this country suffers gets less as you get older. It is much worse at younger ages than it is at older ages. That has some implications for the kinds of data that we want to focus on, and it also made me think about all the responses that I received to this article. I got about 30 e-mails, all of which I answered, and some of which went on to an interchange of about six e-mails on each side. Also, the six letters that I got from the Journal that they decided to publish. One of the people said, he thought the problem was obesity in tech U.S. population that was accounting for the disadvantage in the health statistics. I thought, if that were the case, you wouldn't have the greatest disadvantage early in life; you would have it accumulate and show up later in life.
So there are a lot of things that you want to think about, when you think about how these data play out.
Now, I speculated and it probably was no more than speculation, that one of the things that might have contributed to our poor ranking was disparities in wealth. We are the country among our comparable industrialized countries that has the greatest disparities in wealth, that is, the difference in wealth earned by the top groups, as compared with the lower groups, by far. Whatever measure you use, it is a very consistent finding.
Since disparities in income have been shown to be related to various measures of health, they would be contributing to our disadvantage.
But there are other explanations. By the way, none of the explanations have to do with the behavior of the population. I've given probably 100 talks on this subject over the last five years, and at every one of them, someone says, the population behaves badly, and that accounts for the bad statistics. You can show pretty well that that is not the case. We have among the lowest rates of smoking, the lowest rates of drinking of any of these countries, on average.
One other thing that comes to mind was stimulated by the IOM report on errors in medicine, although I carry it one step further and look at not only errors, but unanticipatable adverse effects, that is, things you can't attribute to errors. Not much data on that. I'll come back to that later when we talk about data implications, but there is some evidence that would lead you to think that iatrogenic effects account for the third leading cause of death in the country, if you add up the statistics, and I did that in the commentary from other published works. It is easily the third leading cause of death, just below heart disease and cancer, and it is well above the fourth leading cause of death. There is a very large margin of error in there. You can reduce the figures from iatrogenic effects, and still have it be the third leading cause of death.
How does this relate to -- does anybody want to ask any questions about that now, or shall I just continue on with the World Health Report?
DR. LUMPKIN: A question. You raised the issue of medical errors and iatrogenic effects. Does that disproportionately impact younger populations?
DR. STARFIELD: I can't sort out the data that way. There is very sparse data on that. But that has got great implications for this committee.
DR. LUMPKIN: Thank you.
DR. COHN: I actually read over your article two days ago or three days ago, and I thought it was very interesting.
DR. STARFIELD: Twelve o'clock at night or something.
DR. COHN: Probably something like that. I was just going through it real quickly. You probably have mentioned it here, but you list toward the end three major impacts or three major causes.
As I looked through this, you didn't seem to mention quite as strongly the impact of the people that are uninsured. I looked, and I see that there is income equality, social equality and all of this stuff, but is this underneath --
DR. STARFIELD: No. Thank you for reminding me of something I forgot to mention.
DR. COHN: I just wonder if that explains a lot of what is going on.
DR. STARFIELD: The impact of the health care system.
DR. COHN: Is that what that is under?
DR. STARFIELD: Yes. I think that the impact of the health care system is different than what we generally think of as the impact of the health care system. Some of the studies that I have done, and that are increasing in frequency in the last five or six years, there really looks to be a big impact of the kind of services that are delivered, particularly primary care versus specialty services.
When I say that, everybody's eyes glaze over. That is because we are not used in this country to thinking about the distinction between primary care and secondary care. Most of the comparable countries really have a clear separation of primary care and specialty care. Primary care is what is delivered in the community and specialty care is what is done in a hospital. In this country we have lots of specialists, and we don't make a clear distinction in ambulatory care between primary care and specialty care.
In fact, very few of our survey data systems will allow you to make that distinction. So you can't tell whether a population is being exposed to primary care or specialty care. Some of the surveys, the national health interview survey and the MEPS survey, have moved slowly in this area, in being able to distinguish the two in the last few years, but not nearly to the extent that they could.
There are still some surveys -- Paul and I were talking about some of the adaptations of the SLATE surveys that do not make this distinction very clearly. But if I look at all the data that I have published and that other people have published, it looks to me that more strongly oriented primary care systems do better. That is not an absolute rule; there are some primary care systems that don't do very well. But on the whole, when you average them out, the primary care oriented systems do better.
That was a finding, by the way, of the World Health Report, although they didn't look at it that way. But if you look at that data that way, you see it also.
I don't know how many of you have read the World Health Report. It is pretty complex. Chris Murray is the major author. He is a physician. He is a New Zealander, has a position at Harvard, but he is now at the World Health Organization. He is very sophisticated. Either he has some economics training, or he works with people that do, because a lot of his work is econometric.
He came up with five main measures of health of all the countries in the world, an incredible feat. A lot of it is very technical, very mathematical, and I'm sure I don't understand all of it, but I think I understand some of it. I had an opportunity to talk to the Manuela Akida, who came to the International Society meeting in Havana, and gave the only public presentation in this report. That was a good opportunity to ask her about some things. Other than press conferences, I don't think it has received a wide recognition verbally; it has been in the press a lot.
Anyway, the five measures of health are the DALE, the disability adjusted life expectancy. That is life lived in good health; it takes into account mortality. The United States is 24th on that. All industrialized nations are above it, 23 industrialized nations are above it.
The second measure is child survival to age five. That has a disparities component in it. The United States is 32nd on that one. Here are two young age measures. Any life years statistic will be heavily weighted by loss at earlier years of life because they have so much more life to be expected. So those are two indicators relatively weighted toward the younger ages.
Then we've got two measures of responsiveness. Responsiveness is basically the supposed response of the health system to those who have contact with the health system. The source of information for this responsiveness was key informants, and no so-called objective data that went into it. They asked people in the country, so-called experts, to tell them about how responsive the health system was to people who had access to the health system. So this explicitly excludes people who don't have access to tech health system. The United States, not surprisingly, ranked number one on this indicator.
Then there was a related indicator, also from key informants, which was the responsiveness distribution. That was a report of the informants as to the extent to which the system is responsive to particularly disadvantaged groups.
DR. GELLMAN: Were they like government people?
DR. STARFIELD: Not a clue, and not a clue as to whether they are comparable across the country.
Then the fifth indicator was the fair financing indicator. That was basically a ratio between family health spending, and total spending of the family minus absolutely necessary expenditures, primarily food expenditures.
I'm not really sure I quite grasp this indicator, because the more money you have, the more money you're going to spend. That goes into the ratio, regardless of need. The United States was somewhere between 54th and 55th in this indicator.
Then there were two summary measures, one overall health indicator, which was the optimum DALES minus the DALES in a system where there were no health services. They used the year 1900 as baseline for that. Also very complicated equations that went into this. The United States is ranked 72 on that indicator.
The overall performance was the five measures weighted. They did weight the measures, and they used experts to come up with the weights. The United States was 37th on that indicator.
So this WH report what I did, and suggested that maybe I was too kind, too generous in my rankings. What it is interesting is that it did confirm the special disadvantage at younger ages of the population in the United States.
DR. COHN: I don't think you have addressed my question, though. I found your analysis very attractive. I am personally in favor of primary care and against all the things that you're against and all of this that you wrote in the paper. But as I thought about it, I was sort of going, gee, could all of this be explained by the fact that you have a lot of people under 65 uninsured, thereby not having access to care. We know that once they turn 65, they seem to have a better life expectancy than most other places.
DR. STARFIELD: Not better than most other places, but certainly not as bad.
DR. COHN: I'm sure it's better as you get older, which would be something that you would say, if iatrogenic disease were a major issue, you would expect as you get more health care, which is typically something that happens when you're older, you die off faster. I'm just trying to be logical about this one.
Once again, you talked about the primary versus specialty piece. Were you able to tease out this uninsured issue and the impact of that? Do you think that that is a major issue or not?
DR. STARFIELD: There is a literature on what insurance does for you. There is no question that insurance is important.
I actually wrote an article in Annals of Public Health this year about the -- I questioned what insurance does for you. It looks to me like if you put having insurance in the regression equation and a good source of primary care in the equation, primary care comes out more important than insurance.
So what insurance does in my reading of the literature is facilitate access to a primary care source. If you don't put primary care in the equation, then obviously insurance looks more -- if you could assure people of good receipt of primary care without the insurance, it turns out you can't, but if you could, you probably would do just as well.
I don't want to minimize the impact of insurance, because it obviously does that. But it is what the insurance does for you rather than the insurance alone.
DR. GELLMAN: Just to comment on the same issue, in a lot of empirical studies, what they show is that insurance has a very significant impact on getting into the health care system, but once you're in the system, then issues like the provider characteristics play a more important role, and insurance is less important at that point.
So I think it depends on whether you are looking at getting into the system or once you're in the system, what kind of care you get. At that point, then primary care and secondary care make a big difference in the provider characteristics.
DR. COHN: I am willing to agree with that. It was just that I couldn't see that this piece was teased out very well in your discussions. I was just trying to figure out in my own mind if there is an effect that has to do with just getting into the system. I think we would all believe that not having insurance is an issue, but you're saying that it is beyond that, that there is obviously a major impact on not having insurance to begin with, because you don't even get into your primary care physician at that point.
DR. STARFIELD: I made reference in the commentary to some studies that were done that look at the impact of the primary care physician to population ratio in the 50 different states, in conjunction with the income disparities in the 50 different states. There is a wide spectrum of income disparities in the 50 states. In every study we've done, we've done it at the state level and at the MSA level and at the county level; the primary care physician to population ratio does reduce the ill effect of large degrees of income inequality. It has an independent effect as well.
The reverse appears to be the case with specialty care. But I don't want to make too much of that, because it depends a lot on the kind of specialty and the kind of outcomes.
DR. LUMPKIN: I assume that is exclusive of emergency medicine.
DR. STARFIELD: We haven't been able to tease out the relationships between professional care and outcomes. Also, it turns out if you look at any disease, no more than 40 percent of people with that disease get to see a specialist. So a lot of primary care in virtually all disease oriented care.
DR. SCANLON: Barbara, if you look at it the other way, what are the characteristics of the countries in the top five, in terms of across most of the measures? What are those countries and what kinds of systems do they have?
DR. STARFIELD: I actually last night did the rankings of the countries, in the three groupings that I have used in my studies, the poor primary care is Belgium, France, Germany as well as the United States. Tech good primary care is Denmark, The Netherlands and the United Kingdom. Also, the good countries, but not quite as good in terms of primary care ranking, is Canada, Japan and Sweden, and the average rankings for the health indicators, the DALEs and the child survival equity, the primary care oriented countries are ranked much higher.
Not so for the responsiveness indicators, the market oriented systems, not higher ranking, and the responsiveness indicators -- it stands to reason, in a marketed oriented system you're going to play to those who can afford the care, and they are going to buy the kind of care they want. Fair financing is also better for the primary care oriented countries.
DR. LUMPKIN: Other questions?
DR. STARFIELD: Jeff, the genesis of the whole thing.
DR. BLAIR: I read the report this weekend. It was so -- thank you. I'm sure there has to be many other good justifications, other than just my question, but I am so grateful for the article, and a concise article. It was so enlightening for me, because some of the assumptions, maybe will become myths, about what was the primary difficulties within the U.S. health care system, were corrected, modified, dispelled.
My thought after reading this was -- because you wound up not having the data to be definitive about exactly what are the things we need to do to fix this, but you zeroed in on different issues, and you began to head us in different directions.
So number one, I want to say thank you very, very much. Number two, my thought was, at least for me, I thought this is helpful for me to try to refocus my thoughts on where we might focus our thoughts on the NCVHS. I was just wondering if that view was shared by those that read the article.
DR. STARFIELD: Thanks very much for your comments. Maybe it is time for me to talk about what the implications are for the data and for this committee. There are several of them, not in any particular order, just the order in which I wrote them in the commentary.
One is the type of health care received. We have got to just do a better job in trying to distinguish types of care received at the least level, to distinguish primary care from specialty care. We now have measures of primary care. We can show that these measures of primary care are in general related to the type of physician, that is, family physicians do best on measures of primary care. General internists and general pediatricians do almost as well but quite as well. And all the specialists basically don't provide primary care, although there are a few that do.
So we need to be thinking about ways that we can distinguish the type of care that people get not only from survey data, but also from administrative data.
I should say that I am not really quite sure -- Marjorie, maybe you can help me with this. I think our core data elements do have the type of physician that provides the care, so we are set up to do that from efforts of this committee, right from the beginning.
DR. GREENBERG: The national provider system includes a taxonomy, an extensive taxonomy. We're not quite sure how it gets updated, but we heard about it yesterday.
DR. STARFIELD: The second thing is, how are we going to detect adverse effects? We don't have a good system for doing that.
Some other countries are doing better than we are. Both Britain and Spain for example routinely monitor drugs that are prescribed. I think France does as well, though I'm not quite sure.
Japan just recently within the last couple of months mandated post-marketing surveillance of drugs. Didn't give the responsibility to provide it, but it gave it to the industry. They must do post-marketing surveillance. We don't have any system for doing that.
In the core data elements, we do have an item for signs and symptoms that people experience when they come to medical care. I think that is probably the only way we're going to do it, is to start systematically documenting signs and symptoms that people have in conjunction with the medications that they are given, to start to begin a science of drug epidemiology that in the long run can help us with a systematic look at adverse effects of medications.
DR. GREENBERG: I just wanted to mention, this came up last year at the annual meeting of the ICD centers around the world. At the 2000 meeting last month, adverse effects and how different countries capture that information is actually on the agenda, and there are going to be some papers on that. So I'll be glad to report back what I find out.
DR. STARFIELD: You're going to report back in November, I hope?
DR. GREENBERG: Yes.
DR. STARFIELD: That would be great. The other thing is e-codes. There is no good consistent use of e-codes. E-codes do enable you to attribute effects of things, interventions. If we had a better way to use those, get the data and use them, I think we could go pretty far ahead.
On the issue of equity, you can tell from the study, the micro studies that we have done as well as the World Health Report that the United States is particularly disadvantaged on those measures that have to do with equity.
Our information doesn't really allow us to get at equity very well. We do have some things in the core data elements that get at peoples' socioeconomic and sociodemographic characteristics, but as you all know, we have had some heated discussions on the best way to do that.
I have come to think over the years that we have been discussing it that we probably will get at it only by geocoding. There is a lot that you can learn about peoples' experiences just by knowing where they are. You make a lot of mistakes at the individual level of attributing geographic characteristics of individuals, but on the whole, you probably do just about as well as you do getting individual level characteristics.
That is something that every health care facility gets. Everybody knows the addresses they send bills to, where people live. So it is not an extra burden for the system to collect information that could be used for geocoding. We now are pretty far ahead in using addresses to characterize where people are. The software does that for you, and to link it with census data.
So I think those are the main implications. For the very long run, I think that we really need to rethink the ICD. We are probably talking a couple of centuries away. The ICD forces you to code the underlying cause of death. I was shocked the other day to realize that patients who die from tuberculosis who have AIDS are coded as dying from AIDS, which is really quite incredible, because tuberculosis is by and large a treatable disease, and AIDS isn't nearly as treatable.
So from the point of view of implications for the health care system, you're getting all these people coded as dying from AIDS, when in fact they die from an acute illness like tuberculosis.
DR. GREENBERG: At least in the U.S., though, and most of the major industrialized countries, there is multiple cause data. So tuberculosis would be on the death certificate.
DR. STARFIELD: I'm not sure we use it very much though, do we?
DR. GREENBERG: We do quite a lot of analysis in this country.
DR. STARFIELD: That is pretty much all I want to say.
DR. LUMPKIN: Simon?
DR. COHN: I want to go back to something you said just a minute ago when you talked about geocoding. I want to make sure I understood what your point was on that.
I think what I hear you saying was that the geocoding could be a reasonable surrogate for race and ethnicity.
DR. STARFIELD: Not for race and ethnicity, but certainly for socioeconomic characteristics. We had a presentation from Nancy Kroeger. In fact, I asked her that question, if you remember.
DR. COHN: I remember, on socioeconomic health but not race and ethnicity. It is something we talked about and we probably need to put back on our radar to handle somehow, looking at some of these things, because clearly we are not doing as well as we might with some of these data.
DR. STARFIELD: There is now an increasing literature, it's not in the medical literature, but within the social science literature, although some of it has gotten into the British Medical Journal, that suggests that characteristics of the area residents themselves have an ecological effect.
In fact, we have started working -- as part of his working group on 21st century, we have started working on the existing models as determinants, and we find that we have to put in specifically ecological characteristics as well as individual ones.
DR. LUMPKIN: I'm almost afraid to call on Jeff, because we're going to get another paper. No, actually I think we are appreciative that your question sparked this, because it is a good discussion. Jeff?
DR. BLAIR: By any chance, on those nations that rank the highest, which I think was Sweden and Japan and --
DR. STARFIELD: Canada.
DR. BLAIR: On those that ranked the highest, was there any population surveys or policy analysis done five, 10 or 15 years ago in those countries that supported the direction in health care that they have taken? Because if there was, then maybe those would be useful studies or surveys that we could initiate in our country, that might put us on a more effective track for health care.
DR. STARFIELD: That is an interesting question. One of the letters that I have to respond to, that will be in JAMA, was a letter that suggested there be a commission to do long-range planning on the goals of the U.S. health care system. I think in general, that is probably a good idea, but my big concern is that the people who would get appointed to that committee probably would make the same assumptions that all previous commissions have made.
I think the answer to your question is yes. Canada for example started a long time ago. Canada has always been interested, much more than we have, in terms of equity. We heard a presentation, I think we did, from Norman Larousse or Charlin Black, about the population oriented data system in Manitoba. There are some provinces in Canada that are decades ahead of us, and as we play it out, probably more than that, even, that are looking at administrative data -- not even survey data, administrative data, to look at the distribution of health care and health in that province. They are able to link the administrative data to survey data, because each province has a unique health identifier.
So yes, the answer is yes. I don't know a lot about details in other countries. I think Australia has done it, U.K. has done it. I'm not sure about the Scandinavian countries or France or Spain. Spain in the mid-1980s made a conscious decision to completely revamp the health care system, so that reoriented it from hospital based to primary care. It was called the primary care reform.
So I think the answer is yes.
DR. BLAIR: Thank you.
DR. SCANLON: Barbara, it looks like the -- in your WHO analysis there was not an attempt to look at population diversity, or I'm not sure that there was. To what extent might that be an issue in the other countries, and how could we better get a handle on it?
DR. STARFIELD: The WHO did make an attempt to do that. Chris Murray is more interested in looking at disparities from individual to individual than by way of social groups, and his methods reflect that. So his equations are all the mean, minus some of the individual differences. So he hasn't looked at it by systematic differences across social groups. Certainly the whole notion of equity comes into the World Health Report.
DR. FRIEDMAN: I just wanted to return to the geocoding issue for a moment. Generally speaking, I completely agree with Barbara that that is really the way for us to go.
At the same time, having said that, as part of my continuing shilling for Canada, let me say that there are a couple of things that have been done that really facilitate both geocoding as well as analyses of economic status, one of which is, their census has a question on it which asks permission for individual level linkage of census data with other specific data sets.
Granted, Canada is a very different country than the U.S. Having said that, there has been a surprising amount -- surprising to me as an American -- of agreement with that question, and people saying, sure, you can link.
Secondly, --
DR. LUMPKIN: Does that include the INS database?
DR. SCANLON: I don't think so. The second thing, on a less sensitive matter, Canadians also have a system where their postal codes, the equivalent of our zip codes, roll up into the equivalent of census block groups and census tracts. Ours are completely disconnected.
And in terms of geocoding administrative data sets, having zips that roll up into census tracts or vice versa would really make a huge difference.
DR. STARFIELD: But you still could aggregate by census tract. You're not forced into postal codes.
DR. SCANLON: Right, absolutely.
DR. LUMPKIN: Marjorie?
DR. GREENBERG: This goes back to the presentation that the committee had in February, but the whole issue of summary health measures. There is now an interagency working group in the Department that Dr. Sondig is chairing, and at least right now, it looks like the summary health measures may drive a fair amount of data collection in the future, or drive at least recommendations for data collection.
I wondered what your thinking was on the extent to which the National Committee should be continuing to look at that, or what role would be appropriate for the committee.
DR. STARFIELD: I don't have any objection to summary measures. However, I think they can obscure a lot of things and interfere with our ability to disaggregate, so that we can attribute responsibility.
So I am in favor of the summary measures only if the had a disaggregatable component. That is what we find in our work. We are working with summary measures in our child health work. We have profiles of health which are one description of each individual in a population. But they come from information in the different domains of health, that we can sort out the impact of one thing versus another. I wouldn't put a lot of effort into the summary measures, unless it were disaggregated.
DR. GREENBERG: I guess in addition to how you saw them, I wonder if you felt there was a role for the committee in following that.
DR. STARFIELD: Oh yes, absolutely.
DR. LUMPKIN: My thought was that it is consistent with some of the things that we discussed as part of the feedback from the retreat. We may not necessarily need to do a report, but we could do a letter to those who are preparing the summary health statistics with certain concerns that we may have in crafting those, and certain desires that ought to be reflected, such as ratio of primary to secondary care.
So we may want to give some thought to doing something along those lines. Jim, you had your hand up?
DR. SCANLON: Barbara, some of the measures you described were included in Healthy People 2010, but many of them were not. Some of the health status ones that relied on the mortality data, it looks like they are also included in the Healthy People framework. Most of the health system ones are not.
DR. STARFIELD: That is an opportunity to mention about infant mortality. Every country uses infant mortality. We use infant mortality. But infant mortality is two different things, and they are influenced by different things within our health care system. Neonatal mortality is influenced very heavily by the birthing process and neonatal intensive care and post-neonatal mortality is not.
So we ought to think about disaggregating, if we are interested in what is responsible for the impact.
DR. SCANLON: The other point had more to do with geocoding. I think HHS would be interested in any recommendations that the committee would want to make, in terms of geocoding. And there have been a couple of developments recently.
Let me just describe -- because it is a little variation of the typical geocoding approach. With the establishment of the research data center at NCHS, it provides us an opportunity to do linkage and analysis that normally -- under protected conditions of privacy and confidentiality, that normally wouldn't be available. NCHS cannot release geocodes for individuals, for example. It would be -- or even information about the geographical location below a certain threshold of population. That has been the policy.
But the research center allows certain kinds of research to be done, where the researcher doesn't walk away with any of this individual information, but the aggregate analysis or tabulations can be done. Actually, there are a couple of projects that have allowed researchers to link with the health interview survey, which is in essence our flagship health survey in the U.S., and our biggest survey, to link neighborhood characteristics for the individuals that would be available from the census or other places, to link those with the individual or the family on the health interview survey, rather than aggregate census tract information from the health interview survey, which is just more complicated.
DR. STARFIELD: I'm glad you said that, because I think it is important for the committee to know that. AHRQ, the Agency for Health Care Research and Quality, isn't quite as far ahead.
DR. SCANLON: No, but they are developing --
DR. STARFIELD: I know, they are moving in that direction, but very slightly. There is the potential to link the survey data to the census data, but we really don't have a mechanism for researchers to do that yet.
The concern is privacy, of course. But if they could set up a mechanism whereby they could do in-house analyses from specifications, I think that would go a long way. Of course, this has cost implications for them, but I don't know any other way to get around it.
Do you have any comments about that, Mike?
DR. FITZMAURICE: Only the self-serving comment that we would welcome the budget to do just what you said.
DR. STARFIELD: You're not going to comment on the research.
DR. FRIEDMAN: I think the committee following up on both the geocoding as well as summary measures, particularly in terms of the capacity of summary measures to be disaggregated, both at a national level for subgroups as well as the state and local level, makes a lot of sense, a great deal of sense.
I also think that sometime in the next six to eight months, we may want to invite Nancy Kroeger back to give us an update. She has got a national infant and child health and development project working with state data, surveillance data, from a couple of different states, and looking at different measures of inequality at zip code, census tract and block group levels. She is much further along than when she spoke to us very generally about it.
DR. STARFIELD: Actually, I think she didn't speak to the whole committee. I think she spoke at the 21st century --
DR. GREENBERG: No, she was on the race/ethnicity panel for the whole committee, a year ago June.
DR. LUMPKIN: As a mechanism to follow up on some of the issues that were raised here, do you think that the work group would be an appropriate place to --
DR. FRIEDMAN: The work group, the Subcommittee on Populations, whatever.
DR. LUMPKIN: Kepa. You mentioned Spain, and I thought for sure that would get a rise out of you.
DR. ZUBELDIA: You mentioned the benefit of using e-codes. I don't know if you are aware, but the HIPAA standards on the professional side don't support e-codes.
DR. STARFIELD: They what?
DR. ZUBELDIA: They do not support e-codes.
DR. STARFIELD: I wasn't aware of that. I don't know whether anybody else on the committee --
DR. ZUBELDIA: On the institutional side, you can send e-codes, but not on the professional side. My recommendation would be, if you are interested in e-codes, to put in a request for the DSML website to have them add the e-codes for the next release.
DR. STARFIELD: Okay. Is that something the committee could do?
DR. GREENBERG: Do they allow -- it is allowed among diagnoses, but there isn't a separate element for it, as there is on the in-patient? And actually, having the separate element on the in-patient really did make a major difference in hospital data, because it went from about five states mandating to now, I think there are nearly 30 or something. But it is definitely something that the Public Health Data Standards Consortium is looking at.
DR. ZUBELDIA: You can send it, but I don't think it is supposed to -- you can send in the same element as the diagnosis, because it is an ICD-9 code.
DR. GREENBERG: Right, it is one of the diagnoses.
DR. ZUBELDIA: But on the institutional side, you have to amend it to supply e-codes.
DR. GREENBERG: Yes, definitely. This committee was very important in getting that to happen.
DR. ZUBELDIA: And maybe we should push for getting that onto the professional --
DR. STARFIELD: I was not aware of that. I think it is important for us to work on that.
DR. GREENBERG: We did recommend that in the core health data elements, in the ambulatory care data sets.
DR. STARFIELD: I think the adverse effects from institutions are going to be heavily weighted towards the procedures rather than towards the medications, although there certainly are adverse effects in medications. But on the out patient side, there are procedure adverse effects. A lot of the medication effects that show up are going to be in the out patient setting rather than the in patient.
DR. GREENBERG: I think there is an exception to that, isn't there? I would do better if Donna Pickett were here, but adverse effects for medicine is supposed to be captured among the diagnosis codes.
DR. STARFIELD: In the out patient?
DR. GREENBERG: In both cases, I think.
DR. STARFIELD: We need to look into that.
DR. GREENBERG: I'm not sure about that. It is something we can report back on.
DR. COHN: I just want to add a note of sanity here. First of all, there are eight diagnoses, and probably more that are included as possibilities in the fields on the professional 837. There is certainly more than enough space to put an e-code in. But I think if one did an audit on professional bills, one would find few and far between.
I think it is unlikely, unless you start encouraging payers to pay for adding an e-code, that you're going to start seeing any e-codes. I don't think adding a field is going to be the right answer. I think most providers are probably feeling pretty overloaded with these various standards in the data requirements already. Just a comment.
DR. STARFIELD: I think Simon is actually right, but as with everything, one strategy is never enough. You have to take a multi-front approach, of which this is one. I think provider education is just to increase their awareness of it.
DR. COHN: Provider education is wonderful.
DR. ZUBELDIA: Practically speaking, all the payers today will reject the claim if it has an e-code on the professional side.
DR. COHN: Not as a first diagnosis.
DR. ZUBELDIA: On the professional side.
DR. STARFIELD: As the first diagnosis, but not -- in fact, the coding guideline is an underlying diagnosis always.
DR. FRAWLEY: I just wanted to follow up on Barbara's point regarding the adverse effects of medication. There is actually a 900 code in ICD-9CM that we would use to report that. In addition, you could use an e-code. So there is specificity there to pick up that information.
The biggest problem both in the institutional side and the provider side of course is documentation, where you would be able to pick up that information and report it accordingly, because if it is not documented it would be very difficult for any coders to be able to identify that and code it using our national coding guidelines.
So I think that is a problem. I think on the provider side, looking at the data reporting that comes out from their claims, you would tend to find less data being captured. It would probably be just the reason for the encounter, and very limited procedures, if there were any procedures performed in the office. I would suspect you would find very little in the way of any information being captured that would utilize an e-code.
DR. STARFIELD: That's true. The few studies that have been done from the out patient setting on adverse effects have used that 900 set of codes, recognizing that you can a vast under estimation.
DR. LUMPKIN: I think at this point, we will look to either the work group or the subcommittee to follow up on this. There are a couple of issues that we need to look at. The e-code issue is one, obviously, the extent to which we believe the summary health measure process will assist us in addressing this issue. So we should have some followup to the committee down the road.
Simon, can we do the letter now?
DR. COHN: I think we can try. We've got copies of it and we have been feverishly making some additional changes, but nothing really very substantial. Is everybody willing to go through the letter at this point? Okay.
I should first of all say that we have attempted to respond to all the comments from earlier today, as well as done a significant amount of wordsmithing. I think whatever we pass probably has to be passed with the recognition that there will be some minor wordsmithing that may occur post.
What you see here, and I'll start reading through it, is a three-page letter, and then there is an optional final bullet that we need to consider, of whether we should include it or not.
Anyway, let me just start reading this letter, since most of you haven't seen this particular version. It might be the best way to handle it.
Dear Secretary Shalala. As part of its responsibilities under HIPAA, the NCVHS will be monitoring closely the implementation of the administrative simplification final rules. Our intent on an ongoing basis is to identify implementation issues and barriers and make recommendations to you that address these issues.
The following concerns and recommendations are based on public hearings conducted by the NCVHS Subcommittee on Standards and Security on July 13 and 14, 2000. The hearings focused on two major areas, issues surrounding code sets, both related to externally maintained code sets, and the widespread use of local, non-standard, non-national codes in health care transactions, and the experience of over-implementers of the health care transactions adopted as standards and the administrative simplification provisions of HIPAA.
Any comments so far? Shall I continue?
One, externally maintained code sets. These are code sets that are maintained by external entities rather than by the designated standard maintenance organizations, that have agreed to maintain the standard transactions. External code sets include codes for provider taxonomy, in parentheses, specialty and type, place of service, claims adjustment reason, claim status, and their admittance advice remarks.
Testifiers expressed concern that the maintenance processes used for external code sets must be effective, timely and responsive national processes that are based on the same principles adopted by the DSMOs. Testifiers stated that these processes should consider all bona fide business needs, permit balanced participation from across the health care industry, and provide a widely available mechanism such as a website to receive requests, publicized deliberations and disseminate decisions.
We recommend examination of the maintenance processes used by the external entities and discussion with the external entities to resolve any weaknesses that are found.
Comments? To move on.
Two, eliminating HCPCS level two codes. According to those testifying, movement away from Health Care Financing Administration common procedure coding system, HCPCS, level three local codes as required by the final rules for standards for electronic transactions is possible, but will require significant effort and resources.
The HCPCS is comprised of level one, level two and level three codes. HCPCS level one codes are based on the American Medical Association's code procedure terminology, CPT, and are used to classify procedures and ancillary services provided to patients. HCPCS level two codes are issued by HCFA and used to identify durable medical equipment, drugs, supplies and services not covered by level one codes. Both level one and level two codes are nationally listed and maintained.
HCPCS level three codes are independently issued by health plans for specific local purposes and vary by locality. Unlike level one and level two codes, they are not nationally listed and maintained.
I'll stop there for a second. This is obviously the information that people wanted as background on what the heck these codes were. Are people satisfied with that so far?
DR. LUMPKIN: I just have a quality.
DR. COHN: Sure.
DR. LUMPKIN: You say they vary by locality. If a health plan, if two health plans are in the same locality, could they have two different local codes?
DR. GREENBERG: Probably.
DR. LUMPKIN: So maybe locality isn't the right word.
DR. BLAIR: Maybe it is locality and facility.
DR. COHN: I actually don't know the facts on that one, I apologize.
DR. ZUBELDIA: Yes, each health plan would issue their own local codes, so maybe we can end the sentence, for specific local purposes.
DR. COHN: Okay. Unlike level one and level two codes, they are not nationally listed and maintained, period. This may still require a little bit of further work, but basically the national Medicaid electronic data interchange HIPAA work group is serving as a valuable focal point for Medicaid programs attempting to consolidate their level three codes and request new level two codes. That is what I wrote right there.
Testifiers mentioned several shortcomings in the current HCPCS level two maintenance process that needs to be addressed.
DR. SCANLON: If I could go back over that suggested correction, does everybody agree to that?
DR. COHN: I guess the question is, does it make more sense. I found attempting to consolidate and standardize their use and procedure codes didn't -- is still a little bit confusing.
DR. SCANLON: Let me hear it again.
DR. COHN: What was there, or what is in here now?
DR. SCANLON: What you proposed.
DR. COHN: I was proposing, efforts to consolidate their level three codes, meaning HCPCS codes, and request new level two HCPCS codes.
DR. SCANLON: Let me see if I understand the process. What we're saying is that with the elimination of level three codes, if someone has a code that seems to be some utility on a national level. That code may have been in many places as a level three code.
DR. COHN: Right, exactly.
DR. SCANLON: It is not going to be put in level one, because that has to go through the CPT process. So it is either going to be in the level one or level two code.
DR. FITZMAURICE: They may be forced to use a CPT code rather than their own level code, if the CPT code fit.
DR. SCANLON: Right. So anything for which there is a business purpose that currently is a level three code will have to be accommodated if that is accepted within level one and level two.
DR. COHN: That's our recommendation, yes.
DR. SCANLON: Right. Do we say that anywhere?
DR. COHN: I think we say that further down, about future use. We handle it more about expected frequency of use. But it's basically the same thing. We also talk about bona fide business use?
DR. SCANLON: That is what we should consider, bona fide business uses.
DR. COHN: We can catch that after we hit all the --
DR. SCANLON: Okay.
DR. COHN: Are we okay generally with where we're going so far?
DR. ZUBELDIA: I don't think they are trying to consolidate level three codes. Level three codes are being eliminated.
DR. COHN: How would you rather -- we can take this offline and work on it. This is a statement more of fact than anything. I just found in wordsmithing that this sentence was not quite right, and there probably needs to be an editorial panel for final wordsmithing, if that is okay, just so we don't spend too much time trying to solve these particular sentences.
DR. LUMPKIN: So let's just put a little asterisk, then we can just have that.
DR. FRAWLEY: Actually, the way you could fix it would just to say, the national Medicaid electronic data interchange HIPAA work group is serving as a valuable focal point for Medicaid programs to standardize their use of HCPCS procedure codes.
DR. LUMPKIN: That works for me. Let's think about it.
DR. ZUBELDIA: Standardize their use of national procedure codes.
DR. COHN: That sounds fine. Mike, have you got that?
DR. FITZMAURICE: It's up on the board. The proposal is to delete what I have in black and to replace it with, attempting to standardize their use of national procedure codes.
DR. COHN: I think maybe calling these procedure
codes is probably -- is it confusing?
DR. STARFIELD: They are really services. Some of these things aren't what you would consider procedures.
DR. LUMPKIN: I don't think we need that modifier, either. Just say national codes.
DR. COHN: National codes, okay. Is that okay with everybody? Good. National codes. Sounds good.
Testifiers mentioned several shortcomings -- about the maintenance and updating process in HCPCS level two codes, as well as the role of the respective code set developers.
DR. FITZMAURICE: Simon, the word about, as well as about the roles of -- just for parallelism.
DR. COHN: As the other code set developers?
DR. FITZMAURICE: Educate the industry about the roles of the respective code set developers.
DR. COHN: Okay, sounds fine. One immediate action is for the HCPCS national panel to clarify this kind of information on their website, but much more needs to be done.
The next one is, recommendation that the HCPCS national panel consist of representatives from the Blue Cross Blue Shield Association, the Health Insurance Association of America, and HCFA. We recommend that representation on the panel should be expanded. For example, state Medicaid representation should be considered. The current mode of representation on the national panel have led to a perception that requests from small entities or other health care segments will not receive the same consideration as those with national panel members.
Next. The process for evaluating and granting requests for new HCPCS level two codes must be able to accommodate a large number of requests in the initial months of HIPAA implementation, as organizations request standard codes to replace their HCPCS level two local codes. The process must also be capable of prompt response in order to avoid the need for temporary codes, and must be able to handle needs that do not pertain to HCFA programs.
Next. Some of the local codes may be considered for inclusion in other code sets. Other national code set developers should be alerted to a potential increased volume of requests for new codes.
Next. The current process for considering requests for new HCPCS level two codes includes a criteria requiring the demonstration of a significant frequency of past use, whereas what is important is its expected future use. We recommend that the criteria be modified to reflect the value of the new code's future use.
Then finally, in addition to these topics mentioned by the testifiers, the committee recommends that HCFA develops instructions to the Medicare contractors on the use of national drug codes and level two HCPCS code that replace level codes, these instructions be quickly disseminated to the public. This will assist the rest of the industry in its implementation of the same or similar procedures.
Are we okay so far? Questions, comments before we move on to three?
DR. FITZMAURICE: Simon, we can probably out the NDC, since I don't believe we refer to national drug codes anyplace else in the letter.
DR. COHN: Sounds fine.
DR. ZUBELDIA: We replaced the J codes.
DR. GREENBERG: The acronym.
DR. ZUBELDIA: Oh, the acronym.
DR. COHN: Are we okay so far?
DR. FITZMAURICE: I just think we're missing an and, replace level codes, and. It is just awkward. It's just a style issue. We can move on.
DR. COHN: Item three, then. HHS resources in promoting industry HIPAA implementation.
A number of experts and industry representatives suggested that HHS should assume a more active role and increase funding to actively promote the implementation of the HIPAA data standards by the industry. We recommend that not only should the Department closely monitor the progress of national implementation, but it must devote sufficient resources to insure that there is adequate technical support, education and testing.
Four. Need to deploy identifiers. Testifiers stressed the need for adequate funding to build the infrastructure and obtain necessary support to deploy HIPAA identifiers, especially the provider and health plan identifiers, in a timely manner. The timely availability of HIPAA identifiers is crucial for obtaining the expected benefits of electronic transactions. We recommend that HHS provide adequate resources to assure that HIPAA identifiers are available for identifying all providers and payers as soon as is possible.
Five. Testing and compliance with HIPAA standards. Testing was identified as a critical component of HIPAA implementation. There was concern from testifiers that different private certifying bodies using different criteria could provide different results to the industry and thereby undermine implementation. We recommend that HHS take an active role in providing support for uniform compliance certification. The role should include activities such as certifying the certifiers, so that purchasers would not be misled by unsupported claims of HIPAA compliance software or services.
Are we okay with that?
Now, this is essentially the letter. I will direct your attention to number six on the next page. This is an issue that was brought forward by one testifier. I personally am a little divided about whether we should put this in the letter, and I would look for the help of the full committee, in terms of making a judgment on this.
This is an issue that I think is of sufficient gravity that we probably need to in November either have the subcommittee hold a small hearing or probably a panel discussion for the full committee to better understand this issue. But let me read to you this issue and see if you think it should be included in this letter at this point or not.
DR. FITZMAURICE: Simon, I'd direct your attention to the board, because I have taken some of the comments that you made and some that I made on it, and put the changes in bold, and put some side brackets I would recommend deleting.
DR. COHN: Let me just turn my head then to see. It says, one testifier informed NCVHS that provider computer systems do not currently store some of the information that will be required in the standard transactions, or store the longer identifiers that are being proposed. Even if a data element is deemed to be situational or conditional based on the type of transaction or circumstance, the provider must have the ability to process that element in the event that the situation or condition applies.
These deficiencies cannot be corrected by the use of clearinghouses or other processors, because they do not have the data. Providers must rely on their own software vendor to make the necessary alterations. We recommend that HHS strongly encourage vendors of provider software to make their products comply with the transactions and identifier standards.
These changes must be done soon and delivered to the providers in order to allow them, the providers, time to perform and necessary training and process changes, so that they can realize the benefits that HIPAA brings. NCVHS plans to hold hearings on this and other issues --
DR. FITZMAURICE: To understand.
DR. COHN: -- to understand the breadth of the problem and to provide additional guidance to you.
Now, the question is, this is an issue that was brought up by one testifier. There is currently further evaluation that is going on right now about all this stuff, and I understand we could probably arrange a panel in November to fully understand this issue. I will look to the committee.
Obviously, this is an issue of major concern, don't get me wrong. I just am concerned about bringing it forward until we have a little better handle on it, but I would defer to the committee in terms of their desire to put this in or not.
DR. LUMPKIN: If I can make a suggestion, --
DR. COHN: Please.
DR. LUMPKIN: -- I think that the way it is, it is too complicated. A very simple statement is that the committee is concerned that vendors may have products that currently cannot produce standard messages. Then, we recommend that HHS urge vendors to bring their product into compliance as soon as possible, so that we just eliminate everything in the initial thing, and say this is a concern that we have, and that we intend to delve into it further at the next hearing.
I would just take everything before, we recommend, and put it into one sentence saying that we are concerned.
DR. COHN: Mike, are you capturing that? Let's take a look at it on the screen here.
DR. LUMPKIN: That vendors of provider systems, and then we recommend what is left in that paragraph.
DR. COHN: That is certainly a lot simpler than what we have right there.
DR. LUMPKIN: Tech standard HIPAA transaction messages.
DR. ROTHSTEIN: I don't see how we can make a recommendation until we have all the facts. If we want to say that we are going to be looking into this further, I think that is fine. You can say we're concerned and that we're looking into it. But it seems that the recommendation ought to follow the fact finding.
It may turn out that when we actually have a hearing, that we'll learn from the vendors that they just can't wait to get their new products on the market, because they are going to generate all these sales, and they are ready, willing and able to go, or some other thing that would move our recommendation.
DR. LUMPKIN: So you would suggest that we eliminate the, we recommend, and then just say that we are concerned and therefore we're holding hearings?
DR. ROTHSTEIN: Yes.
DR. COHN: That certainly represents some of my concerns that I think I was expressing, so thank you.
DR. LUMPKIN: So we're agreed, the committee is concerned that vendors of provider systems have products that cannot produce HIPAA transaction messages.
DR. GREENBERG: Or the committee heard testimony?
DR. LUMPKIN: No, I think -- the NCVHS plans to hold hearings focused on this and other issues to understand the breadth of the problem and to produce additional guidance, which was the last sentence that was being proposed. So it would just be two sentences in that paragraph.
DR. COHN: So this actually is not a major thing, but is more like finally, so there would be no recommendation to the Secretary?
DR. LUMPKIN: That's correct.
DR. COHN: We go back to a finally, comma, a sentence or two before, thank you very much for your consideration, or whatever.
DR. LUMPKIN: Right. We currently say we appreciate the opportunity to offer these comments and recommendations. We would say finally, and then leave.
There is a motion by Simon, seconded by Kepa, that we send the letter as we have now modified.
DR. COHN: There is a wordsmithing recommendation here on the one -- we basically said we recommend examination, on the item one. This is obviously a recommendation directed to the Secretary. We'll make a wordsmithing change.
DR. LUMPKIN: The motion is to adopt the letter with the leeway to do a few editorial changes for readability purposes.
DR. COHN: Thank you.
DR. LUMPKIN: All those in favor, say aye.
(Chorus of Aye's.)
DR. LUMPKIN: All those opposed, say nay. Abstentions? It carries.
At this point we have a break for lunch. We will return from lunch, and then have reports from the various subcommittees and work groups. Let me just understand, if we have any additional action items. I know we have one from Lisa's committee, on a letter that needs to be prepared, and that we need to respond to before November.
DR. IEZZONI: We need to respond in a matter of a couple of weeks, but we don't have a letter and we don't know exactly what it is going to say.
DR. LUMPKIN: Right, but we need to approve the process.
(The meeting was recessed for lunch at 12:00 p.m., to reconvene at 1:00 p.m.)
DR. LUMPKIN: The next item we have on the agenda for this afternoon is an update on the GCPR project. We have David Camacho, who is senior management advisor to the GCPR Executive Committee, Dr. Steve Brown, GCPR, from Lexicon Terminology, lead, and Dr. William York, project lead, Lytton PRC, GCPR prime contractor. Welcome.
DR. CAMACHO: Hi. I'm David Camacho. On behalf of the GCPR Board of Governors and the Executive Committee, we are pleased and honored to be able to present you with an update on a program that we think is very special.
Today we will cover a short program overview. We will talk a little bit about the state of tech project, the promise of GCPR.
I think now one of the major things that has happened since the last time that you were briefed, as I understand it, is that we have something tangible. There was a proof of concept done, and we did transmit information across disparate systems, although it wasn't in a laboratory setting.
Then we're going to talk about, consider the possibilities. That is pursuant to the idea that, now that we have developed some of what we think is pretty exciting technology, we think we should try to use it now, particularly within the purview of our internal partners.
Our roots are from Presidential Review Directive Number Five, which says that VA, DoD should get together and create a comprehensive -- we have added virtual -- patient medical record.
After these groups came together to begin to talk, it changed to virtual, because it was realized that there are existing legacy systems that have the information, but the task at hand really was to find a way to get the disparate systems to talk to each other.
Our vision is to provide public information, clinical information to the point of service, I guess we would say. Our goal as we mentioned earlier is this lifelong virtual medical record.
Our program goals are to create a secure and encrypted system, patient focused national architecture which I understand we got from NCVHS, if I've got that right, and interoperability, which has turned out to be our major product. We tried to make computer systems be interoperable.
This is our leadership structure. Board of Directors, Dr. Garthwaite, Rear Admiral Church from Indian Health. I report directly to the Executive Committee. Each of the partners has two members assigned to the Executive Committee. We now have a GCPR program office in Skyline; if you ever have a chance, please come out and visit us.
These are our customers. Of course, the three partners. VA, DoD and HHS are the three primary signatories, and then the patients, providers, labs, ancillary services, Congress, GAO. GAO has taken a special interest in this; they are there about once a month, and we have a lot of opportunity to excel, and it is a wonderful thing.
The customer focus. This is something that we have added. We feel that whatever we do must be transparent and user friendly and should create no additional work when we're talking of the clinical setting.
One of the things is, GCPR does not solve world hunger, and we are not in competition with any other existing system. We are a developmental project, and sometimes we may have to make course corrections.
Why a joint venture? For one reason, we were told so. Number two, the Congressional guidance as I'm sure you folks know is that there is a need for more interagency sharing, federal sharing, to avoid redundant activities, and because it makes sense to do so.
This is how we visualize the framework happening. We have what I call a black box. Dr. Kolodner doesn't like me to call it a black box. I have Dr. York and Dr. Brown who are going to go on the technical side of the house, but all the information would go through our framework and it would go to DoD facilities, even support military deployment, civilian facilities, VA facilities and of course IHS partners. But it is much more than that, and that is what creates the excitement with the project, that it is much more than that.
The functional requirements. These are the functional requirements that were the basis for a statement of objectives, which was what was the statement of work for the GCPR contact. These were the things that -- any framework would have to be able to address multiple levels of data. It was point of care decision making, so it would be meaningful to the commissions that would it. It would meet or exceed HIPAA requirements, which is a tough one, and to provide complete, accurate and timely information.
The framework as I mentioned earlier was developed based on existing heritage systems. The goal was to develop a common vision between the three partners to identify unique activities, and then to design interfaces. From that, we would develop the generic security standards. As it turns out, we decided to use the DRD, because they were the most stringent. That is one of the rules we follow. When in doubt, we take the toughest requirement.
Then we hope to create in this synergy a federated master patient index, and then working together we are aiming to create a test environment to test out the framework.
These are some of the key components of the GCPR, common information components. I won't go through the list, because Dr. York and Dr. Brown will be talking about them, so I'll just move on.
DR. BROWN: Because of our need to interchange information between the three agencies' information systems, terminology and mediation really play a critical role for GCPR. We developed a two-pronged approach to our terminology strategy. It has been a challenge, but we are making progress.
The short term model or STM was created to meet our immediate GCPR needs through the creation of a low cost registry of terms in actual use by the agencies. The STM was designed as a necessary but not sufficient for the first step in GCPR terminology strategies.
The STM is a database that GCPR has developed, is a registry of the agency code sets in use, and standards codes as well. Presently in contains some 650,000 terms from some 8,000 code sets, and all the terms are linked to reference information model attributes.
The STM is a partial meta data registry. It allows us to begin meeting recommendation 3D of the NCVHS data standards recommendations regarding an open meta data registry. We have resource dependent plans to allow us to fully participate in that effort.
With regard to addressing recommendation 3A of the NCVHS report, the state data report, the members of the RTM working group are actively involved with HL-7, and we are trying to figure out what components of STM should be contributed through the HL-7 V3 terminology.
The reference terminology model or RTM is GCPR's planned long term approach to the terminology mediation question. The RTM will be concept based and have 40 definitions. We are defining the RTM giving due consideration to terminology quality standards and guidelines as put forth by ASCM, the anti-HISBE CIPRI paper, and of course NCVHS.
The RTM planning process includes evaluations of available tools, terminology models, maintenance issues and other core planning areas include collaboration planning and an RTM demonstration project.
We have completed approximately four months of a six month demonstration project that is designed to help us operationalize the RTM process. The RTM demonstration project has developed and begun to populate a draft reference terminology model for drugs that allows for aggregation based on active ingredients.
One of the first areas we need to address for a production GCPR RTM is a transmission of disparate information about medications, and we look forward to a collaborative effort in achieving this. Based on the demonstration project, the GCPR is ready to operationalize building RTM for drugs, in line with the frameworks on needs.
We have recently also begun working with Kaiser Permanente, NCI and other federal agencies. We believe the terminology issue is one that must be addressed in a collaborative fashion.
DR. CAMACHO: Thanks, Steve. As far as the project time lines, we are right now in alpha pilot testing. The pilot testing will go on through about June. That is where we will test to see if GCPR is not a foreign body in the VA, Indian Health, IHS systems, make sure that we obey everybody's rules or whatever the rules are, security, whatever, telecommunications.
It is a rigorous thing, and it is again one of the more joyful parts of the project, to work with each of the agencies. We hope to get something out in the field about September of next year. That will be in Alaska. It will be about a six-month test. It will be limited.
What we will be doing, folks, is, we will be testing four partitions, pharmacy, radiology, a little demographics and a little laboratory. From that, we will then evolve into a beta which will be a much more robust test, with about 10 partitions, which include things like immunizations, physical exams and other things. The goal is to make a more useful product for the commissions, to encourage their use. We don't know what we'll do after beta.
One of the things that has come about as a result of the amalgam of the three federal agencies is the idea of common sources architecture. It was kind of a hard concept, so I got this diagram to try to help explain what we are talking about.
If you consider GCPR as the amplifier in a stereo system, the GCPR controls amplification, volume control, tone quality, et cetera, and with that you can connect other equipment, other systems. This could be Vista, this could be CACS-2. But you can see that we see us as, we provide the common services that anybody would have to have if they were going to try to create an interoperative system.
DR. LUMPKIN: What's the turntable?
DR. CAMACHO: IHS. To put it in terms I understand, I think of the GCPR framework as a wondrous black box. We get things in from VA, DoD, IHS and hopefully the private sector, and we put it in our wondrous black box, and we mix it up and we make it intelligible to the person that is drawing the information. It is stored and cached, ready to be used.
In our tests, we will be putting out these products, the lab, the out patient, the lab and demographics, as I discussed.
Dr. York?
DR. YORK: Thank you. You just saw the black box. We're going to make it a little more colorful now, and use it to talk about some very interesting parts of the interior of that black box.
The problem is -- and everybody has a great deal of concern about the security aspects of moving medical information. I think we all would have that about our own records.
As Dave has said, we're going to follow the DoD lead because of how stringent it is. So we will use that as our guideline. But security isn't everything. Just because you know who is looking at it, does that mean they should be looking at it? So there is another piece of this called privacy that needs to be dealt with.
There aren't any guidelines out there, so we have come up with what we think is a unique way of dealing with the fact that this is going to be an evolving field, because as we go along, things are going to happen. Right now, the only thing that happens in privacy is, you sign a little piece of paper that says you can move the record from one institution to another, and you depend on the institution's medical record side to take care of the privacy issue. Now that we have a system that can move records from many sites all at once and then glue them all back together again seamlessly, privacy becomes a much bigger issue.
So if we look at this as being the framework and in this is that it is secure. The third is an extra secure area here, and we have to have an area where we perform all of the functions that make it work. But in order to do this, the issues of policies around the privacy act issue are an evolving field.
So in order to deal with an evolving field, you don't know what it is going to look like. So we have had to go out and figure out how can we put together a way of instituting policies as they come out from multiple sources.
First of all, we have three agencies to deal with, and we know they are all going to be somewhat different. But each institution has rules about how they allow their data to be used. Within that, the issue of the rules. An administrative person can't see the same kinds of data as a physician. They have no need to see all that information.
So what we have to do is incorporate a way of talking about roles and policy issues, plus our finding out who that user is. That is the security piece. We put all that together in a policy role security evaluation module, which then makes a decision. The decision is, do I let the gatekeeper pass information through, and if so, what information can pass through. If not, no information passes through.
Now, to make it work so that it is efficient, we pre-position the heritage system information in the cache. That gatekeeper will then allow that information if appropriate to go on to the user in the amounts and the types that are dictated by the policies and the roles. So we feel this is fairly unique.
Let me reiterate again. The framework has two regions. It has a user region. This is where all the activities and processes occur within the framework. This is secure, but not as secure as the second region, which we call a trusted extension of the heritage system. No one can see the information in that area, not even the system administrator of the framework. So it is as if it still remains in the original system.
Now, the gatekeeper is just that, it is a gatekeeper. It is either going to allow information through, or it is not going to allow information through. We have a policy evaluator, which looks at all of the policies that are out there. Those policies come from those that own the data, the roles, policies about individual roles, and the use of the information within the institution that is receiving it from their policy aspects. Policies reside with data owners and with the people that are going to use it.
Now, if I take policy role and security, I can combine them to either open a gate or keep it closed. That is how we plan to deal with the privacy issue. We feel this solution is unique enough that we are going to pass on how we do this to the Culvermed group for incorporation into their RAD service, and provide all the information that we are working through with this demonstration to develop a new standard to help with tech policy issues.
DR. CAMACHO: Thanks, Bill. As I mentioned at the beginning, our mission is to create a lifelong comprehensive medical record that started out with the veterans on active duty. We have expanded that.
But here is one way that we are looking at getting to that goal. First of all, we are looking at, where is information required in the continuum from pre-service all the way through to retirement. For example, the VA would want entrance physical information. Right now, they don't have access to that. That would result in reducing a lot of the redundant stuff.
So there are various things that we are doing that I will talk about a little later on towards beginning to create that lifelong virtual medical record. A synopsis would be to pull information from medical pertinent pre-service information that would go to the active duty side and to the VA side at the same time. We would do that in conjunction with the VA and the DoD and all other interested agencies.
We have this slide called Consider the Possibilities. What we mean here is that now that we have a tangible product, it opens it up to a lot of other possibilities that it wasn't really designed for. Kind of like the Internet, in a way; the more that people get involved with the project, the more new ideas can be associated with it.
We believe that we can become the foundation and should become the foundation for the lifelong virtual medical record. In house sees it as a way of connecting rural health, and that is really going to be tested in Alaska. Then as the means of common services architecture, which is a really big deal as far as the DoD is concerned. In fact, many of the DoD folks think that is the most important value of GCPR, the common services architecture piece.
Since I have been with the project in the past five, six months now, Dr. Rodner said I had to go earn myself. So what we have done is, we are beginning to try to look for partners. We are starting internally with our three partners, where we are looking at within and without. We are looking for working together, either sharing costs of development, or even providing services, being the government wires, perhaps, for some of the things.
Here are some of the things we are doing. The VA disability system; we are meeting with the other side of the VA, the VBA, and they have asked to meet with us with a view towards getting the demographics from the DoD to the VA, which is a problem right now.
The VA director in Alaska has asked us to see if we can transmit information from civilian providers in Anchorage to the VA in Anchorage, so we are going to try to test that. Also, we have reached a tentative agreement with the Tri-Care contractor, Foundation Health, who have also agreed in concept to bring information from the Tri-Care contractor to the DoD test sites in Alaska.
Another thing that we are delighted with the initial meetings is, we are meeting with the Kaiser Permanente folks in Oakland. Steve and Dave have been the lead on that, with a view towards perhaps developing an MOU towards terminology development.
We are working also with the folks as perhaps being the wires for the federal credentials program, where information has to go across systems. The Military and Veterans Health Board is that board that is directly responsible for creation of the lifelong virtual medical record.
We are also meeting with the Garden Reserve folks. Where that comes in right now is, the Army Guard has made a decision to contract out all of their physical support to the VA, and they need a way of getting information from the VA to the Guard.
Other federal agencies have also been in contact with us, and we are now ready to start talking to other federal agencies. Then the common sources; that is an underlying theme that we try to do with all of the partners.
Other initiatives that we are looking at are the readiness piece, in particular I mentioned earlier the work about the lifelong medical record. How that is tied in real well is that we were in Hawaii recently, and Hawaii has been assigned as a test site for an accelerated physical exams program, and they have asked us to see if we can support them making that happen. So you can see, as we go out and start talking about what we can do, people are asking, could you do this with us.
If you have connected IHS, VA and DoD, you have wonderful opportunities for research, and if you had more, you'd really have great opportunities. So now that is beginning to be considered. We are also looking for support to other projects, for example, DoD, IHS and VA are getting hot and heavy into population health and studies there, too.
Other major benefits. We hope to be the genesis of creating that federated master patient index, because you will have to know where the patients are in whatever the system is that you have. We certainly encourage standardization. We are sending people everywhere, and in some cases, people are now even recording some of the stuff that our smart guys like Steve and Bill are doing.
We foster the use of the state of tech art tools, so that we will be meaningful to as many systems as we can. Our goal is to be as open as we can, so people will want to use us. And of course, hopefully that will reduce costs.
Future patient benefits. I really think it is a wonderful goal to say that we are trying to provide information at the point of service. That is something that is so needed in our country, that I think it is what makes such devotion to the project.
Another thing is, this improved continuity of care across time and location. If you are really concerned about adverse actions or something like that, you need the information for wherever the patient is, if you are responsible for that patient.
So you can see, it is very easy to believe in the project.
Future government benefits include the transfer of information between disparate systems. I mentioned the source of medical research in population studies. I think just with VA, DoD and IHS, there are 1777 medical facilities. That gives you an idea. And this idea of just in time medical information available anytime, anywhere, is again a worthy cause.
Where are we now? We are getting ready for a pilot. Actually, we are in pilot. We are going north to Alaska sometime next year, weather permitting. We are now looking for a short-term return on investment. We are looking for some relationships, and it doesn't always have to be money, but we certainly could use the money, but we are looking for getting more federal partners involved and even civilian partners involved in the project.
We are looking first inward at VA, DoD and IHS. One of the things that is beginning to be needed to be discussed is what will be the ultimate use of the framework within each of the agencies. In our view, we hope that folks like you and others will see that we should be a framework utility for America.
We'll now open it up to any questions you may have for us.
DR. GELLMAN: I've got a couple of questions. First I want to ask you how you guys are planning to comply with the privacy act publication requirement of a system of records notice. Is there going to be one for the system? Are the agencies going to do it separately? Is it going to be coordinated? Have you thought about that at all?
DR. YORK: Yes, we have thought about that. Unlike most medical records, this is a patchwork of many medical records. So our logging and audit systems have to be a little more complicated than what you would look at if you picked up a single record as defined in a single institution. You could say who looked at it and what went on with it.
In order for this system to work, you would have to look at who is looking at all of the pieces of the record. At this point, the determination whether those audit records would be passed back to the originating source or would be maintained in a central repository, has not been made. But both possibilities are there to be used.
DR. GELLMAN: Okay, but I was asking about system of record publication requirements.
DR. CAMACHO: We are still working with our health information people, but because we are not storing a permanent record there, it is there temporarily, and then it disappears. Our initial read is that a system of records is not indicated for this particular system. If that is different than other groups, it would be --
DR. GELLMAN: I would give a different opinion, but I think I need to know more facts.
The second question. In the DoD appropriation bill this year, Congress passed a provision that says no medical record can be disclosed outside of the Department of Defense except for law enforcement or national security purposes, without the consent of the individual. I want to know how that provision of law will affect this system. Have you thought about that? Have you seen the language?
DR. CAMACHO: Sir, as we speak, and I can speak to DoD a little bit, as we speak, the way DoD handles that is that at the bottom of each record, there is a form signed by the individual which gives authority to release information under certain circumstances. We have been told that we need to have a legal review of all that, but we were told that with that approach, we would probably be able to get the information that we need. But we don't have a final answer yet.
DR. GELLMAN: Given the way that most medical records are used and disclosed, especially in a system of this type where you've got all this joint activity, I'm not quite sure who has what, where it is, whatever. All the secondary and tertiary users, in order to get consent for that, you're going to have to give people about a 20-page consent form in order to get consent, and then you're going to have to deal with the prospect of people denying you consent for various disclosures, for various purposes. It would seem to me to be pretty cumbersome to deal with.
DR. KOLODNER: I think what you are referring to is the Boxer amendment. If we're not given permission to move the data, then we're not given permission to move the data.
In fact, because of the flexibility of the system, the way that it was designed to meet HIPAA requirements with them not yet finalized, and because it knows who the data is about, we have looked at the fact that the system could be implemented in the future, so that an individual has to electronically give permission to move the data. So this particular system actually has the possibility of restoring back to an individual control of the flow of information, at least to the extent this framework is used to move their information.
DR. GELLMAN: I am interested as this thing develops and as you figure it out, because the Boxer amendment doesn't make any sense, but everyone has got to live with it, at least for a year.
DR. LUMPKIN: Jeff.
DR. BLAIR: A couple of questions. Steve, where are you? One area that I want to understand a little bit better was -- I have two questions, first of the two. You indicated that you created a reference terminology model, and it seemed to be focused primarily on medications or drug codes. I wanted to have a little bit of an understanding as to why you felt it was necessary to create your own model, and related to that is, to what degree is the FDA working with you or assisting you on that model, or any other drug knowledge based vendor?
DR. BROWN: We have created a draft model as part of the demonstration project. The main purpose of the demonstration project is to learn the lessons necessary to operationalize the process. So the main goal of it is not to say, this is the final medication model for the world, but one to get us rolling along.
So the reason -- we did look at what was available, and have borrowed as heavily as possible, and of course that would be our intent all along. But what is available doesn't always hit the mark. I think if the recent anti-HISBE meetings are any indication that NDC level codes for instance really aren't exactly what we need, and clearly in your recommendations of having at terminology that is aggregable by active ingredient, for instance, it is fairly obvious that it is not yet done.
So we took a stab at it, and we would love to share and get as much input as possible, there is no doubt about that.
DR. BLAIR: Is there anything that you could give us as advice or guidance? Obviously you read the recommendations we had; you might have even heard the testimony, in terms of going forward with either -- well, it would be nice to be able to select drug codes for patient care, but falling short of that, are we on the right target with the recommendations we came up with? Or would you suggest that we revise our thinking in some way?
DR. BROWN: I think surely beginning with aggregating around active ingredients is a good approach. From there, what you consider definitional depends upon the needs of the information system. The active ingredient is a great first step.
The whole issue of drug effects and separating effects from structure and intent is a bit of a difficult area, and one that I think maybe hasn't been done as cleanly and as well as it could be. Where the terminology ends and where knowledge bases begin is an overlapping and gray area. We would sure love to have help in thinking this through, so it can meet our needs and others' needs as well.
DR. BLAIR: The other question is, as you may have noticed in our report, some of the primary candidates when you start thinking of medical terminologies, clinically specific terminologies, were LOINK and MEDSIN and SNOMED-RT. I was wondering what observations you may have come across with respect to those clinically specifically medical terminologies in respect of whether you expect to be able to use them or use them with modifications, or whether you are running into some problems.
DR. BROWN: We hope to be able to use -- not invent the wheel as often as possible, of course. So looking at the structure of the SNOMED reference terminology, for instance, it being based on description logics, you would hope that the federal license that is being negotiated now would go through, and we would simply be able to use that, because it has been done well, to the extent that I have been able to evaluate it, and it meets most of the criteria that have been set forth in ASTM and all the other standards.
So my understanding is also that they have incorporated good pieces of LOINK into it already, and taking it from a second to a third generation terminology system. So anything that has already been done in that fashion, great, bring it on. We'd love to use it. I think our goal is to try to patch together other pieces as suggested in the framework paper, so that it fits together and is all consistent, and to try to do so using computer based tools to reduce our maintenance costs.
DR. BLAIR: The last piece, does MEDSIN fit in there for history and physicals, or have you not had a chance to look at that?
DR. BROWN: We haven't had a chance to formally evaluate it. We have created terminology model criteria that are general and then partition by partition, area by area, we will create and make available specific criteria, for instance, drug terminologies or exam terminologies and whatever, and then apply them with a fair market survey, and try to include those as foundational terminologies to determine where they are, their content, their structure, and all the important things, to see how that fits into this patchwork that we are trying to piece together and keep consistent.
DR. KOLODNER: Jeff, you had asked about FDA. About two weeks ago, the VA did meet to talk about our internal needs for selecting adverse drug events. We had people from FDA there. We had a very lively discussion, talked about the direction they were going with MEDRA, the direction we're going with GCPR. Also had people from -- I'm not sure what the initials stand for, but NCCMERP? I'll see if I can find out what it means, but it is another group. I think it is the USP activity.
In any case, we had the discussion. We talked about the need at least -- since VA was needing a terminology for adverse drug events that would fit into a larger structure of signs and symptoms that we needed to be sure that there was, to the degree that we can, the harmonization between any terms that were using GCPR and MEDRA and the other, so that we either translate those that are translatable, or know that we have to collect separately things that don't map across those.
That is a charge that has been given to us, to look at how we can link those, because otherwise, it could be a situation we don't want to have if FDA moves MEDRA out, which I believe is geared to a standard created by the pharmaceutical companies and the regulatory agencies around the world, and working toward HL-7 and some of the other standards. We are going to want to be able to move those back and forth, anyhow.
DR. LUMPKIN: I have a couple of quick questions. This may not take very long to answer.
The first is that every state has what are called veterans homes, which are long term care facilities for veterans. Would these be covered under this system?
DR. CAMACHO: I would not dare to speak for the VA. Just a moment.
DR. KOLODNER: Right now, we have no data exchange with them. Those particular veterans homes are often covered by the states, not by VA. VA on the other hand spends one billion of our budget for care that is delivered outside our rolls. Whether it is to get more complete clinical information about the people that we pay for on the outside, or to provide information on veteran through providers outside, again all of us with the patient's permission, the idea is that this system would provide a vehicle for it, and what we have to do is make sure that there are the appropriate safeguards for sharing this.
DR. LUMPKIN: The other question was, you are engaged in a pilot. I assume that one portion of the evaluation is the feasibility, does it actually work.
DR. KOLODNER: Yes, sir.
DR. LUMPKIN: But are you also doing an evaluation on the return on investment and the business model, and whether or not this in fact is a successful business model for this particular venture?
DR. CAMACHO: One of the things that we met on last week is what I call what constitutes success. We worked on performance measures, where were we trying to measure, how will we do it, how will we get outside corroboration and those kinds of things. So to be honest with you, we are probably a little beyond the power curve, but yes, we are going to be looking at it.
One of the things that is inherent in everything we are doing is, we are now being asked by our partners what is in it for VA, what is in it for DoD, what is in it for IHS. That is why you saw the rest towards internal first.
But yet, we are beginning to build up a test structure that we will coordinate and get everybody to sign up to.
DR. LUMPKIN: Thank you. My last question is, much of what I've heard you describe is involved in sharing of information, but we have an increasing focus on reduction of errors as well as a desire to add knowledges and decisional support to that. To what extent do you see that being incorporated in your current activities?
DR. BROWN: Providing decision support for things such as discovering errors and the like really is dependent on a high level of interoperability, where we understand the data that is being transmitted, so that we can act on it appropriately.
So I think that I can address it from the terminologist's point of view, which is that with formal definitions in a concept based terminology, we will have a much better handle on that. I think part of our design process for learning about it is that it is important to include hooks that are available for knowledge bases to easily share in the information as it is passed, or to act upon it.
So if we were to include maybe cast numbers, and then we can get access to materials data safety sheets, those kinds of things. I think part of our design process is to think very clearly about the boundary between terminology models, information models and knowledge bases, and then try to make it as open as possible.
DR. YORK: Basically right now, as we look at that problem as a contractor, we are building a mechanism to move information. We are looking towards the individual heritage systems to take the additional information with their decision support, where it is there, and incorporate it where the activity is actually occurring.
The framework itself inherently does not have applications internal to it, to look at the various aspects of medicine. It is geared primarily to move that data, mediate it, get it into a form that someone can use, as Steve said, through that terminology process, and then pass it to that system in a form that it is capable of taking it and incorporating it into its own built-in decision systems.
DR. LUMPKIN: Thank you. Simon?
DR. COHN: I think this is a question that is meant to be a follow-on to the terminology discussions that you were having.
Obviously, GCPR is about glues. There is a technology glue, there is a privacy glue, but it is really a lot about the data, making sure that the data that moves through has meaning, doesn't change meaning, whatever, as it moves through the various systems.
You have talked some about terminologies. I was just reflecting that there is a couple of different approaches one can take as one deals with myriad systems. One is coming up with wonderful terminologies that we then populate all of the systems with. In other words, you are basically replacing the local terminologies, sort of a grand mapping project. I am curious about which approach you are taking, not so much approach, but where you see you are going with this. Do you see eventually that things are going to be pulled out and replaced, or are you just going to be mapping and mapping and mapping?
DR. BROWN: I think that at least for the time being, pulling out and replacing everything is not really achievable.
DR. COHN: Right. I'm talking about where you're going, though.
DR. BROWN: Well, we are designed to be middle ware, and to interact with the heritage systems. I don't think it is fair for us to say, you have to use this or you have to use that.
I think the important thing is if you have the right concepts in concept-based terminology with clear definitions. I think that gets us around the forever mapping problem, or at least allows us to begin using automated approaches to reduce that burden.
In terms of saying that -- I don't know that we have the authority or the budget to begin to tell people that they have to do this or they have to do that. But we want to make things as flexible and as useful as possible.
DR. COHN: Thank you.
DR. LUMPKIN: Barbara?
DR. STARFIELD: My question was answered, thanks.
DR. LUMPKIN: Mark?
DR. ROTHSTEIN: I have a question about the policy evaluator role in this. I'm trying to see if my understanding is correct. That is, your system imposes no new substantive rules on the use of the information, but only is an application of the rules that are already in place, that have been developed by DoD and VA and other customers or users or however you define it, data owners. Is that correct?
DR. YORK: That's correct. We have no policies within it that are built in that are in force. If you noticed in the diagram, the policy issues are outside of the framework. We said they reside with data owners or the people that are using them. That concept was put in because of the very nature of the fact that this moves towards the user to make those policies, then it is available.
So we implement policies, but we do not enforce anything or put anything into the system that is, if you will, hard wired. Those policies then can evolve as the agencies evolve their own policies.
DR. ROTHSTEIN: So it seems to me then that your overall goal would be to have a security, a data security system that is better than currently exists, but your goal also would be to have a privacy and confidentiality application that is at best no worse than what exists.
DR. YORK: I would agree it is no worse, and possibly better than what does exist, if you know how medical records get moved around. This one always has someone watching over it, and a user, anytime they go in, will not be able to look at records if those policies are written correctly.
DR. ROTHSTEIN: Right, but the application of this system can be obviously built better than the rules under which it is operating, and those are in someone else's purview. So if one of your data owners wants to allow researchers, for example, unfettered access to patient data without their consent, that would go right through your system. Or if one of the hospitals that are a part of your system wanted to operate under a rule that said any doctor at any facility can get the medical records of any patient, without it necessarily being their own patient, that would go through your system, because that is not your rule.
DR. YORK: Yes and no. Given the number of facilities that we are dealing with, and in talking to the individuals, the biggest problem is going to be allowing us to see any data whatsoever. They have a great deal of concern over this privacy issue, where if you will, they feel like they are losing control. There are two policies that get dealt with. When you talk about the user and say the institution says, we're going to let our researchers look at anything, that is at the user site. But as I pointed out, there is another set of policies at the site of the information, which may not be at the user site. Those policies also have to be taken into account.
So unless everyone within the system said you have free access to everything, it won't happen.
DR. ROTHSTEIN: No, but what I'm saying is, if the VA central administration took the view that research is very important and we don't need individual consent, then it is not the role of your system to make that call. So the system is -- and this is not a criticism, I'm just trying to get clear in my mind -- it is never going to be better than the rules under which it operates.
Therefore, if we are concerned about the substantive privacy and confidentiality rules, your system is really not the place to focus our efforts.
DR. YORK: Yes, in a simple word.
DR. KOLODNER: One extension to that, though. Because we actually have NCVHS or at least the Subcommittee on CPR as a committee that guides us and we interact with them, when we go into these areas that have not yet been well defined, because people haven't implemented these, the idea is for us to be a practical implementation and to provide feedback through the subcommittee.
As we in fact did -- although we need to pursue it further -- on this issue of roles, and the question of, if my facility doesn't allow our secretaries to see clinical data, but a remote facility that the patient has given permission to have their data, does allow their secretaries to see their data because they may pull it up for their providers ahead of time, what is the meaning of that to me when I release that.
Those are some questions that are ones that are fundamental policy issues that I certainly look to NCVHS to wrestle with, as we wrestle with it internally, and for us to hopefully be a leading edge, where we can bring back issues as we trip across them, so that they can be dealt with.
Once that is established, or if a new regulation were passed by the Congress to implement such a suggestion or regulation, that then can be put in the system. The system remains flexible and can be updated with the things that are felt to known to be required. We believe that the privacy and security is fundamental to the whole framework, and it has been from the initial concept in terms of moving data. It is something where I think we are going to need to also be looking at issues of researchers getting data that is de-identified, knowing all the problems of whether you can really effectively de-identify it.
But if the system can aggregate that data and not release any demographic information, then it is possible that this actually becomes a better system for doing research. Those are things for us to discover and to learn and to interact with the committee on as we move down that road.
So what you are hearing is where it is currently, if there is a new bar that is raised through HIPAA or through other changes, that can then be built into the system.
DR. LUMPKIN: Mike for the last question, and then we'll have to move on.
DR. FITZMAURICE: John had mentioned earlier about what is the return on equity. Barbara Starfield mentioned earlier today that iatrogenic effects is probably the third largest cause of death in the country. As this project goes on, it probably is going to have to latch on to a killer application.
So I want to suggest that the killer application be reducing adverse events. The laboratory data are pretty well coded and the results pretty well known. Pharmacy data aren't very far behind, although you have pointed up some problems with the drug codes. We have seen those in the NCVHS here, and the difficulty with aggregating drugs by different ways, whether it is the active ingredient, whether it is the function they are supposed -- try to go into a system and pull out all the beta blockers. Unless you've got a data dictionary and you have pinpointed each one of them at that time, you can't do it.
You are also collecting patient data, the diagnoses and the procedures are pretty well done, but as Barbara pointed out earlier, we don't have a lot of codes for the signs, symptoms and the conditions of the patients as they present to the physician.
I would suggest that as you move along the GCPR project, you might want to consider the information that is critical for picking the low-hanging fruit of reducing adverse events, the drug interactions with the patient, and the choices of drugs and the illegible ordering of drugs.
Then as you do that, you come back to NCVHS and point to us where the informatics needs to be pushed, where health care data development is most critical for achieving the benefits that you are after. We will be happy to work in partnership with you to get that information out to the public, since we find from the public that a lot of what they say is needed to feed that information back to you as well, that we are jointly and mutually compatible forums.
DR. LUMPKIN: And hopefully when you talk to them, you will share that with our work group on computerized patient records.
DR. CAMACHO: We will do that, sir.
DR. LUMPKIN: Thank you. This has been a very fascinating and enlightening discussion, so thank you for coming.
We will move on to subcommittee and work group reports. We will start off with the Subcommittee on Populations.
DR. IEZZONI: Thank you, John. I guess we have suddenly spawned an action item. We hadn't anticipated this.
As you all recall from my prior reports, our subcommittee is currently looking at the appropriateness of using the international classification on functioning and disabilities to fill the placeholder in the core data element functional status.
We had been on target to get a draft report to you by November. However, yesterday we learned something, and that is that the WHO vetting process for finalizing what is termed the ICIDH has been changed somewhat recently by a variety of different factors that have politics involved, have research findings involved, have a variety of different motivations.
Because of that, it is not at all obvious yet what the final version of ICIDH will look like, the version that will be approved and signed off on by WHO next spring. However, there are processes that are in place for finalizing that version that we in the subcommittee and we hope the full committee will want to weigh in on.
So the action item is as follows. The subcommittee will over the next week or so, because this needs to be done quickly, prepare a letter to Dr. Shalala about our advice for the procedures that the U.S. contingent may wish to follow for finalizing the ICIDH.
This letter obviously will have to be signed by Dr. Lumpkin, representing the full committee, but there is a very stringent time line for this. This really needs to be done within the next two to three weeks.
So the issue that we need to resolve is one which John has actually spoken about in the past, which is, how does our committee turn around on a dime to react to events that it did not anticipate, which need a quick response.
DR. LUMPKIN: So your motion is?
DR. IEZZONI: So my motion is that the full committee give our subcommittee the approval to move forward a letter that we would work on, that we would probably need to have the full committee on a conference call to review, and then it would be sent forward by Dr. Shalala under John's signature.
DR. LUMPKIN: So under our previously adopted procedure, the motion would be to adopt the procedure to make a decision between meetings, which would be -- the subcommittee would meet by conference call, they would forward a letter to the full committee to review, and then we will discuss and if it is approved, adopt on a conference call that we will begin scheduling very quickly.
DR. IEZZONI: Right, because this letter needs to go to Shalala's office within two to three weeks.
DR. LUMPKIN: It has been moved. Is there a second? It has been moved and seconded that we adopt the procedure to adopt this issue in between meetings. Any discussion on that?
DR. COHN: Can I just ask a question? Or am I not allowed to ask a question?
DR. LUMPKIN: No, we are actually at the point when you are supposed to be discussing and asking questions.
DR. COHN: Good, okay. This is just a question about, we don't have the letter yet.
DR. IEZZONI: Right.
DR. COHN: But do you have a sense of the flavor of the types of things that would be in that letter?
DR. IEZZONI: Well, we will probably say that we feel that due process needs to be followed, and that some of the procedures that are currently being pursued for finalizing ICIDH by WHO are not taking into complete consideration research that has been done, and opinions of various different organizations in a consistent and objective fashion.
So it is more a letter that will be, let's follow a reasonable process here, and let's try to make sure the final version of IDICH reflects research findings and is not based upon --
DR. COHN: Thanks.
DR. LUMPKIN: Any other questions or comments? All in favor, say aye.
(Chorus of Aye's.)
DR. LUMPKIN: Opposed, say nay. Any abstentions? That is the procedure we will follow. Anything else to report from your subcommittee?
DR. IEZZONI: Our subcommittee felt very challenged in a good way by John's presentation at the executive subcommittee yesterday. As soon as we get this ICIDH thing launched along a little bit further, our subcommittee will be thinking about presenting priorities to the full committee as future efforts that we will undertake.
DR. LUMPKIN: Dan?
DR. FRIEDMAN: There are three to four activities that the health statistics 21st century process has been engaging in. First of all, we have been working to generate feedback to the interim report. Most of the feedback so far has been positive. At the same time, I am increasingly thinking of it a little bit as like asking a four-year-old whether or not she likes Christmas, because the report is general enough so that it has been hard for us to get really constructive feedback. That is obviously the task that lies ahead of us in developing the final report, is coming up with something which is much more specific, which can be much more of a planning document.
But in any case, we are trying to generate feedback in two different ways. First of all, in coordination with the NHII work group, the public hearings, the regional public hearings that we have been holding, and there will be a CDC briefing later this month. Then in addition, there have been and will continue to be a series of presentations at professional forums. I'm going to be doing one tomorrow for my colleagues in other states.
A second activity that we have been engaging in is trying to develop what is loosely speaking a model of the determinants, or a template more than a model, a template of the determinants of population health. The purpose of this is not to generate an empirical model, not to generate a model that can generate hypotheses, but to develop a model that can be used to evaluate data gaps and data needs in health statistics at local, state and national levels.
A third activity that we need to engage in, and that we have just started thinking about is trying to identify health statistics specific privacy and confidentiality issues. One of the things that we have briefly discussed is the possibility of commissioning a paper on that, which would basically be a cataloguing, or as Bob Gellman said, a market basket of potential issues and potential solutions.
A lot of those issues revolve around what I think of as boundary issues, having to do with data being used for multiple purposes. In other words, increasingly it is hard to -- except at the national level, it is hard to differentiate sometimes health statistics data, as it were, from data used for intervention or data used for research purposes. So increasingly, we are finding data that if it is used for one purpose, it is health statistics, if it is used for another purpose it is programmatic data. We need to try to pin down some of those issues in terms of privacy and confidentiality, much more than they have been done.
Those are really the three essential activities that have been going on. In addition to that, we are also proceeding to think about developing an edited volume that is partially a reflection of this process.
DR. LUMPKIN: Any questions? Lisa.
DR. IEZZONI: I've been told that I need to mention something. Because of this little hitch that has been thrown in our thinking about ICIDH, we have cancelled the October 24 meeting of the meeting of the subcommittee on populations. That is for the record. That has been cancelled.
DR. LUMPKIN: Thank you. Simon?
DR. COHN: Obviously you have already dealt with the action item, thank you. I guess I first of all wanted to announce for the public record that the subcommittee has reabsorbed the work group on patient medical record information into the full subcommittee. Jeff Blair has graciously accepted the position of vice chair of the subcommittee and will be focusing a lot on the next steps for patient medical record information. Jeff, thank you for your willingness to take on that additional role.
I also wanted to announce that obviously the staff -- many to most of the staff that have been working on the work group is also going to be working with the subcommittee once again, focused on PMRI next steps. Michael Fitzmaurice has graciously accepted to continue to lead that effort, obviously once again focused on PMRI issues. So Michael, thank you. Hopefully the others will also be willing to continue with those responsibilities.
So I think that is issue number one. I also wanted to announce that we are going to be holding public hearings on October 26 and 27 on tech issues of electronic and digital signature here in Washington. I think you all have copies of proposed questions in your packet. Certainly if you want to send to Kepa Zubeldia additional questions that you may have, we would invite them.
I think this is an issue that goes across committees, so I would encourage any of you who can come to the session to certainly come. Be aware that this is not meant to be a terribly technical discussion. We are starting at the basis of looking around at what is the business piece for what may be needed in trying to build from there, recognizing that we would all observe that the short discussions we have had here at the full committee on electronic and digital signature have not done a lot to inform and clarify our views on all of this stuff.
Obviously, the objective here, recognizing the security final rule, will not include anything on digital electronic signature, is to try to advise the Secretary on what steps she ought to be taking in moving forward to some sort of a standard in this area. So that is really the intent here.
I don't know if we'll get all the way there in two days, but I think it will be a good start.
Any questions about that before I move forward?
The other comment I would just make, and this is not by way of a discussion of a work plan, but just to be aware that there are a lot of issues related to standards and security that are certainly on the subcommittee to-do list, and probably some of them need to be actually discussed at the full committee, and maybe talking about that as we plan our future subcommittee and full committee activities.
We are as both a subcommittee and full committee tracking implementation of HIPAA standards, both the financial and administrative standards, the upcoming security standards once they come out, and the identifiers. There has been a lot of activity going around that.
There is also the digital signature that we just talked about. There is the issues of enforcement and compliance, and there will be an MPRM that will have to be talked about or hearings held next year to get into that area.
There is also the issue of new standards, a la including first report of injury, that need to be somehow discussed as well as any additional standards that need to be handled. Jeff will be leading the efforts around PMRI standards, and we will be getting in November or February a full discussion of a potential work plan on how to address that over the next 18-month period.
I'm sure I have missed something here, but last but not least is the whole issue of code sets, and the issues around ICD-10CM. We will have to hold some sort of a panel, hopefully early next year, beginning to look at what is happening with ICD-10CM. We also have a request on the table for some sort of hearings or discussion around the issue of maintenance of code sets, those that have been selected as national standards. So that will be another issue that will be coming up.
It is going to take a little bit of action to make sure we can put all of this together. This is not all subcommittee activity. I think some of this is going to have to be elevated to the full committee, and I think we have looked at any of your views on things I have just discussed should be heard by the full committee. Mike?
DR. FITZMAURICE: Implementation issues?
DR. COHN: Oh, I'm sorry, tracking and implementation was the first thing. That was item one, identifying implementation issues. So I just wanted to alert everybody that there is a lot on the plate.
DR. LUMPKIN: Is this a concerned look or is this a question?
DR. STARFIELD: It's a question. Kepa raised the issue about the potential inadequacy of the e-code. It seems to me that probably should come from your subcommittee. It is a population issue, but you have got the knowhow to prepare a letter. Maybe you could ask Kepa.
DR. COHN: Let me just clarify that. I didn't hear that the issue was the inadequacy of the e-code. The issue was whether or not there should be a simpler field on the A-37 for the professional --
DR. STARFIELD: You got it right. You just showed your committee is more able to deal with it than the population subcommittee. You just proved that.
DR. COHN: Let me make a comment on this one for discussion, and we can decide one way or another. There are lots of things that could be part of the standards subcommittee platter. Geocoding is certainly an area that is a standards issue. This is another one. But I don't think that either of those should properly be right now part of the subcommittee's plate.
I think we can try to investigate and ask to see if there is an issue. My own view in terms of the professional implementation guide, that that actually does support e-codes. I'd have to hear a strong argument from some other group that isn't me saying that there needs to be a separate field for e-codes as opposed to having them be accommodated in the professional forum. But I will check and see if there is a problem there.
DR. LUMPKIN: Simon, I think that maybe the way that this issue should be managed is that if your subcommittee could frame the discussion, and then we can discuss it in the full committee. There are a number of issues that are raised. There is a benefit versus cost, but there is also utility, given the source, and there are a whole number of issues to debate, whether or not this particular source of e-coding will bear any fruit, even if there were a spot that was available for it.
DR. IEZZONI: Barbara, do you know if e-codes are in ICD-10?
DR. STARFIELD: What?
DR. IEZZONI: Is there a separate e-code in ICD-10?
DR. STARFIELD: Yes, they are a different letter, and they are not a supplemental classification.
DR. IEZZONI: That is what I was thinking. So maybe instead of focusing on it for ICD-9CM, which is theoretically out the door, realize that for ICD-10 it might or might not even be an issue.
DR. STARFIELD: Well, it is still in the chapter.
DR. IEZZONI: Yes, but it is part of the classification, not a separate --
DR. STARFIELD: Right, so it is really considered part of 9-CM also. I guess the question is if they are going to deal with that chapter the way they deal with the supplementary chapter in ICD-9CM, and it looks like they probably are.
DR. LUMPKIN: But I think that part of framing the issue is to look at the facility transaction and how it is dealt with, and then compare it to the 837 and see if in fact there is a utility to treating this particular code, the cause of injury or illness separately or some other way that may in fact yield more information than we currently are getting, versus the reliability of the people we are looking at fill it out and actually doing the coding.
So there is a whole host of issues, and I just think we need to frame them. If we were to at some point make a decision to try to influence the standards development organization, those are exactly the same issues they are going to raise. To the extent that we can define them for them, it would be the extent to which we would expect that they would follow our recommendations.
DR. STARFIELD: I didn't plant that, I really didn't.
DR. LUMPKIN: You may want to -- since he is not here, he wasn't feeling well and had to leave early, you may want to give him the charge.
DR. COHN: I think my concern on this one, I think I may have to ask Kepa to try to put something together. I think as you can tell, I have an opinion on this one, and it is going to be very hard for me to make a case for an additional element on probably something that is already pretty large as it is already, where this one is already being captured already. But I will check to make sure that there is a capacity to capture this on the professional forum, and I will ask Kepa if he wants to delve into it further.
DR. LUMPKIN: Okay, thank you. I'm glad you said that, because I hadn't been able to pick up that you had an opinion on this issue.
DR. COHN: Yes, I am biased, I admit.
DR. LUMPKIN: Anything else on tech standards and security? Confidentiality and privacy?
DR. FRAWLEY: Our subcommittee is obviously anxiously awaiting publication of the privacy regulations. So we did spend time talking this morning about what our plan will be, when the regulations are published, and looking at our responsibility regarding monitoring implementation and advising the Secretary on issues related to those regulations.
What we decided to do is, in the interim prior to the publication of the final regulations, is look at the issues that we raised in the initial report to the Secretary on privacy and confidentiality, with the result being recommendations to the Secretary on issues that have been raised by industry in implementation of its regulations.
The other area that we talked about was unintended consequences of the regulations. There is the possibility that there may be issues that come out of the regulations or that perhaps were not addressed in regulations that have unintended consequences, which could result in a breach of confidentiality of health information. So that is something else that we will be looking at.
We do have a work plan that we need to look at, that has a number of issues that the subcommittee has expressed interest in. We have decided that before our November meeting that we will re-look at those issues and prioritize them. We have been doing some work in the area of e-mails, and did look at a report today from the Federal Trade Commission on some of their recommendations on online access and security.
We decided that we have a list of issues that need monitoring. As Simon pointed out in standards and security, there is only so many people that populate our subcommittee and that are able to take on some of these issues, so we will try to look at the issues and decide whether we want to recommend briefing sessions, commission papers, what we need to do on some of those issues.
For our next meeting, we have decided that we have to keep track at the state level, because in the absence of federal legislation -- and we know the federal regulation will not be totally comprehensive -- we need to watch what the states are doing in the area of privacy and confidentiality.
So we are going to invite a representative from the National Conference of State Legislatures to come to brief the subcommittee on what happened in 2000 in the state legislatures on privacy issues, and what the forecast is for the coming year.
The initial conference of state legislators, also working on the revision of the Uniform Health Care Information Act, which was a model developed in the mid-80s for the states to develop legislation in this area. So we will be holding that briefing session.
However, during our subcommittee session, it did come up that this might be something of interest to the full committee. So we did want to bring it to the full committee, as to whether this should come forward on the agenda to the full committee, or whether we should keep it in the subcommittee. So I think it would be appropriate to flow that over for discussion to other members who are not members of our subcommittee.
DR. LUMPKIN: Should we hear that as a full committee? The only ones I see shaking their heads right now are folks who are on the subcommittee. So can I see a show of hands of how many people think that that would be something that should be moved from the subcommittee agenda to the full committee agenda?
DR. STARFIELD: It would help me if you restated exactly what would come to the full committee.
DR. FRAWLEY: It would be a briefing by the National Conference of State Legislatures on privacy activity at the state level, in terms of what legislation has been enacted, the forecast for legislation, the issue of pre-emption in terms of how the states view the federal activity.
DR. LUMPKIN: I think you have whetted everyone's appetite, and I think we should do this at the full committee.
DR. SCANLON: Kathy, have we briefed the full committee on the analysis of state laws on the health policy project? I'm not sure.
DR. FRAWLEY: No, and we really probably should. The health privacy project of Georgetown University did a compendium of state by state analysis of privacy legislation and regulations. Probably it would be worthwhile to have that as part of the briefing. I think the thing that we are struggling with is that we can't keep ourselves focused on what is happening at the federal level if there is a lot happening at the state level that we need to monitor as a committee.
So I guess someone can put together that briefing for our November meeting.
DR. LUMPKIN: Thank you. Anything else?
DR. FRAWLEY: Nope, that's it.
DR. LUMPKIN: The NHII work group had a joint meeting with the 21st Century Vision for Health Statistics, and we will continue to have joint hearings.
I think Dan made the best description of our concern about the first one we held in Chicago, which was, we were basically asking children over four years old whether or not they liked Christmas. So we will in the hearings in San Francisco and North Carolina and Maui be focusing on asking individuals about recommendations and barriers and obstacles that they see to implementing the vision, and we hope we can get more useful information for our ultimate document.
One of the issues that we did raise is that we thought there would be some benefit in having a joint work group session, extended session, so we can discuss a number of issues and deal with issues of coordination between the two work groups, the work products that we are developing and the issues that we are managing, as well as begin to flesh out the obstacles and recommendations sections for both documents. We will be looking for some time to have some extended time to do some work on that.
We are as Dan mentioned having a briefing of the director of CDC and the senior staff at CDC on both the projects. That will occur next month, on the 10th, I think. So we will be doing that in order to bring the message to them. They have been doing some very interesting work on standards for disease reporting, so it will be important for them to understand what we are thinking, and get some exchange of ideas.
I think we in the reports covered most of the materials for the future agenda. Are there any other points we need for future agenda items?
DR. STARFIELD: No, I just want to make sure that we get Marjorie on the agenda in November to report on the two meetings that she is going to in October.
DR. LUMPKIN: Okay.
DR. COHN: What did you just say?
DR. LUMPKIN: Marjorie is going to -- I guess she is doing some travel, and she is looking for some business reason to do that, so she has two trips that hopefully we can get a report and feedback on.
DR. GREENBERG: The annual meeting of ICD --
DR. LUMPKIN: If you're going to defend yourself, you have to speak into the microphone.
DR. GREENBERG: I won't respond. October is the annual meeting of the heads of WHO collaborating centers on the ICD. Then in November also prior to your next meeting is the annual WHO meeting on the ICIDH.
DR. COHN: Have a great time. Can I put another thing on the full agenda? I think we talked this morning at the end of the morning about that item six that we were talking about. The question is, how much of a disparity is there between all the systems and processes that everybody else has out there, versus what these new standards will be requiring. We hope that by November we should be able to report back for a couple of the industry groups as well as, it may be useful for the committee to hear the initial information that we receive. So I think we can put it on the full committee agenda for at least some education and whether we have to write something further to the Secretary would be helpful.
DR. LUMPKIN: And we're talking about the provider side?
DR. COHN: The provider side?
DR. LUMPKIN: Vendors, as opposed to -- because we have heard as we were doing the hearings, probably more so on the payer side about the changes that they have to make, but this will be more of a focus on the provider side.
DR. COHN: Exactly.
DR. LUMPKIN: That was the clarification.
DR. SCANLON: A couple of population based and focused issues possibly for the November meeting. I think we have under development in HHS an evaluation for S-chip. I think we will be able to have at least the framework for that either for the populations committee or the full committee in November.
In addition, as we spoke yesterday about some issues, if the committee would like, we can get a briefing on the eliminating disparities initiative and some of the data issues and the leading health indicators.
The other area is, if the committee is interested, we could -- there is a draft HHS strategic plan that brings all of the HHS objectives and missions together. We could either have a briefing for the committee on that, and certainly provide copies of whatever draft is available at that meeting. I don't know which would be the better.
DR. LUMPKIN: I think we would have to look at the agenda and we'll have to make a call about how much subcommittee time we'll have at that meeting.
DR. COHN: This is just for information as well as an offering to the committee members. Marjorie made a set of overheads for the TMRI recommendations, as well as, there is some written text that people could use for those presentations. Jeff and I have used these at various times, and we wanted to make them available to the full committee to use in presentations as you need.
DR. GREENBERG: Do you want the Power Point presentation on the NCVHS website? Or would you rather just --
DR. LUMPKIN: I'd rather have people read the report itself, but we can maybe e-mail it to the members.
DR. COHN: Yes, I think e-mailing it alone would be -- the words that go along with the pictures would be useful to help people spread the message ago.
DR. GREENBERG: In talking about possible presentations from the Department, one of the things that we discussed at the executive subcommittee retreat was providing more information to the committee about the budget process and how it ties in with the data planning. It was suggested that November was a good time. Jim, can you help me out there? Actually, you may not have been on that part of it.
DR. SCANLON: We could schedule something, you think, for the full committee?
DR. LUMPKIN: Let's talk about that, because I think I need to see a list of everything that we're talking about doing, because we are --
DR. GREENBERG: It's a lot.
DR. LUMPKIN: Right, and we may want to begin to implement that process, and we may want to take some time in thinking about how our time can best be used in reviewing the budget process. It may be too late to start it for the planning that is going to be beginning in November. We may want to throughout the year have a few background pieces, so when we actually sit down and do it a year from November, we can have the foundation of understanding the process and what kind of input would be best able to have an impact on what HHS is planning.
DR. SCANLON: That is where we can relate the strategic planning process to the budget process, to the IT planning process as well. At any rate, this year's budget process is much different, because of the Presidential election. There will be a technical budget submitted. There will be a policy budget in January or whatever. So it is not typical this year.
DR. GREENBERG: Also, Stan, I know we had a panel obviously at this meeting on quality, and there is also one planned for the November meeting. Is that still the case?
PARTICIPANT: (Comments off mike.)
DR. GREENBERG: When we had our conference call, we had talked about organizing a panel on medical errors, including quick activities. You will be participating in the executive committee conference call? Okay, fine. That is October 5, so anyone who is not going to be on that call has any other ideas for the agenda, even for November or even February, so we can start thinking ahead.
DR. LUMPKIN: Please e-mail.
DR. GREENBERG: Please e-mail, yes.
DR. LUMPKIN: Having completed our business, we stand adjourned.
(Whereupon, the meeting was adjourned at 2:35 p.m.)