[This Transcript is Unedited]
Hubert H. Humphrey Building
Room 705A
200 Independence Avenue
Washington, D.C. 20201
JOHN R. LUMPKIN, MD, MPH, Chair. Director, Illinois Department of Public Health, Springfield, Illinois
JEFFREY S. BLAIR, MBA, Vice President, Medical Records Institute, Albuquerque, New Mexico
SIMON COHN, MD, MPH, National Director for Data Warehousing, Kaiser Permanente Medical Care Program, Oakland, California
KATHRYN L. COLTIN, MPH, Director, External Quality and Data Initiatives, Harvard Pilgrim Health Care, Wellesley, MA
KATHLEEN FYFFE, MHA, Federal Regulatory Director, Health Insurance Association of America, Washington, D.C.
EUGENE J. LENGERICH, VMD, Penn State University, Hershey, Pennsylvania
DANIEL FRIEDMAN, PhD, Assistant Commissioner, Bureau of Health Statistics, Research and Evaluation, Massachusetts Department of Public Health, Boston, Massachusetts
RICHARD K. HARDING, MD, Professor of Clinical Psychiatry and Pediatrics, Vice Chair, Clinical Services, Neuropsychiatry and Behavioral Science, University of South Carolina, Columbia, South Carolina
VICKIE M. MAYS, PhD, Department of Psychology, University of California, Los Angeles, Los Angeles, California
CLEMENT JOSEPH MC DONALD, MD, Distinguished Professor of Medicine, Indiana University School of Medicine, Director, Regenstrief Institute, Indianapolis, Indiana
PAUL NEWACHECK, DrPH, Professor of Health Policy Studies, School of Medicine, University of California, San Francisco CA
MARK A. ROTHSTEIN, JD, Herbert F. Boehl Chair of Law and Medicine, Director, Institute for Bioethics, Health Policy and Law, University of Louisville School of Medicine, Louisville, Kentucky
PAUL M. SCHWARTZ, JD, Brooklyn Law School, Brooklyn, NY
EDWARD H. SHORTLIFFE, MD, Department of Medical Informatics, Columbia University, New York, New York
BARBARA STARFIELD, MD, MPH, Distinguished University Professor and Professor of Health Policy and Pediatrics, School of Hygiene/Public Health, The Johns Hopkins University, Baltimore, Maryland
KEPA ZUBELDIA, MD, President, Claredi, Kaysville, Utah
JAMES SCANLON, HHS Executive Staff Director. Director, Division of Data Policy, Office of the Assistant Secretary for Planning and Evaluation, DHHS, Washington, D.C.
MARJORIE S. GREENBERG, Executive Secretary. Chief, Data Policy and Standards Staff, Office of the Director, NCHS, CDC, Hyattsville, Maryland
GARY G. CHRISTOPH, PhD, Chief Information Officer, Director, Office of Information Services, HCFA, Baltimore, Maryland
J. MICHAEL FITZMAURICE, PhD, Senior Science Advisor for Information Technology, AHRQuality, Rockville, Maryland
Agenda Item: Call to Order, Welcome, Agenda.
DR. COHN: Good morning. I want to call this meeting to order. This is the first day of two days of meeting of the National Committee on Vital and Health Statistics.
The committee, as most of you know, is the main public advisory committee to HHS on health information policy.
I am Simon Cohn, a member of the committee, and I also chair the subcommittee on standards and security. I am not John Lumpkin, chair of the committee, as I think you all recognize.
John asked me to stand in or, as you notice, I am sitting in, for him. He is unable to be here today. He is, as I understand it, in Illinois doing his job. So, you will be seeing him tomorrow -- his main job as director of the State Department of Public Health for Illinois.
We will be seeing him tomorrow and most of the action items will be occurring tomorrow afternoon.
I do want to welcome the fellow committee members, HHS staff and others here today. I also want to welcome those on the internet.
I want to once again remind all of our committee members and others that we are on the internet and you need to speak into the microphone. If you can, identify yourself when you are talking, so that people on the internet understand and appreciate who is saying what.
With that, let's have introductions first around the table and then around the room. As I commented, I am Simon Cohn, a member of the committee, and national director for health information policy from Kaiser Permanente.
MS. GREENBERG: I am Marjorie Greenberg from the National Center for Health Statistics, CDC, and executive secretary to the committee.
DR. STARFIELD: I am Barbara Starfield, member of the committee. I am from Johns Hopkins University.
MS. FYFFE: Kathleen Fyffe. I work for the Health Insurance Association of America and I am a member of the committee.
DR. FRIEDMAN: Dan Friedman, Massachusetts Department of Public Health, member of the committee.
DR. MAYS: Vickie Mays, UCLA, member of the committee.
DR. FROHBOESE: Good morning. I am Robinsue Frohboese, the deputy director of the Office for Civil Rights here at HHS.
DR. ZUBELDIA: Kepa Zubeldia. I work for Claredi Corporation, a member of the committee.
DR. NEWACHECK: Paul Newacheck with the University of California and a member of the committee.
DR. LENGERICH: Gene Lengerich, Penn State University, member of the committee.
DR. FITZMAURICE: Michael Fitzmaurice, Agency for Healthcare Research and Quality, liaison to the committee.
MS. COLTIN: Kathryn Coltin, Harvard Pilgrim Health Care, member of the committee.
MR. SCANLON: I am Jim Scanlon from HHS. I am the executive staff director for the committee.
MS. WHITE: I am Gracie White, NCHS staff.
MS. ADLER: Jackie Adler, NCHS staff.
MS. CABRELLI: Marie Cabrelli, Microsystems, Contractor to CDC.
MS. BEBE: Susan Bebe, NCHS staff.
MS. WILLIAMSON: Michelle Williamson, NCHS staff.
MS. JACKSON: Debbie Jackson, NCHS staff.
MR. WANG: Eric Wang, Social Security Administration.
MR. HITCHCOCK: Dale Hitchcock, HHS staff.
MS. PENNINGTON: Anita Pennington, ACR.
MS. JONES: Kathy Jones, NCHS staff.
MR. HANDLER: Aaron Handler, Indian Health Service of the Public Health Service. I am a staff committee support person to the population subcommittee.
MS. HAWK: Susan Hawk, ASPE.
MS. HORLICK: Gail Horlick, Center for Disease Control and Prevention, staff to the subcommittee on privacy and confidentiality.
MR. STEINDEL: Steve Steindel, Center for Disease Control and Prevention, staff to the NHII subcommittee.
MS. CHAVEZ: Ilke Chavez, HHS.
MR. MOORE: Bob Moore, Titan Systems Corporation.
MR. CABALE: Tom Cabale, Titan Systems Corporation.
MR. PETTICOURT: Doug Petticourt for the Coalition for Health Information Policy.
MR. FANNING: John Fanning, HHS.
MR. TATE: Michael Tate, American Dental Association.
MR. RODIE: Dan Rodie, American Health Information Management Association.
MR. ETTINGER: Dan Ettinger, Agency for Healthcare Research and Quality, and staff to the quality work group.
MR. HALBERSON: I am John Halberson, Office for Civil Rights staff.
MS. YELSEE: Judy Yelsee with the National Quality Forum.
MS. DEERING: Mary Jo Deering, HHS staff to the NHII work group.
MS. GILFOY: Helene Gilfoy, Phoenix Health Systems.
MS. CANAAN: Susan Canaan, writer for the committee.
DR. BRAITHWAITE: Bill Braithwaite, HHS staff to the committee.
DR. COHN: Thank you for all being here. Let me just review the agenda for just a second. The agenda is pretty much as you see it in the agenda book with just one or two changes.
We are going to start this morning with an update from the department followed by some discussions and updates on the privacy regulations which have just recently come out.
After our morning break, we will have a report from the WHO executive board meeting, and then review a number of documents which will be primarily for action tomorrow.
I do want to point out that there is noted in your agenda a document that says, review letter to Secretary on national drug code issues.
Now, that will be an action item for tomorrow. We do not have a copy of this letter. It will be being discussed at the subcommittee this afternoon. It will come back to the committee for discussion and hopefully vote tomorrow.
Hopefully, instead of that item, we will be able to cover the NCVHS 50-year history, which actually is an action item for this morning.
After lunch, we will be having a presentation on evaluation of the National Health Information Survey's ability to collect and report on a select set of seven health indicators. Then we are going to break into subcommittees for the remainder of the afternoon.
Based on this information, are there any members of the committee who need to recuse themselves from discussion or vote on any of the issues before us today?
Well, with that, Jim Scanlon, are you going to lead off on the update from the department?
Agenda Item: Update from the Department.
MR. SCANLON: Thank you, Simon, and good morning, everyone. Obviously a number of things have happened since we met in late November.
I am going to bring you up to date on a couple of things in the broader data policy here at HHS and related areas.
In a few minutes, you will also hear about our privacy regulation and on the status of our HIPAA data standards as well.
One of the first items to report, obviously, is that we had a Presidential election and we have a new Secretary of Health and Human Services. I have left a press release at each of your places.
Our Secretary is Tommy Thompson, the former governor of Wisconsin, actually four-term, elected four times as the governor of Wisconsin.
The secretary took office early in February. Several of the members of the Subcommittee on Data Standards were here when he addressed the HHS employees.
At any rate, in the press release, you will see some of the areas that he has indicated for his own agenda, and many of these are familiar to you.
You will see Medicare, obviously, is a major aspect of the agenda, the patient's bill of rights, affordable health insurance for those who don't have insurance and so on. You can take a look at the press release.
Hopefully, when things settle down a bit more, we will have the Secretary or an emissary from his office come and talk to the committee and really, I think, we want to brief the Secretary's office on where the committee is on some of the recommendations and reports that will be coming up later this year.
In terms of other activities related to data policy and HHS, I would bring to your attention a study that HHS has been directed to support at the National Academy of Sciences.
Again, I left a one-page description. I think this would be of special interest to the population subcommittee, but to the overall committee as well.
In the health disparities research act that was passed late last year by Congress, HHS was directed to fund a study and report at the National Academy of Sciences that deals with race and ethnicity data in the health area.
Let me say a little bit more about this, because the population subcommittee, particularly, may want to become more involved.
Actually, it directs the National Academy of Sciences to conduct a comprehensive study of HHS data collection systems and practices, any data collection and reporting systems under any of our programs or other programs relating to the data on race and ethnicity.
Then, report back to the Congress within one year of enactment of the law, a report that you will see outlines various elements.
It identifies the data needed to support efforts to evaluate the effects of socioeconomic status, race and ethnicity on access to health care and other services, and on disparities in health and other social outcomes, and the data needed to enforce existing protections for equal access to care.
Obviously that is a tall order, but we have actually had some work already begun in this area to help with the Academy.
Secondly, it would examine the effectiveness of our systems here in HHS as well as other systems in the private sector.
Here, I think it would be the data that is maintained by providers, by health insurance companies, by institutional providers and so on, where I think we have actually had some difficulty and we don't have a whole lot of control over, relating to race and ethnicity.
Obviously, develop recommendations for ensuring that HHS, in administering its entire array of programs and activities, collects or causes to be collected reliable and complete information relating to race and ethnicity.
Finally, it includes protections about the costs associated with implementing those recommendations and the possible effects on program operations.
We are currently in early talks with the National Academy of Sciences on how to set up this study. It probably would be an expert panel at the National Research Council.
I think we would have the national committee here involved in any way that we can, but we are in early stages here of planning that study. Questions about that before we go on?
DR. NEWACHECK: Jim, can you comment on how you think the population subcommittee might be involved, if this has been assigned -- or at least looks like it appears to be -- to the National Academy of Sciences function?
MR. SCANLON: I can think of a couple ways and then of course there is an aspect of this that is under the control only of the Academy.
I think a lot of the work done by the subcommittee we probably would want to bring to the notice of whatever panel is established at the academy.
There have been a number of recommendations already. There have been a number of issues raised. I think we would want to bring those to the Academy.
We may even be able to work out -- I just don't know this would work -- having someone from the committee testify to the National Academy of Sciences committee and possibly even participate as a liaison. We will just have to see how this works.
It is a little more difficult because the Academy, as you know, sets up an independent process. That is a strength, obviously, so that we can assure that there are objective recommendations.
We have less influence over the members of their expert panel.
I can think of other ways. I think maybe we can have, when the panel does get underway, have them come and brief the committee. I think there should be several ways for the full committee and the subpopulations committee to have input into that.
DR. STARFIELD: This work goes back, I think, three years.
MR. SCANLON: It actually goes back to the 1970s.
DR. STARFIELD: I mean, the last work. The question is, can we identify the continuity person.
MR. SCANLON: At the National Research Council?
DR. STARFIELD: No, at the subcommittee level.
MR. SCANLON: That is a good question.
MS. GREENBERG: We will work with the current lead staff as well as others. We can certainly pull together the reports and the recommendations.
Some of them, I know, are currently under review in the department.
MR. SCANLON: I might add, as Barbara intimated, everyone has been working on this area for a long time. Even in looking at race ethnicity disparities and so on in access to care and health insurance and actual treatment, even that goes back several years.
There is a research literature building up. In fact, we at HHS sponsored a research review by a colleague of Barbara's at Johns Hopkins, which is just about final now.
In addition, our data council's race and ethnicity working group actually came up with a very extensive and comprehensive analysis and set of recommendations for moving forward in this area.
We also have the data director, web-based data directory, where we pretty much describe what we are able to collect in terms of race ethnicity.
We have the new OMB standards. We have the new notice of tabulation guidelines that OMB has published and is currently getting guidance on relating to how to use those standards.
I think the real area where they will probably have to focus a lot of attention is how there are areas where HHS simply doesn't have control over what a doctor's office does or an institutional provider.
Obviously, this is not an area where anyone particularly wants to regulate, but I think here it is probably a matter of getting the industry together to see where a consensus might be reached on what the obstacles may be and how to move forward.
Even where HHS buys data from hospital discharge services, we are pretty much limited by what existed in the record. The same thing with most other administrative data.
Where we obviously have some direct control are surveys in our research and epidemiology systems. Obviously, there we generally use the OMB standards and we tabulate whenever we can, although there are other issues there about having sufficient numbers for all the groups.
At any rate, we are in the very early stages, and we will keep the full committee up.
DR. COHN: Other questions or comments?
MR. SCANLON: Let me go on with a couple of other things quickly. Within the data council, we have actually started a couple of new things that I wanted to bring to the full committee's attention.
In the international arena, particularly with the WHO, there have been a couple of activities which the WHO has undertaken that have gotten a lot of attention.
Last year's world health report included a ranking of 161 countries based on the performance of the health systems.
You will hear more about this from Ed Sondik today. I think there were a number of problems that were raised by the countries, justifiably so, in my view, about both the data and the methods and the framework that were used.
The publication of that report prompted a lot of concern on the part of the member countries, prompting the director of the WHO to ask for advice on what would be the best way to proceed in that whole area.
I think everyone was interested in describing health systems in the various countries and looking at characteristics that may contribute to health outcomes and health status.
I think that is something all of us around the table are already interested in. It is very difficult to do.
They did a fairly simplified version in this world report and I think it really concerned a number of people.
So, our international health office asked the data council to set up an expert group within HHS. We really have representation from all other agencies.
Actually, there is a lot of expertise within HHS on these issues. The data council did set up the group. It is operational now. It has provided some advice to our international health office.
They actually prepared a position paper on some of the issues, and the positions were used at the discussions at the World Health Organization executive board meeting in January.
Actually, we will share that position paper. We have it in the book here as well.
Ed Sondik, who will be speaking later this morning, I think, will go into a lot more detail about what exactly the WHO framework included and then what some of the issues were and areas we might proceed in the future.
In addition to that world health report assessment, the WHO was trying to bring together some related issues relating to data standards for the ICIDH-2, for how to standardize or at least harmonize measures of health status through questionnaires, and finally, standardized some approaches to summary measures of population health.
I think Ed -- I am not sure how those resolutions fared in the executive board meeting, but Ed Sondik will talk to you a little bit later about those.
At any rate, there should be a lot of activity in the international health area.
I think I may have just briefly reported this last time. I think all of you are familiar with our national surveys of substance abuse among youth that HHS sponsors.
They are very high profile surveys. They provide data on the level and the trends and the distribution of substance abuse among youth and among the general population, and they usually get a lot of press attention.
Actually, we have several of these surveys. I think each year when we release some of these findings, it is a bit confusing about how all these various trends fit together.
So, again, the data council was asked to set up a working group to look at the three major surveys that involve youth, to put together kind of a framework that describes how they fit together, to look for opportunities for coordination and integration, and to make some recommendations about how can we better explain a broad, comprehensive picture of substance abuse and youth in the country from all the surveys, a more comprehensive view.
Just briefly, let me talk about our data strategy work group and the budget. The budget process for fiscal year 2002 is drawing to a close, and probably within a month the President will send his 2002 budget up to the Hill.
I am happy to say that, over the past year, the data council has worked very closely with our budget office in terms of trying to get a collective, department-wide perspective on data improvements and investments that agencies might be thinking about.
This began when the data council developed kind of an overall framework for planning investments to improve data and information for decision making.
I think I mentioned that framework basically to you. The major themes were how do you address critical data gaps in health and human services, how do you promote and encourage information systems in health and human services for the future, the next generation, and how do you assure that the decision makers and consumers and others actually get the data, good data, when they need it.
This was part of the framework that agencies were provided. They were asked to think about, what sort of issues exist in this area and what kind of investments would be needed later in the budget process.
The agencies proposed a number of enhancements. Our data council strategy group actually pulled together all of those into a comprehensive cross-cutting analysis and then looked for a subset of those activities that would make the most difference in terms of a department-wide impact and addressing critical department-wide data needs, some of those.
So, that small subset of priorities was communicated to the budget office, which then asked the agency heads to give special consideration to that subset of activities.
They included such things as investments to stabilize and rebuild the NCHS core data systems. They included some robust approaches to increase capacity for state and local data.
They included some enhancements, for example, to the behavioral risk factor surveillance system.
They included some interagency initiatives aimed at a new generation of long-term care data and surveys and quality and patient safety data as well.
No one can predict what the ultimate outcome will be in the budget process. I think in terms of a planning and strategy approach, we were actually fairly pleased with this approach the past year, and we would hope to build on that for the year ahead.
Not only surveys were included, or surveys and data collection, but the NHII aspects as well, including support for HIPAA and some other activities at the National Library of Medicine that sort of support NHII activities.
Again, there is no way to predict what the outcome for the budget process will be. I think in terms of a planning process, this seems to work pretty well in bringing the agencies together across HHS on common data needs.
Let me briefly, if I have a minute or two, Simon, bring the committee up on responses to recommendations from the committee, and let me just work backwards here.
The NCVHS, you recall, sent in some recommendations relating to the ICIDH-2 process. That letter was received and it actually led to some intervention in the WHO.
I believe we have a letter back to you from the Secretary in your agenda book this morning.
This committee also sent a fairly detailed letter on implementation issues in HIPAA transaction standards and code sets. I think this was in late December, a very lengthy letter, a number of issues.
We actually took these issues to our health data standards committee, which is chaired by Bill and Richard Dell(?), and we have a similarly long letter response back to you that is in the agenda book this morning, that responds to some of those recommendations.
On the standards for patient medical record information, these were submitted, if you remember, in the summer and presented to our data council in, I think, August or September.
I think we sent a letter of acknowledgement back and then began a process of kind of looking at the recommendations.
It is a little more complicated because there were different kinds of recommendations. Some of them were very specific activities suggesting that agencies undertake this or that.
Others had to do with a much broader process and strategy for the next wave of standards and, really, for policy.
I think with the transition and the budget change, that has really slowed down some of those.
We actually had, after the data council meeting, asked all the agencies to look at the recommendations, to identify any issues and opportunities, and to identify any activities that they were taking or could take that addressed those 10 specific recommendations. There were a number of sub-recommendations, if you recall.
We also asked them to think of things that HHS, as a whole, collectively could do to advance some of the recommendations.
We have heard from just about all the agencies. Clearly, the agencies have a number of activities that are moving forward in this direction.
There were a number of reservations as well, understandably so. We pulled together a matrix that we are working through the data standards committee, and we have asked people to refresh and update what the agencies might be doing.
I think if I can just briefly react, I think many of our agencies are looking to the government computer-based patient record framework as a way to work on some of these standards. Several of our agencies are participating.
In addition, I think a number of our agencies view the whole HIPAA data standards process as kind of a framework and a platform for moving to the next step.
I think privacy and timing were still a concern. I think there was some reluctance to move forward on a wholesale basis even with potentially standards, until everyone was assured that we had adequate privacy and confidentiality protections in place.
At least that was mentioned. That was brought up several times about obviously keeping the big picture in mind as we moved forward.
Obviously, both legally and policy wise, HHS couldn't commit in advance to saying, yes, we will adopt whatever standard the committee sends.
I think we will have to see what standards are actually proposed and then evaluate them and see where we go from there.
I think there was a lot of support for the committee to sort of do fact finding in terms of its plan to see where the industry now stands on this, both industry consensus and industry readiness, and low hanging fruit, potentially, in that standards area for the months ahead.
I think everyone said, obviously, whatever the committee recommends, we always give serious attention to the recommendations.
Obviously you can't say you will adopt a standard. I mean, that is just not legally possible without seeing what it is and going through the process.
Finally, I think in a more sort of future-oriented vein as well, there is a growing interest in the NHII report, and in the whole area of NHII in terms of moving forward not just with the EMRI but with the NHII more generally.
Again, this will depend on a lot of policy decisions and budget decisions and other things later this year.
I think clearly now interest is growing and there is beginning to be increased press attention to the committee's report on the NHII, as well as what others are recommending, the IOM report and the President's PTAC and some other groups in terms of future work in these directions.
Obviously, they are not recommending that HHS take over this area. This is an area where there has to be close work with the industry and with the public health research community, with the vendors and with everyone else.
Clearly, there is more attention in this area. Let me stop there, and I will take any questions.
DR. NEWACHECK: Paul, are you aware of any further action in the department regarding our committee reports on Medicaid managed care or the islands and territories?
MR. SCANLON: I have asked our HCFA folks to look into that. Gary is not here this morning. I will try to get a report back.
DR. COHN: Any other questions? Dan?
DR. FRIEDMAN: I just want to have an acronym check. PTAC?
MR. SCANLON: President's Information Technology Advisory Committee. Ted Shortliffe, who is a member of this committee, actually is on that committee and co-chairs the health section.
They have recently been extended until June to finish their report. I think people are beginning to see reports from various commissions and sectors looking forward to the national health information infrastructure, or whatever we may want to call that.
It is basically the capabilities that information technology holds for improving health and moving forward. It is called the President's Information Technology Advisory Committee. Ted Shortliffe can actually brief the committee on that. We may even have a copy of their draft report.
DR. COHN: Other questions or comments? Mike?
DR. FITZMAURICE: Jim, I would note that on February 9, they put up on their web site a portion that applies to health care. I will get the web site address and give it to you.
MR. SCANLON: Okay, good.
DR. COHN: Okay, Dr. Braithwaite?
DR. BRAITHWAITE: Good morning. Some big activities since our last meeting. At the last meeting we announced the publication of the first final rule for transactions and code sets.
We were all excited about it, but out there in the industry nothing much seemed to happen, until on December 20 we announced that the privacy final rule was coming out.
It got published on December 28 and that seemed to be the kick start necessary to make HIPAA real out there in the real world.
We will be hearing our featured briefing this morning from Linda and Robinsue about that final rule.
The kick start has had a number of effects. The process that we call a DSMO, the designated standards maintenance organizations, that have been designated to maintain the standards, have set up their web site, become active, and are starting to get increasing numbers of suggestions for new standards and for changes.
That seems to be operating quite well although, according to some folks in the industry, not fast enough. This is always the lament, of course.
It is sort of one of those be careful what you ask for scenarios, where they ask for control over it and we give them control or we take control or whatever, and then it is not enough and then it is not what they really want.
Anyway, in addition to that, WEDI, the work group on electronic data interchange, held hearings in Chicago for the last couple of days, where I attended, to hear, in an open public forum, much like the hearings this committee holds, from industry folks, anybody who wanted to come and talk about what experiences and problems they were having implementing the HIPAA standards.
We got some very interesting feedback and the WEDI board will be digesting all of that and coming up with a recommendation, I presume, both to the committee and to the Secretary.
It is clear that, as more and more people delve into implementing these standards, they realize how difficult and complex it really is.
Their estimates of how long it will take keep ratcheting up. Since we have a statutory two-year implementation time frame, they are very afraid that they are running already two to three years late, before they had even started to actually implement much of what they had been investigating.
So, speed of implementation is a major issue and coordination is somewhat of an issue, particularly between privacy and security rules.
They feel, from the information technology side, they can't begin to implement the privacy rule until they get a security rule that gives them the means with which to implement some of these things.
They have some problems with the code sets decisions that we made in the final transaction rule and problems with a few of the data elements which they themselves built on a consensus basis as the implementation guide that we adopted.
So, the industry is going back to their own standards developing organizations and beating them about the head and shoulders to re-evaluate what they set as the standard that got adopted.
Everybody is worried that this process is going to take too long, so that people will waste a lot of time and money trying to implement something that is not reasonable to implement.
Meanwhile, the Secretary and the department are trying to get our people and act together to find a more rapid means of actually issuing, through the rule making process, a maintenance and modifications to the standards necessary to allow compliance. So, all of this is going on simultaneously.
It is mostly reiteration, but just to review, we have three final rules that we expect to put out this year, one for the provider identifier, one for the employer identifier and one for security.
We have three proposed rules that we expect to get out this year, one for the plan identifier, one for the claims attachment standard transaction, and one for the general enforcement of the other HIPAA standards besides privacy.
Because we don't have our full bureaucracy in place and things are not really rolling again yet, I can't really estimate when these things might come up and get published. But that is sort of the schedule that we are working on. Any questions?
DR. COHN: Actually, I have a question, because this is the first time I have actually heard you mention this year for the enforcement and compliance as a notice of proposed rule making. Are you confident that it will be out by year end?
DR. BRAITHWAITE: No.
DR. COHN: Just thought I would ask.
DR. BRAITHWAITE: This is the estimated schedule we are working on. As always, these things are totally unpredictable sometimes, and we are just trying to keep you informed as best we can, without committing to anything.
DR. COHN: Okay, Kathleen Fyffe?
MS. FYFFE: Thanks, Bill. There have been press reports over the past few business days that the effective date for the final privacy rule is going to be in April instead of February. Do you have any comment to make about that?
DR. BRAITHWAITE: I am aware of that and I think Robinsue and Linda will be addressing that in their privacy rule briefing.
DR. COHN: Okay, Kepa?
DR. ZUBELDIA: If the standard setting organizations and the DSMOs come up with a revised implementation guide to correct some of these data issues that are identified, and they do so in June like they are talking about, what will be the time frame to get that adopted by the Secretary as the current implementation guide?
DR. BRAITHWAITE: The statute allows the Secretary to update those standards within the first 12 months after they have been adopted, under the circumstance, where changes are required to allow compliance with the standards.
If the changes being proposed are required for compliance, then that is something the Secretary could move on very quickly to adopt those changes.
We are not allowed to change the standards more than once a year for sort of maintenance and general changes. So, we would have to divide those two up and see how that comes out.
DR. COHN: One final question. Bill, do you want to mention the changes that have come up as a result of the BIPAA law and how it affects the codes and the administrative financial transactions?
DR. BRAITHWAITE: No, I am not prepared to do that.
DR. COHN: Maybe we will talk about it this afternoon at the subcommittee, then. Okay, any final questions before we move on to the discussion around privacy?
We are sort of running a little late. Dr. Robinsue -- I am afraid I am going to ruin your last name.
DR. FROHBOESE: It is a very difficult one. It is pronounced Frohboese. I am used to getting many different pronunciations and long pauses after the Robinsue.
DR. FROHBOESE: I am very pleased to be able to meet all of you and to be part of this meeting this morning, to discuss the privacy rule.
I am actually new to HHS. I started in November as the senior career person within the Office for Civil Rights.
I am the deputy director in that office and, during this time of transition, am the acting head of the office.
What we would like to do this morning is both to give you a very broad overview of the final rule that was published in the Federal Register on December 28, as Bill said, as well as to just give you an overall sense of how HHS is moving forward on the implementation of this rule.
I would also like to explain how the Office for Civil Rights, as well as other components of HHS, fits into the whole privacy rule context.
As you know, ASPE took the lead on actually drafting the regulation and brought together experts from across HHS to work on drafting the proposed rule and certainly reviewing in detail the more than 50,000 comments that HHS received on the proposed rule, and then finally to put together the rule that was published in December.
Linda Sanches, who many of you may know, was integral in the group that was responsible for drafting the rule. She will cover an overview of the rule.
I know that tomorrow we are also scheduled to meet with the privacy subcommittee. We are looking forward to being able to discuss perhaps some more nuts and bolts of the rule, as well as how we can work with the privacy subcommittee as we work on implementing the rule.
I actually was not involved in the drafting of the rule. However, I was in the Justice Department in the civil rights division, where I was involved in the Justice Department perspective in providing input to the rule.
My training and background is actually from a joint degree program, a JD PhD program, in law and psychology.
So, I have much experience and sensitivity to the issues that I know this committee, as well as a number of others, have really been grappling with in trying to develop this privacy regulation, to get the right balance between protection but, by the same token, allowing providers to do what they need to do, and researchers to collect the kind of data that they need to collect.
We hope that we really have been able to strike that balance.
As you may know -- and I think as one of the questions earlier got into -- there has been a lot of recent publicity about the rule.
There have been a lot of letters that have come in to the Secretary to put this issue at the top of his agenda.
The letters have really been mixed. We have received a number of letters that have expressed concerns about the rule and particular areas about the rule.
We have received a lot of numbers really praising the rule and urging the Secretary and the department to move forward with all due speed.
The rule was scheduled to go into effect this Monday, February 26, which was 60 days after it was published in the Federal Register on December 28.
However, a glitch developed. GAO notified the department recently that Congress actually never received a copy, formally received a copy of the rule, and some required information about implementation of the rule.
So, another package was sent over to Congress last week. It was received on February 13.
What this means is that, under the Congressional Review Act, a federal statute which says that major rules will go into effect either 60 days after they are published in the Federal Register or receipt of the necessary information by Congress, the later of those two dates.
Under that federal law, it means that the privacy rule will not go into effect now until April 14.
So, that is about a 45-day period longer than we had originally anticipated.
As you also know, however, the actual enforcement of the rule was not scheduled to go into effect until about two years of the date of actual enactment of the rule for most providers, and three years for small providers.
What the department has been focusing on during this interim period is public education, outreach and technical assistance to both inform consumers about the rule, as well as providers, as to their responsibilities and how to come into compliance with the rule.
When the rule was published in the Federal Register, then-Secretary Shalala delegated the authority for enforcement to the Office for Civil Rights. So, that is how we fit into the overall scheme of the privacy rule.
What we have done since the rule was published is to really capitalize on the expertise within the department and the individuals who were responsible for drafting the rules.
So, we have really kept intact the body of individuals who were the experts in putting the rule together.
So, Bill has been very involved. Linda has been very involved. We pulled together people from ASPE, people from ASMB and other components -- NIH -- within HHS, to assist OCR during these initial phases, in getting the rule off the ground.
We also have been very involved with the Data Council. John Fanning and Jim Scanlon have been very helpful to us as we think through a lot of the issues about implementation.
We have met with the Data Council. We have met with the Data Council's privacy committee.
We will be using the privacy committee as a tool to facilitate both intra-agency as well as interagency implementation of the rule.
For example, I think next week we have a meeting where John has pulled together representatives from the Social Security Administration, from the Department of Veterans Affairs and the Department of Defense. We will be meeting with these entities to discuss implementation.
All of that has been very valuable in terms of ensuring coordinating as well as collaboration, both within the department as well as with the other federal agencies.
As you also may be aware, several weeks ago the Senate Health Committee held oversight hearings on the privacy rule.
At that hearing, GAO provided testimony. If you haven't seen that testimony, we would be happy to give you copies of it.
It is, I think, very useful and important testimony in terms of providing a broad overview. That was the thrust of GAO's testimony.
In addition to that, there was a panel of providers as well as a panel representing consumer groups.
In addition to all of the activities that are going on with coordination and also on the Hill, OCR has been very active in trying to get out information, both to the public, as well as to providers and various associations, trade groups.
We have done that through several different mechanisms. When the rule was published, we established a privacy hot line, a privacy resource line with a toll free number.
The toll free number is 1-866-627-7748. That number is staffed by individual staff from the Office of Civil Rights, who answer general questions about the rule and then refer specific questions to a panel of experts that we have compiled that really is basically the individuals who drafted the rule.
To date, we have received close to 800 calls on that line in the past couple of months. The calls break down roughly -- I would say that initially most of the phone calls we received were from consumers, wanting to know when their rights will be protected and when this rule will go into effect or where they could get a copy of the rule.
As time has gone on, we have received more and more detailed questions from basically providers or associations who have specific questions about how the rule affects them.
So, we have been relying more and more on the experts to return the phone calls.
In addition to this resource line, we also have a web site. That is at www.hhs.gov/ocr/hipaa.
That web site has a copy of the rule and also we have been receiving a number of questions via e mail, to date about 250 e mails.
We have been in the process of analyzing these e mails to then develop frequently asked questions and responses that we will be posting on the web.
Along with these activities, we also have working groups that range from a speakers bureau -- we have received many, many requests for presentations across the country.
To date, we have probably made about 40 presentations, with many more scheduled for the upcoming months.
We, again, are drawing upon the expertise of individuals who drafted the rule, to make these presentations.
We also have a policy subcommittee, and that subcommittee is examining questions that are arising through multiple sources, through letters that have been written, through the e mails, through questions that we get on the resource line.
We are examining also areas where the rule may need to be retooled. As Bill had mentioned, the Secretary has the discretion, during the first year, to make amendments to the rule.
We really do want to know if there are parts of the rule that need to be fine tooled to make it workable. So, any suggestions along those lines that the committee has would be very helpful.
We also have a strategic planning committee. That committee has a two-fold purpose. One is to put in place the kinds of technical assistance, materials that we will be developing in the short term to be able to help providers come into compliance with the rule.
We really do want to focus on getting out information that will be very helpful and will ensure that covered entities can come into compliance.
In addition, that committee is working on an enforcement plan for when we do start enforcing the rule several years from now, of exactly how OCR will be conducting both complaint-driven as well as compliance reviews to determine compliance with the rule.
I think that that gives you an overall sense of what is happening on the department level. I do want to open it up for any questions that you might have, and then turn it over to Linda, who has a very slick power point presentation that she has packaged and taken across the country.
All of you will also be able to see the kind of overview that we have been presenting. Are there any questions right now?
DR. COHN: Kathleen Fyffe?
MS. FYFFE: You have mentioned that the Secretary has the discretion during the first year to modify the rule.
Given that so many pieces of correspondence have come into the Secretary's office, in general, what would the process be for the involvement of a national committee to advise the Secretary that there be a potential modification?
DR. FROHBOESE: I think there very definitely is an important role for the committee to play. I think tomorrow, when we meet with the privacy subcommittee, we can talk about some particular issues.
We really do want to hear from the committee if there are areas of concern or suggestions that you have, on how to make the rule more workable. So, I think there is a very important role for the committee in its advisory capacity.
DR. COHN: I was actually reflecting that, for the administrative and financial transactions, the subcommittee on standards and security, I think, has been very helpful to the department in terms of helping focus issues where retooling may be indicated.
DR. FROHBOESE: Absolutely.
DR. COHN: Obviously, we will talk about that tomorrow in the subcommittee.
DR. FROHBOESE: Sure. I also know that the committee was very helpful as the rule was being drafted and as various ideas were bounced off the committee when the group would come up against complex or thorny issues. I think that there is a really important role there.
DR. COHN: Other questions or comments?
Before we move on to Linda, Robin, I want to thank you.
DR. FROHBOESE: Thank you.
DR. COHN: Certainly the work that you have set out for you is very impressive in scope.
DR. FROHBOESE: It is ambitious to say the least. We have a lot on our plate right now.
DR. COHN: Linda?
MS. SANCHES: Thank you. First, I am wondering how to turn this on for you.
I am sorry to disappoint you. It is actually not very slick. I have yet to figure out how to make all those sort of nifty colors and exciting backgrounds appear. I am open to technical assistance afterwards, if anybody wants to tell me how to do that.
I am going to go over several topics today, and I realize we are running a little bit behind in the agenda, so I will try not to dwell too much on any one topic.
I will be talking about who is covered by the rule and what some of the new individual rights and requirements for covered entities are contained in the regulation, some new requirements for business associates and what our enforcement capacity is.
I think you all are probably very familiar with the background. This is a result of HIPAA. It is from subtitle F, administrative simplification.
I just want to note, again, that we did get over 52,000 comments and we had a very fun year last year reading all those comments.
We did read them all. Some were very short. Some were only a paragraph. Some went on for well over 100 pages and were from people in this audience. Thanks very much. We needed the help.
Obviously the compliance data on that slide is wrong, as you just heard.
So, who is covered by the rule? As you probably already know from the data standards discussions -- there is some issue with the projector. I don't know if anyone is here who can work with that.
The people covered by the rule are health care providers who transmit health information in electronic transactions, the ones that are standardized by the rule, health plans, which is a defined term in the statute, and health care clearinghouses. I hope you have copies.
So, what is covered by the rule? We cover protected health information. What we mean by that term is individually-identifiable health information that is transmitted or maintained in any form or medium.
As you may be aware, in the proposed rule, we only covered information that had been part of the transaction at some point, or had ever been electronic.
That is not the case in the final rule. We cover any information that a covered entity holds.
We also cover that information if its business associate holds the information.
We do have special provisions in the rule for de-identifying data. This is a method by which covered entities can take information, get rid of the identifiers and, therefore, it is no longer subject to these protections.
I want to make some key points, because these are issues that I think have been subject to some misunderstanding out in the world.
The only required disclosures in the regulations are disclosures to the individual who is the subject of the information and disclosures to the Office of Civil Rights for compliance purposes.
We discuss a lot of other uses and disclosures in the regulation themselves, but those are only permissive. They are not required by the rule itself.
We basically work through a lot of uses and disclosures that are common in the health care industry today and set rules and guidelines for those types of uses and disclosures, but we don't require them.
As I am sure this group is aware, this is the first comprehensive set of health privacy protections that affects most every provider, all health plans, et cetera.
It sets a national floor. We do not preempt more stringent or more protective state laws. That, I know, has been the subject of some consternation by many of the people who are subject to this rule, but that is part of the statute itself.
As you are thinking about these issues, keep in mind that the rule affects disclosures and the uses of information and the relationships between covered entities and others who might receive health information from them.
The regulation sets up a set of rights for individuals who are the subject for health information.
The first is the right to inspect and copy their own information. The second is a right to amend that information if it is inaccurate or incomplete.
Individuals have a right to obtain from a covered entity an accounting of any disclosures that were made of their information for purposes not related to health care, not related to treatment or payment or running the covered entity itself.
For any other types of disclosures that are made, upon request, they have a right to receive an accounting of all those other disclosures.
They have a right to have a covered entity accommodate their reasonable request for confidential communications.
What that means is that if they would prefer to be called at one phone number and not another one, that that request should be accommodated.
They have a right to file a complaint with the Office for Civil Rights or with the covered entity itself, if they believe that their rights have been violated or any of the requirements of this rule have been violated.
They have a right to obtain, from providers and health plans, a written notice of the information practices of those covered entities.
What that notice would include would be statements of how their information is used and shared and what their rights are.
So, what does the rule do? In general, we say that all uses and disclosures that a covered entity may make must be limited to what we actually permit in the rule.
In the rule we set a variety of types. We classify the uses and disclosures by various ways of thinking of them.
Not everyone has agreed with the way we put them together, but it made sense to us at the time and we hope it continues to make sense years from now.
The first would be, we have special rules for uses and disclosures for treatment, payment and health care operations purposes.
We have a separate set of requirements for uses and disclosures for which an individual needs to have an opportunity to agree or object to a use and disclosure before they are made.
We have a set of criteria for allowing disclosures for some specific public purposes.
Anything else that doesn't fit in those requirement require that the covered entity obtain a written authorization before that disclosure can be made.
As I mentioned, there are special requirements based on the type of use or disclosure that is going to be made.
I wanted to point out, the first category were uses and disclosures for treatment, payment and health care operations.
Treatment, I think people can somewhat guess what that is, although it is a defined term in the rule. It is actually providing health care treatment or services or managing their treatment.
Payment has caused a little more confusion. Payment includes a variety of activities related to paying for health care, and that includes utilization review, review of services for medical necessity and determinations of coverage.
Health care operations is also a long list of related activities, that are related to providing treatment and payment, and actually running the entity itself.
Health care operations include certain quality assessment activities, improving the competence of health care professionals, accreditation, medical review, fraud and abuse protection programs, and certain customer service uses.
For any of these uses and disclosures, covered entities must apply a minimum necessary analysis.
What does that mean? This is actually greatly revised from the proposed rule. So, if you knew what was in the proposed rule, put that aside for the moment.
In general, covered entities must restrict information that they use or disclose to the minimum amount necessary to accomplish that purpose.
This requirement does not apply to disclosures to health care providers for treatment purposes. We did this because there was some concern that there may be inappropriate analyses of what is minimally necessary for a provider.
We didn't want to, in any way, cause any concern about what could actually be released to a health care providers. So, disclosures to health care providers for treatment are not subject to the minimum necessary requirement.
For internal uses within a covered entity, we require that covered entities establish what is often called a role-based access system.
They need to establish policies and procedures by which they would identify in general the types or classes of workers they have, the types of information that these workers need to successfully accomplish their functions, and any conditions for access.
That would mean, for instance, that a hospital might decide that all billing clerks could have access to certain information but not, for instance, the entire medical record.
Physicians could have access to the entire medical record of anyone who is being seen in the hospital where a record might be kept of everyone's records they requested, if they were not actually their own patients.
These are the types of policies that need to be worked out. Covered entities do not need to go through individually each piece of information and make a decision for each and every use and disclosure, is this the minimum necessary for this particular reason I am giving this information to this nurse or to this filing clerk.
They need to set up general policies and procedures and they need to make sure that they are followed.
For disclosures of health information -- i.e., sharing of information outside the covered entity itself -- they need to, as well, set up general policies and procedures for routine disclosures.
If they routinely are sending bills to a particular health plan, they can set up general guidelines for how they send information to that health plan.
For non-routine disclosures, they need to set up procedures by which they are going to review those requests for disclosures, or the way they are going to make those disclosures and they need to actually individually review those disclosures, if it is not a routine, common occurrence.
The minimum necessary also applies to requests by covered entities for information to other covered entities.
That means that if a health plan is requesting information from a health care provider, they need to only request the minimum necessary, and they need to set up policies and procedures to determine, in general, what the minimum necessary is.
Providers need to establish what is the minimum necessary to disclose. If the request is from another covered entity, they don't need to -- they can rely on what the other covered entity has requested.
If a health plan requests information from a provider, the provider does not then need to figure out, well, is this the minimum necessary that this other health plan needs. They can rely on the fact that the health plan is subject to this rule and is requesting what is minimum.
I want to note that the minimum necessary does not apply to disclosures required by law.
In the final rule, we require written consents in certain situations. A written consent applies to treatment, payment and health care operations, or uses and disclosures for those activities.
Consents are required of health care providers for any direct treatment relationship with an individual.
Health care providers who are not in a direct treatment relationship or health care providers in an emergency treatment situation and other covered entities are not required to obtain a written consent in order to use and disclose the information.
So, what is in the consent? A consent is a one-time-only form and it is fairly brief. It informs the information that protected health information may be used or disclosed for treatment, payment or health care operations purposes.
It refers the individual to the notice of information practices, so that the individual can learn more about how the information is actually used or disclosed by this covered entity.
It notes that the individual may request restrictions on how information is used for treatment, payment or health care operations.
The consent may be revoked, and a covered entity may condition treatment on obtaining the consent.
I want to contrast that with individual authorization. I mentioned earlier that, for uses and disclosures that are not for treatment, payment or health care operations, or that are not for another class I am going to talk about soon where they are required to get some sort of statement of agreement, or if they are not for public purposes, a written authorization is required.
That includes anything else that doesn't fall in those general categories -- pre-employment physicals, employment determinations, life insurance, a request made by the individual themselves for the information to be disclosed.
For all those things, an individual authorization must be obtained before the information can be used or disclosed.
A written authorization is much more detailed than a consent. It needs to specifically lay out what information is going to be disclosed or used, who the recipient of the information will be. It must have an expiration date.
All uses and disclosures must be consistent with the authorization, and a covered entity may not condition treatment on the individual authorizing disclosure of information for other purposes.
A health care provider cannot condition providing treatment on the individual agreeing to have their name sold to someone else for some other marketing purpose, for instance.
Opportunities to agree or object. There is a small set of disclosures or uses for which we believe it is important for the individual to provide some sort of assent where possible, but where it didn't make sense to actually require written documentation.
The first of those is for facility directories. When an individual is going into a hospital, often their name will go up on a list.
If someone calls in, they can then find out where the person is. They can come visit. They can have flowers sent. They can get a general statement of how the person is doing.
For those types of disclosures, a written consent or authorization is required. However, when someone is coming in, they need to be provided with an opportunity to agree or object.
They need to be told, your name will go into the facility directory; is that all right with you. The person can then say yes or no.
If a provider wants to disclose information to a person involved in care of the individual, someone who perhaps is picking them up to take them home, their spouse or companion, someone who might, in fact, be taking their prescription to a pharmacy to be filled, for that, a written consent or authorization is not required.
They merely need to make an attempt to get an oral consent from the individual.
If the individual actually brings the person into the consulting room with them and says, hi, this is my sister and I want her to be here, you have pretty much got an oral consent to having information disclosed.
If someone else walks up in the hallway and says, hi, how is my cousin doing, if you can, you should probably find out from the patient first whether it is okay to be disclosing information to everyone in the hallway who asks about your condition.
The last category is disaster relief. What this means is that in disaster relief situations a health care provider may provide information to, say, the Red Cross so that you can be located, so that they can tell your family members where you are, that you have survived and that you are actually in the hospital, for instance.
For those types of situations where it is possible to get your assent, they need to attempt to. Otherwise, they can use professional judgement as to whether the disclosure can be made.
We have a long list of specific what we call public purposes, for which uses and disclosures can be made without the individual's consent or authorization, and we have conditions for all of them.
We have long discussions of each one and definitions of all these terms. For more detail on all of this, I suggest you actually read the rule. It does get rather complicated.
Covered entities may disclose as required by law. They may disclose for public health purposes, to avert serious threats to health or safety, for health oversight purposes, for research, for law enforcement in certain conditions and there are a few others as well. For all these, no authorization is required.
I want to talk a little bit more about a few of them that are probably of interest to this committee.
First, what is a disclosure that is required by law? Required by law means that there is a mandate in law that compels a public entity to actually use or disclose information in a certain way.
These include Medicare conditions of participation or statutes or regulations that require information to be disclosed if payment is sought under a government program providing public benefits.
What is health oversight? Covered entities may make disclosures to health oversight agencies and also to their contractors and agents, for oversight activities that are authorized by law and that actually require the information, e.g., for compliance with government programs.
I want to point out that one entity may be both a covered entity -- for instance, a government health plan like Medicare -- and a health oversight agency.
That can get a little complex in the public sphere because of this.
We allow disclosures for public health, and the public health provision is actually quite extensive.
First, we allow disclosures for the kinds of things that you always commonly assume are public health, to prevent or control disease, injury, disability.
We also allow disclosures to report adverse events, product defects, to track products, to enable product recalls.
Some of these are within the general discussion or purview of FDA, but they may not in fact be disclosures to FDA.
They may be disclosures to others who generally gather this type of information. Those are discussed at length in the rule itself.
Disclosures can be made to persons as authorized by law in a public health intervention or investigation. In a certain small set of circumstances, public health disclosures can be made to employers for medical surveillance of the work place or work-related illness or injury.
Disclosures can be made without authorization for research in a circumscribed set of circumstances.
The first is if the researcher obtains approval from an IRB or a new entity is sort of created in this rule called a privacy board.
There is a set of criteria that need to be determined by this group. A couple of them are the fact that they have determined that the privacy risks are reasonable in relation to anticipated benefits from the research itself, and that the researcher has a plan to protect the identifiers of the individual.
Disclosures can also be made to researchers so that they can review the information to actually prepare a research protocol, and for research on the health information of people who are dead.
For these last two types of disclosures, the researchers need to provide certain representations regarding how they are going to use the information, and there is certain documentation that is actually required.
The regulation sets up a set of safeguards within covered entities. It requires them to establish minimum necessary provisions, to set up policies and procedures to safeguard information, to set up a bunch of administrative procedures to safeguard information.
A lot of work is actually done in the health care industry that is outsourced, that is done by others, by contractors, agents or others.
To protect that information when it goes to people working for covered entities, we have created a set of requirements for what we call business associates.
Business associates perform functions on behalf of a covered entity using or disclosing identifiable health information, or they may provide services, such as accreditation services, that involve using identifiable health information for the covered entity itself.
When a covered entity engages a business associate, they need to obtain satisfactory assurance that the business associate will safeguard that information.
I wanted to point out that there are a lot of relationships between health care providers for treatment purposes, where it gets a little bit difficult to figure out who is actually working for whom.
So, we basically have stated that you do not need a business associate contract or assurances for these treatment purposes between health care providers.
So, a covered entity needs to get satisfactory assurance from the business associate that they will safeguard the information.
What is satisfactory assurance? In general, it means a contract. The contracts need to establish permitted uses and disclosures of the information that is needed for these functions.
It needs to make certain assurances, that the business associate will make no other uses unless required by law, that they will use appropriate safeguards, that they will report to the covered entity any other covered uses and disclosures they discover have been made by their employees or subcontractors, and that at termination they will return, destroy or limit further uses.
There are also some other requirements regarding -- there are certain rights that individuals have regarding accounting for disclosures and access to information. Those rights continue if that information is actually held by a business associate.
So, the contract needs to state that they will, in fact, honor those obligations.
Now, I said in general these assurances mean a contract. There are certain situations among government entities where a governmental entity may have a business associate that is another governmental agency.
In general, we don't have contracts with each other. In general, we have memorandums of understanding or we can actually use regulations that establish in law these same requirements.
The rule would allow, when you have two government agencies, one is the covered entity and the other is a business associate, that they can use a memorandum of understanding or a regulation, if those documents actually contain the same type of requirements that would be in the contract.
A covered entity is generally responsible for the actions of a business associate, if they know that there has been a violation of the agreement and failed to act.
What that means is that, if the business associate is actually doing terrible things with the information -- they are selling it on the side or posting it on a web site and this was not allowed in the contract -- the covered entity is responsible for that when they discover that it is occurring. If they do nothing about it then, in fact, they are in violation of the rule.
The statute itself actually establishes certain monetary and criminal penalties. The civil monetary penalties for violation of this rule are $100 per violation, capped at $25,000 per year per each requirement that is violated. This will be enforced by the Office of Civil Rights.
There are also criminal penalties for certain violations. The penalties increase depending on just how bad you were, what your intent was in terms of using or disclosing the information, than what you did with it. These criminal penalties will be enforced by the Department of Justice.
In terms of thinking about what this all means, obviously people need to go and read the rule. The rule itself is not very long.
It is just the preamble that goes on for many, many pages. The rule itself can be read in a single sitting without too much difficulty.
Covered entities obviously need to follow these rules and they need to think about their relationship with others.
There are lots of entities in the world, though, that are not covered entities, but will be affected by the rule.
I think everyone in the health care industry needs to be thinking about what their relationships are with covered entities, and what are the purposes of the disclosures that are made, because as you notice, there are different requirements depending on the purpose of the disclosure.
A covered entity could be making disclosures to a particular entity for operations purposes, for payment. It could be required by law. It could be for health care oversight.
It can get confusing. I think a lot of groups need to really sit down and think through how they get information today from covered entities.
They could actually, if they are lucky, just be a recipient of information under 512. They could be a public health authority. So, information can come to them if they meet the requirements.
They may be a business associate if they are actually doing work for the covered entity, whether or not they are actually paid for it.
They could be a partner in an organized health care arrangement, which I haven't gone into and I can take questions if you would like.
Basically, we allow disclosures among participating covered entities if they are involved jointly in quality assessment activities.
If you are involved in joint quality assessment activities, you need to figure out if you meet this criteria.
There is a lot to think about in this rule. If you want to think about it with some documentation, you can go to the web site, which is listed on the back page, or you can call the telephone numbers that are listed here.
I am happy to take questions now. I realize that there was a lot of information and you didn't actually have a visual that you could rely on.
DR. COHN: First of all, I want to thank you for your presentation. As one who has actually read the reg now, and literally only made it through a couple hundred pages of the preamble, I actually really appreciate your trying to distill this.
Actually, I am reminded that this is complex. The many hundreds of pages of regulations are influenced by the information in the preamble.
I suspect that your FAQs are going to be really critical, because I think there is a lot of confusion in the industry about exactly how this all plays out. That is one comment, obviously FAQs are going to be really important.
Questions? Barbara?
DR. STARFIELD: What is the biggest change over the current situation? Who is going to be most impacted and in what way?
For example, I go in to a physician and I sign a form which allows that physician to disclose the information to a payer. What is the major impact?
MS. SANCHES: What is the biggest change? I think it sets for the first time requirements that the information be protected.
While a lot of providers may now be doing very similar things -- they may be protecting the information, they may be providing their patients an opportunity to provide consent -- that is not necessarily standard across the United States. This will set, for the first time, a standard of safeguards.
I think what is also different is that it does provide individuals with these rights to gain access to their records.
I have been amazed how many people have come to me since we put the rule out asking exactly how they can get access, because they have been denied access to their own records.
I think for individuals it provides them with certain rights to get information that they didn't have before.
Hopefully it will provide them a little bit more insight into how their information is actually used.
DR. COHN: Other questions? Kathleen and then Vickie.
MS. FYFFE: Thanks, Linda. For purposes of the folks that are on the internet, can you say what the new compliance date is?
MS. SANCHES: Sure. This is really late breaking news, that the implementation date now will be April 15. As of late last night, the papers were still being prepared to be sent to the Federal Register, where there will be a notice of the new enforcement date. I think we are trying to make sure it coincides with the public notice.
MS. FYFFE: So, the new compliance date, except for small plans, would be April 15, 2003.
MS. SANCHES: I believe that is right.
MS. FYFFE: Okay, thank you.
DR. MAYS: I am a little naive to some of this, so I want to ask a question. In terms of the research, when you talked about going to either an IRB or privacy board, can you tell me two things? What is a privacy board and two, what is really the change for a researcher.
MS. SANCHES: I was really afraid I was going to get a research question.
A privacy board is something that was created by this rule. It doesn't exist today. It is basically analogous to an IRB.
We are basically saying an IRB-like structure for reviewing privacy issues should be extended to all research, whether or not it is federally funded. Right now, an IRB is only for most publicly funded research.
The privacy board requirement would apply to all research done in the United States, no matter how it is funded, if it uses protected health information.
The privacy board concept builds on an IRB. It contains many of the same requirements and processes, but it adds certain questions regarding whether or not -- well, it adds questions that don't exist today regarding whether the privacy risk is outweighed by the benefits of the research itself.
So, it adds a few other questions that need to be looked at by an IRB. Then, if an IRB doesn't exist because it is in a private sector situation, then an IRB-like structure, the privacy board, needs to be created. Did that satisfy your question?
DR. MAYS: I need to think some more about the issue of burden and to also understand whether or not it is requiring the generation of another structure that, indeed, one would have to submit to.
DR. FITZMAURICE: Could I elaborate, Simon, a little bit on that? I think what this does, we are kind of familiar with IRBs as they apply to federally-funded research.
What this does is get a set of criteria to IRBs that cover privacy. So, it is separate from the common rule that describes IRBs.
In addition, those in the private sector who undertake research using individually identifiable health information, and they want to get it from covered entities, have to set up a privacy board, which is an IRB-like structure.
It has to be divorced from conflicts of interest. It has to have at least one independent person on it. It has to approve the waiver of getting an individual authorization.
When you come to a covered entity like a provider and you say, I want to get access to patient records, you either have to present an IRB waiver of the individual authorization or you have to present a privacy board waiver of the individual authorization.
For privacy, the criteria are the same for IRBs or for the privacy board, and they are listed in the privacy rule.
DR. MAYS: Then who does the privacy board report to? We know who IRBs report to, and there is a whole set of regulations for IRBs. Then who does the privacy board -- particularly if we are talking about the private sector, independently set up?
DR. FITZMAURICE: That is left up to the private sector. It could be an independent privacy board. It could be a privacy board set up by a particular organization, say, a particular health plan or a particular research foundation, for example.
DR. FRIEDMAN: Could you comment on the circumstances under which treatment can be conditioned on the provision of consent by the individual patient?
MS. SANCHES: Yes. The consent is for any use or disclosure of information, for treatment, payment or health care operations.
If a provider is not able to get the consent, then it is difficult to see how they could provide any treatments, if they are not able to use any information, and that includes using the information to set the next appointment or to actually look at them or make a diagnosis. They are not allowed to use the information. They cannot disclose it to someone else.
In those situations, they may refuse to provide treatment if the individual won't consent to allow them to use the information to provide the treatment.
There are exceptions to this for emergency situations or where a provider is required by law to provide treatment. Then the consent provision is set aside, and there are rules for attempting to get it anyway.
In general, the reason the consent -- you are allowed to condition treatment in this circumstance, because it is unclear how you could provide treatment without obtaining the consent.
DR. STARFIELD: I want to give you a little vignette, and it is very complicated and involves some of the things that you said.
A health insurer asks an employer to provide information on the percentage of employees who have various medical conditions -- the percentage who have heart disease, cancer, diabetes, et cetera, et cetera, not individually identifiable.
In order to get that information, the employer has to get individually identifiable information; right? You have to get information from the employee.
You have got to either consult records or do a survey or whatever; is that not correct? How is an employer going to get that information except by contacting individual people.
MS. SANCHES: I am not sure what your question is, but employers are not covered by our rule.
DR. STARFIELD: Employers are not covered.
MS. SANCHES: Employers are not covered.
DR. STARFIELD: That answers it.
MS. SANCHES: The statute, as you recall, covers plans, providers, clearinghouses. We do have special rules about the interaction between group health plans and employers.
Basically, we all know that employers who sponsor health plans are running these health plans to a certain degree and need access, to a certain degree, to some information, depending on the circumstances.
This would limit how the employer would then use the information for purposes other than actually running their health plan.
So, we do deal with employer access in regards to their relationship with their health plan in that way.
I think you are asking a slightly different question, though. You are suggesting that if an employer is actually perhaps seeking coverage by a health plan, they may need to provide some information to the plan.
DR. STARFIELD: For setting of premiums that mostly would occur.
MS. SANCHES: Again, we don't actually cover employers. An employer can have whatever interaction they want with their employee.
MS. COLTIN: I just was interested in the bullet on your last slide about what I think you called partner in an organized health care arrangement.
Could you just say a little bit more about what that is, what you envision that looking like?
I didn't know if you were talking about sort of vertically organized systems of care, like physician hospital organizations, or if you were talking about public private partnerships and entities that might engage in sharing information for purposes of measuring or monitoring quality.
MS. SANCHES: It is an interesting question. The general rubric for the rule is that you can only disclose information for treatment -- again, this is under the consent -- for treatment, payment or health care operations.
It is your own health care operations. That would mean that, for quality improvement or quality assurance purposes, you could use that information for your own quality program, or you could hire somebody else and you could disclose it to the person you hired to do your quality improvement program.
Often there are systems that are integrated. As you noted, you could have two organizations that work very closely together that want to share information for these quality improvement purposes.
What we have done is, we have said that part of health care operations involves disclosing when you are part of these organized health care arrangements where, in fact, you do all this work together and you do quality assurance together.
There are a couple other types of organized arrangements that are listed in the rule itself.
The idea is to allow disclosure for these purposes when groups are integrally linked but are actually separate organizations, where it makes a lot of sense to share the information.
DR. COHN: I think you resemble one of those, actually.
MS. COLTIN: Dan and I are part of one of those types of arrangements. I was interested in how this would apply.
DR. COHN: I realize that we are going a little bit over, but I think it is really important that everyone understand as much as they can about the privacy final rule.
DR. ZUBELDIA: Linda, I have a question, the minimum necessary issue. How does that interact with the transaction requirements?
We have heard extensively about the transaction requirements being very extensive, way beyond what is being used today.
If a provider files a claim with the health plan using the standard transactions, do they have to be concerned about the minimum necessary, or just by following the standard implementation guide they are automatically complying.
MS. SANCHES: Actually, that is a good point. I should add that to my examples. Required by law also includes if it is required by another regulation.
If a data element is required in a transaction that we have adopted, the minimum necessary does not apply. It only applies if it is an optional element, which may come up in the future.
Right now, I think for the transactions that are out there, minimum necessary does not apply. I think the claims attachment might raise other issues.
DR. NEWACHECK: Can you explain the process by which a consumer would be able to inspect their own medical record and make amendments to it?
MS. SANCHES: I can't go too deeply into the process because I didn't review it just before this meeting.
They are allowed to request access and they must be given access. If they request that copies be made, there are time frames by which that request must be met. The individual -- they can be charged for making copies that is a reasonable cost based charge. There are time frames for meeting all these requests.
In terms of making amendments or corrections, they can request amendments or corrections if the information is incomplete or incorrect.
The provider or plan, if it is incomplete or incorrect -- I am stalling because I can't remember exactly how the language goes in the reg text -- should make the change.
They may refuse, but then if requested, they need to put down a statement of why they are not making a change.
For instance, if the individual says, actually I was born in 1964 and not 1968, it seems unlikely to me that a provider is not going to want to make a change on a fact-based issue.
I think there are cases where there will be concerns where the individual might disagree with a diagnosis or might feel like the information doesn't adequately reflect everything that happened.
There it is more likely that the provider, if they refused to make the change, would need to put a statement in about why they are not going to be making the change.
They don't need to make the change if they don't agree with it. They do need to provide a statement as to why they are not doing so.
DR. COHN: I want to thank both of you for coming and joining with us and providing a briefing on this. As I understand it, we will be seeing both of you tomorrow morning bright and early about 8:00 for the subcommittee on privacy and confidentiality.
As one who is from California, I am physiologically challenged at that hour on the east coast. I do understand your issues. I do want to thank you both for being here. We will see you in the morning.
With that, we will break for 15 minutes. We will come back at five minutes after 11:00.
[Brief recess.]
DR. COHN: Dr. Sondik, thank you for joining us to give us an update on the WHO executive board meeting.
Agenda Item: Report from WHO Executive Board Meeting.
DR. SONDIK: Thank you. I am happy to do that. What I am going to talk about today has to do with the world health report. I hope everyone here has seen this, has a copy of it.
It is entitled, World Health Report 2000: Health Systems Improving Performance.
This document has generated a great deal of interest, really, across the world. I was part of a delegation that attended the executive board of the World Health Organization.
There are really two topics that are interrelated that came up at the executive board. One has to do with this document and the second has to do with a framework that is very much related to the document which is essential for a continued evaluation of health status and health systems.
Let me give you a little bit of background on this, although I think most of you are quite familiar with this.
This document is, in essence, a framework for improving health system performance. It does this by guidance on stewardship and the function of leadership of a country in its health care system.
The bulk of this document is actually concerned with those types of issues. What has gotten the most attention is the analytic framework that is included in this document, that culminates in a ranking of countries, or countries' health care systems, by a measure of efficiency or performance.
Now, there are three -- actually a number of issues related to this. Let me summarize them briefly.
The first issue that the executive board addressed had to do with -- I shouldn't put it in those terms, because it sounds as if they were addressed one by one.
One of the issues that was addressed was the analytic framework itself, and I will spend a little bit of time on that, and the question of its validity, the validity of its efficiency measure.
The second had to do with data. This document ranks all 191 member countries of the World Health Organization.
At the time the report was put together, they had data on 30. So, they extrapolated from the 30 in a variety of different ways, and used these extrapolations, estimates, imputations, whatever words you want to use here, to come up with the rankings.
The third point had to do with the release of the report. It turns out that member countries, member states, did not see the report prior to its release.
They also had not had any kind of a discussion of the framework itself, which I must say I was totally unaware of any of this until the report was actually released and Chris Murray, who led the effort, came to the United States.
I just assumed that it hadn't made its way to me, when I assumed everybody else had seen it.
It turns out that countries hadn't, and this was a very, very sore point in Geneva.
Another related issue had to do with this framework for the continued evaluation of health status and health systems.
This has to do with the ICIDH-2. This is very crucial to this because it is through ICIDH-2, which has now become something along these lines -- and you all who are intimate to this can correct me.
You will see that actually one of the slides from WHO that I am going to show you has a different title for it.
In this case, we can call it the International Classification of Functioning Disability and Health Two. In any case, it is the only one, I suppose.
It is the core of how health is viewed in this, and it is very important conceptually to this effort.
The second component of this had to do with health surveys, that in order to apply this framework on a continuing basis, it is important that countries conduct surveys of their population as to their health status, and it is ICIDH-2 or this document that provides the framework for that, in essence provides -- not directly but indirectly -- the questions that would be asked, as to how healthy individuals are.
Then summary health measures, as you will see, are very important to this.
There were questions that had to do with all of these issues, and there was a meeting of the executive board, which happens prior to the meeting of the World Health Assembly, which I guess takes place in April or May.
The U.S. delegation included Tom Novotny, who heads the Office of International and Refugee Health, Mary Lou Valdez, who is a very critical person to all of this, who works in the office, a number of other people, and Greg Myer from AHRQ and myself from NCHS CDC.
We went, two of us as, in essence, technical advisors in parts of the delegation.
Let me give you a little bit, now, of background about the report. This is, in part, a real tour de force, this document.
It is very complex. The analytic framework is very complex. It contains an enormous amount of research that was done at WHO by many different researchers.
I think its goal, in terms of evaluating and tracking health system efficiency, but the use of these measures to in fact improve health systems, is a very important goal.
It is critical, though, that the process and the analytic framework be as valid, of course, as it can be, and that was a critical issue here.
The breadth of this document, though, is much broader than the analytic framework. It includes -- and this is a slide from the World Health Organization -- information about stewardship, about the financing of health care systems, resource generation, a variety of things, how to maintain and uphold the public interest.
So, the document is more than the analytic framework, but the focus has really been on this analytic framework.
Now, the objectives of this effort, again the slide actually is from Chris Murray, who visited us here on January 5, and spent the better part of the day with a subcommittee of the Data Council that had been put together and, at that point, had met a few times and met with him, to review this document and look at a variety of issues in it.
It is actually one of the best committees that I have been involved with.
The insights that the group generated were enormous and the zeal for evaluating this, analyzing it, was actually quite incredible.
One of the important purposes of this document, which is to stimulate interest in health systems, I think certainly has happened here. We have already had an interest in it.
I think it has created a base of interest and activity that I am sure will continue.
Well, this health system performance analysis objectives -- and the underlining and italics are mine on Chris' slide -- were to monitor and evaluate outcomes and efficiency of systems, build a system that can do that, build this in a way so that it is evidence based, looking at the relationship between health systems and performance.
That is a very tall order. It may be two lines, but it is a very tall order, to understand the relationship between a health care system and how it is structured and its outcomes.
Finally, to empower the public with information that is relevant to their well being from all of this effort.
I think it is three objectives that certainly nobody, I think, can argue with.
The framework has six pieces that are illustrated on this slide by the five checks and the word efficiency, on the right-hand side.
Let me try to explain this very briefly. The three critical elements of this are measuring health, measuring responsiveness and measuring financial contribution.
Now, when you measure health, there are sort of two dimensions, if you will.
One dimension is, if I can use that term, is the average level, say, in the United States, France, Canada, Zimbabwe.
The other is the distribution of health in any of these countries. It gets at the issue of equity. It gets at the issue of disparities and so forth.
So, this framework uses five elements, the average level and the distribution of health in a country, the average level and distribution of responsiveness which has a very particular definition here, not exactly what you might think, and finally, financial contribution, which is not looked at directly in terms of amount, but is looked at in terms of the distribution. Does everyone pay the same, or what are the differences and so forth.
All of these are put together into the sixth element, efficiency, which does include the amount of resources that are spent per capita in a country.
The ICIDH is very important to this. This was a background slide that Chris used about this.
It is very important because, as I said, it provides the core domains of health, as shown here, which go into the assessment of how healthy a population is.
So, these are the domains which, at first blush, certainly look very comprehensive -- vision, hearing and so forth, mobility, communication.
What it doesn't include -- and this is a source of debate and it is not just a conceptual debate -- it doesn't include information on our internal health, for example, whether or not we have diabetes but it doesn't affect our functioning at that point in time in any way.
It doesn't include risk factors, for example, smoking, use of alcohol, drug use, is not included in this.
This really looks at -- it is almost a kind of black box approach which looks at the individual and the dimensions of functioning.
That, in this approach, is what goes into the definition of health.
This is put together into a summary health measure. In this case, they use DAILs(?) as opposed to DALYs because this is a positive measure, in a sense, looking at life expectancy as opposed to looking at DALYs which, in essence, is a negative measure.
You all have probably heard of DAILs. WHO has a web site where there are a number of background papers that are included, that go into great detail on all of the elements that I am giving you here.
One of the criticisms, though, while there is a lot of detail on methods, they don't go into a sufficient amount of detail to allow a member state to actually re-calculate the estimates that were used in the report.
So, in something this complex, even more definition turns out to be -- even more detail turns out to be -- important.
Responsiveness was defined in this way. It has to do with the legitimate expectations of the population regarding the non-health-improving dimensions -- and I underline that -- of their interactions with the health care system.
So, this is somewhat different than one might think of in terms of responsiveness. So, it is related to the health care system but, on the other hand, it is non-health-improving dimensions.
The components are, for example, respect for persons including dignity, confidentiality, autonomy, client orientation, access to social support networks and so forth.
How one differentiates this from -- I will go back to the prior slide -- how these are non-health improving versus health improving may be a matter of definition on the part of the analyst here.
One could certainly point out that, for example, access to social support networks could be very crucial to the health of the individual.
What they were trying to do was to get another dimension of the health care system here, that is viewed as an important set of values related to the health care system, if you are trying to analyze or evaluate that health care system.
Now, the fairness and financial contribution had a number of components, but let me get right to the heart of this.
Health expenditures is defined to include all the contributions to the cost of the health system, including taxation, social insurance, private insurance and out-of-pocket payments.
They took this, then, on a per capita basis as health expenditures.
They then considered this with respect to non-subsistence income, defined on this slide -- which I will read for the people on the internet -- we define it as effective non-subsistence income, as the income the household would seek to consume, given financial market constraints minus expenditure required for subsistence.
In simple terms, it is everything other than the core amount of money that one needs to live in that particular country.
The contribution for an individual is the ratio between these health expenditures and the non-subsistence income.
They then -- if you recall from the slide I used earlier, which had the five elements and then efficiency -- financial contribution is not considered in terms of average level. It is considered in terms of the distribution.
So, the measure of fairness or equity or the distribution they used was this measure in which they looked at the difference between the individual financial contributions and the average, and used this formula.
I think essentially any measure of deviation could have been used here. In doing this, they tried to emphasize the tails of the distribution.
What is important here really is nothing more than the fact that they looked at the individual financial contribution to the health care contribution, with health care system taken broadly, so it included more than simply, for example, your insurance payments.
Now, we had these five elements, then. How were they put together?
They did some experiments to try to understand how to put them together. They surveyed people in WHO and they surveyed people outside of WHO and asked them how they would weight these five elements, the health level, the health distribution, responsiveness, that distribution and fairness of financial contribution.
What they got were results that were very similar, whether they were people in WHO or outside WHO.
They combined these into the all column here. Then they made just a slight adjustment to that. So, they weighted these so that health level got 25 percent, distribution 25 percent, responsiveness 12.5, distribution 12.5 and fairness of financial contribution 25. So, if you put it all together, obviously, it all adds up.
That became the measure of performance. Let me just pause and say again, it is this combination of five elements.
The analysis in here is actually done on the basis of two measures. One is just DAILs alone for countries and the other is this composite measure.
There is some difference in the overall ranking that takes place.
What is not clear in the document, but what is really crucial to understanding the rankings is how efficiency is defined and estimated.
Frankly, this was not clear to me, even though I had gone through this a couple of times, until Chris actually spent a day with us, and we talked about this framework and where it comes from.
The way performance or efficiency is defined here is in terms of a measure. In this case it says goal achievement.
Think of this in terms of that overall performance measure or of DAILs and ask what, on the basis of the resources that a country has to spend, what is the best that could be achieved, what is the maximum number of DAILs that could be achieved and, in terms of this performance index that weighted some of the five elements, what is the maximum that that could achieve.
This upper curve that says maximum attainable, is estimated actually in this document. It is estimated through an economic analysis that is based on two factors principally.
One is the dollar level of resources that go into the system, per capita dollars, that go into it, and secondly, the average education level of the population. Those are the principal two factors that go into this upper curve.
The lower curve falls back to an assessment of what health was like at the beginning of the 20th Century and, through their analysis, they defined what amounts to the minimum that is possible.
The way that I like to think of it is perhaps a little more reductu ad absurdum is, suppose we didn't have any resources going in at all, what would be the level of the population.
Well, that is probably too few, but in effect, it defines a lower line.
So, we have two countries here, A and B. What is the efficiency of the health care system of country A versus country B.
Now, country B, since it is more health care resources going to the right, let's say country B is France and let's say country A is Zimbabwe, which has fewer resources going in.
Both their health care systems may be operating at peak efficiency, but they are not putting in the same amount of resources.
So, they were focusing on efficiency and not, if you will, the amount of units of health that were produced.
The way they defined efficiency is by looking at the difference for A, the difference between the two curves, and then looking at how far up this difference country A actually was performing.
So, this distance here, over the big distance, was the measure of efficiency.
In this example, then, A and B are at about the same level of efficiency, even though B is doing better in terms of goal achievement and has more resources going into it.
So, this is very important to understand. It raises the question, okay, but what relationship does the health systems resources and the goal achievement have to do with the efficiency of the health care system in this year, or least year or over the last decade.
This is, in essence, a production function approach, particularly for the maximum attainable, but what relationship is that production function to what is happening in the health care system at this point in time.
Well, they calculated DAILs and plotted them, as you can see, here, versus -- this happens to be the log of the health expenditure per capita.
It is very interesting that, when you look at this diagram, what you see is, in effect, a frontier. That is the upper part of this mass of dots, which kind of says that, for any given expenditure on the lower axis, there is a maximum that countries seem to be able to do.
So, if you are over here, you can't do much better than this. If you are over here, you can't do much better than this. Nobody is up here.
So, you have a frontier. So, this seems to add credence, if you will, to their framework that is being used.
Again, if you abstract it, the diagram that we just saw, in effect, looks like this. This is actual data of DAILs versus the expenditure per capita.
So, you can see that this frontier analysis is very critical. The inputs into that were, as you can read on the slide, health expenditures in dollars, the average years of schooling.
International dollars means adjusted for the economies in the country.
Then there was a variety of other, but very small factors, that they put into this variable called positive other.
The key factors were the first, the dollars and the schooling.
The result, in terms of overall performance, is shown on this slide, which gives, again, on the horizontal axis, again on a log scale, health expenditures per capita, and the vertical scale is the measure.
A measure of uncertainty was put around each of these points. Then they calculated efficiency based on the argument and analysis that I mentioned earlier.
In doing this, they feel that this effort will stimulate the development of more and better data resources, and this has already happened.
Many countries, frankly, complained about the data that was used and said, you didn't use the best data. We have got better data. They started sending in better data.
Others said, well, we are not collecting that sort of thing, but we will do it. WHO is working on a survey that could be applied across the member states.
WHO makes the point -- again, this is their slide with my underlining -- that they are committed to helping the member states get the best available evidence.
They are also going to be working -- at least this is the plan -- with two or more countries in each of the regions at a subnational level to apply this.
For example, if we did it in the United States it would be, perhaps, on a state or a county basis, to improve the health care systems in these countries.
Well, what were the issues that the WHO examined? This is my slide now. What were the issues that the WHO executive board was hearing?
First, it was ranking across 191 countries. Many countries felt this was really not possible to do. It occurred to me -- I mentioned this to the Data Council when I talked to them -- that I was walking over one morning to the meeting and I was thinking, we have 200 countries and on a scale between zero and 100, that gives us, if this was spread out as much as it could be, it means that we would have half a point out of 100 between each country.
There is clearly a lot of uncertainty in the numbers that go into this. So, ranking is a pretty tricky proposition here. You have got a lot of things to rank and you don't have a lot of room, if you will.
So, it was an issue for many countries. The missing data or the imputations and estimations was a great, very significant issue.
There were questions about the use of DAILs as a proxy for health. As I mentioned before, DAILs don't contain certain things that are critical for any health minister in deciding what to do next in their country, whether it is about the health care system and improving its efficiency or changing the behavior or practice.
The responsiveness index was also an issue. The framework for efficiency, which I have mentioned, was a very significant question.
The measure -- it was a question about what the measure really is measuring. Is it efficiency in kind of a common sense way as we think about the operation of the health care system.
Since DAILs don't change very much per unit time, it seems to say that if we change the health care system radically, it is not going to show up in the efficiency measure.
Say, for example, we totally revamp the health care system within the next two years and we measure DAILs again.
Are they going to look any different two years from now from the way they look today. For most countries, they are not going to look very different.
Since they are not going to look very different, then how is the efficiency or how is a measure of the health care system going to show up.
I think it will show up in this framework over time. Whether it shows up in the notion of efficiency on an annual biannual or even every 10 year basis is something I think we need research on.
Another issue that was placed before the board was the need for peer review. I mention the release of the report was clearly a political issue.
Also, some countries questioned as to how you actually use this. How do you use the rankings, anyway, that come from this document.
Some countries were concerned that they were not involved in the development of the work.
Lastly, I put the rankings again, because the rankings rankled, I can't tell you how much.
Now, the U.S. position was that we felt -- this is the royal we, since Novotny headed the delegation he should really be here speaking -- but the U.S. felt it was important to build confidence, to link the report to health system improvement.
In other words, issues of validity and reliability needed to be proved and summary measures to improve health system performance were a good idea, it would seem, but exactly what did those summary measures mean in terms of health system performance.
We emphasized strongly the need for peer review, because the document had not received the kind of peer review that almost every report that I can think of in the United States receives.
We also felt that it was important that, over time, we evaluate the report's impact on changes in the health care system.
We thought it was also a very important and very positive benefit from this if this would work to harmonize efforts across the world, but particularly in the developed countries.
I must say that we are now talking, not solely because of this report, but in part of this, we are talking much more actively with Canada and with Europe on doing joint studies that will help us to make better comparisons of health status, for example, within our health care systems.
I think in that sense this has had a very positive results.
Now, the WHR, the World Health Report resolution that came out of the executive board, first of all, started with a preface that stated that comparisons should be predicated on broad agreement, on framework, design and data sources, and seeking input from all the members states.
This is the result of these deliberations on this particular issue at the executive board.
The resolution also incorporated a set of measures that were proposed at the beginning of the meeting by the Director General, who was responding to criticisms from a number of countries that she had received, particularly Brazil, but many other countries. Brazil was quite vocal in their concerns about this.
She endorsed, without any discussion ahead of time, at the beginning of the meeting, establishing a technical consultation process across all the WHO regions, consulting with states on the data that are used, providing information early, and establishing a small advisory group.
Rather than producing the report every year, produce the report every two years.
The executive board then added -- their component was, they requested the Director General to initiate a scientific peer review, to ensure the consultation process, to develop a multi-year plan for research and development which, I must say, I think is very important.
We really don't have the measures that we need in order to do this. In fact, we don't have a good set of national health accounts, which are critical in understanding the input that is going into a national health care system.
They emphasized a plan to improve data quality, that there should be a report on the impact of these reports on policy and practice, and they were very specific about providing information ahead of time.
After a several-hour discussion, they arrived at 15 days before time to publication.
They also dealt separately with ICIDH-2. What they did, which showed their concerns about this as a framework is, they separated from ICIDH -- I wish I had a copy to hold up -- they separated ICIDH-2 from the idea of doing surveys in every country and developing a standardized survey and summary health measures and focused solely on ICIDH-2, and gave it really a strong, ringing endorsement.
Their resolution here was not to the Director General but to the World Health Assembly to endorse ICIDH-2, and I guess that happens in April or May.
The next steps with respect to the World Health Report is some sort of a regional review of the framework. We would like to do that in the United States, and we have had a number of proposals that we are considering within the department, some of which involve the World Health Organization and some really involve just researchers in this country.
Particularly, we have talked about a U.S. meeting on this framework. And, as I mentioned, this has stimulated cooperative activity on survey development between us and Canada, and some countries in Europe.
So, that, in perhaps too long a nutshell, is where this stands. I must say this was a wonderful learning experience for me and for Greg Meyer, neither of which had been to the executive board.
I have to say two things. One, the professional delegation was really terrific from the United States. Novotny, Valdez, and many others who were either there or worked on it here, put in an enormous amount of time on these issues, and these were not the only issues that they were dealing with.
I think they did a wonderful service to the department in getting the department to really focus on these things. I see this focus as continuing.
The other thing was that I was very impressed with -- who am I to judge this -- but with the executive board of WHO.
It took some ideas which were not all that focused and, after a good deal of debate, actually put them into resolutions that I think were much better than the original ideas that had been put before them, which of course is exactly what you would expect.
Sometimes one wonders, especially in an organization that is dealing with six languages, simultaneous translations, documents that are written in the six languages almost as fast as activities take place.
I came away with a great deal of respect for what can be done on a topic that is as technical as this, and dealing with it on an international basis. Thanks.
DR. COHN: Questions? Barbara and then Dan.
DR. STARFIELD: Thank you. Having spent a lot of time trying to read this report and making sense of it, I can appreciate the value of what you have done. It is terrific.
There are just three other observations that I will make about it, and one is the ICIDH. We talked a fair amount in this committee, particularly the subcommittee on populations, about the summary measure, because we are thinking about how to fit a disability impairment framework in the core data elements.
Do we have to put in each one like we put in each individual ICD code, or is there some summary measure. That is one thing.
It is interesting to know there already is something that is being used somewhere.
Second of all is the whole notion of equity as it is used in this report. We tend to use equities and disparities sort of synonymously.
The way WHO has used equity is very different than we in this country have used disparities. They define equity as basically the difference between the lowest and the highest, the person who achieves the best health, the best something, and the person who achieves the bottom. That difference is the measure of inequity. The larger, the more inequity.
Now, we focused in this country on disparities, systematic disparities between population groups, subgroups.
There is nothing, absolutely nothing of that in the WHO report, nothing, because you don't know who the bottom person is.
You don't know whether it is a member of a group that is systematically deprived or not. So, the whole notion of equity as systematic disparities across population groups is missing completely from the report.
There are some basic assumptions that go into them. Measures are not only the measures themselves that are so difficult, but it is the assumptions that go into them that are pretty much unstated.
DR. SONDIK: The documents that Brazil prepared focused on at least a couple of those issues that you have outlined.
They felt that there was a strong sense of ideology, particularly ideology, in the report. I personally take it more in terms of this is one shot, if you will, one approach.
That is why we need the research and development. If something like this is a core, we can build on it and around it.
For example, I do think we need to look at more in terms of the distribution or the notion of equity here, and think of it more in terms of the way we think of disparities in this country.
Their purpose is to rank the countries and, in so ranking, stimulate the appropriate actions to try to move the countries that are lower to higher positions.
At one point in the meeting, the minister from Zimbabwe said to the minister from France -- France ranked number one and Zimbabwe ranked number 191 in this -- he said, I feel sorry for you because you have got no place to go.
That, in essence, is the purpose. There is always a danger in this that when you produce something like this, every element is taken as, this is the way to do this.
I must say that the report does not provide a variety of alternatives to it. It basically says, this is the way to do it.
Certainly in talking to the researchers involved in putting it together, they are well aware of a variety of issues and would welcome people building on the framework.
DR. FRIEDMAN: Ed, I have reviewed the report a couple of times and, like Barbara, I really appreciate your attempts to bring clarity to it.
When I was reviewing it, on the one hand, I was struck because I thought the framework overall was very helpful and I thought the goal of that report was laudable.
At the same time, I found the methodology absolutely opaque, absolutely opaque. You indicated part of this.
When I went back, and I think it was the responsiveness measure, I was really surprised by the extent to which that seemed to be based essentially on a delphi method of talking to a few folks in each country and obtaining what was essentially a subjective opinion from some undetermined number of people in each country about some of those dimensions.
DR. SONDIK: That is a good point. There are two elements here. One is the definition of the measure. That is sort of issue A.
Then issue B is, okay, if you accept that definition, at least in terms of the production of the first report, how do you get the data for that.
The data was from a set of, in essence, key informants. There are questions about how you choose, obviously, the key informants.
You know, the argument for this is to say, we wanted to produce -- this is why I consider this sort of a tour de force when you put it all together -- the argument is, we wanted to be able to work on all these elements simultaneously and put them together. So, this is what we did.
We had very good agreement among these informants, key informants, that were used.
On the other hand, they are in a very weak position, of course, when one asks the question, well, you know, how valid is this, or how reliable, for that matter, even among the key informants.
What they may say this year may differ wildly from next year because it is highly subjective.
So, when you do this sort of thing you clearly open yourself up to lots of criticism.
To be positive about it or charitable about it, putting it all together was an accomplishment. Not providing a sort of desiderata of issues that need to be dealt with, a menu of issues that need to be dealt with and improvements in this, I think, was a mistake. That is not in the report.
It probably should have been there up front and in conclusion. But it really came across that this is the methodology that we are going to be using. We will refine it, but this is it.
I am glad you mentioned that. The lack of solid data in this is striking, but that is only one of the elements.
The fact that data was only available on 30 countries and was estimated, imputated, first of all, basically by taking a few countries in each region and then extrapolating to the region, and then particularizing it to specific numbers that could be made country specific, like the amount of money that was being used to fund the health care system or mortality measures or that sort of thing.
The distribution of health was another issue, which was really not the distribution of health. They don't have a way of doing distribution of health because they don't have solid surveys from these countries at sufficiently high pixel levels, if you will, that you could actually get a distribution.
What they used was mortality data for children, for five year olds and under, that were used to give the distribution.
DR. COHN: I guess I am hearing two of our members thinking that you need to initiate a scientific peer review sort of thing, is the conclusion I am hearing.
I am just going to open it up for other questions. I do want to warn the committee that we do have other issues after this session to discuss.
Probably it is not a good use of our time to pick at the report as much as to make comments on whether we think the recommendations that have come out from the executive committee are the right recommendations.
It is just a thought about how to use our time. I am sure there are tremendous amounts that could be picked at in this report. Jim, you had a question or a comment?
MR. SCANLON: I have two quick observations in that spirit. Ed, when you are ranking countries and you only have data for 90 of them, I think we euphemistically call them actuarial techniques.
The other thing is, I think there were good parts and bad parts of this whole story. The framework itself, in my own view, was just appalling not well done in terms of method, data or consultation or almost anything. It was more of a political advocacy kind of attempt and hopefully that sort of thing can be avoided.
I think in terms of a good positive outcome, I think in the United States, as Ed said, as well as in other countries, I think people are beginning to think, well, what is it about the health system that you can measure in terms of performance and how does it really relate to health outcomes.
Maybe the DAIL is not a good measure. It is probably a good measure across countries. It is probably not a great measure in the United States and maybe there are other aspects to make it a little more sophisticated analysis.
Hopefully it will point to a research agenda, Ed, that will make more sense.
The other thing, for the perverse and unintended consequences, the production function actually argues that many countries should be spending less, because you can't improve the outcome that much anyway beyond a certain point here, in the middle even less.
It is actually almost the opposite of where you would want to have a World Health Organization sending dollars.
DR. COHN: I think that is a comment. I don't think that is really a question. Kathleen?
MS. FYFFE: Thank you, Ed. Very quickly, the scatter diagram, do you know if the United States and Canada were in the frontier on the scatter diagram? Were they sort of as outliers, as little islands?
DR. SONDIK: I would think that in this one, which is looking at the DAILs -- we would have to check -- that that is probably us on the right, because we spend more per capita than any other country.
Now, another one of these sort of, not perverse but peculiar thoughts you get when you are walking from the hotel to the WHO was that, I was thinking about, well, we do spend more than any other country.
On the other hand, what we spend also influences the health in other countries. What we are spending is money on research and a variety of other things, drugs and so forth.
How do you actually work to separate all of this and should we be, in some sense, penalized because we are spending too much. I don't know.
I don't think there is any answer to this. One has to view this, in my view, as a research effort, and you use that to give insights.
In terms of giving measures such as the GHP, the gross health product, if you will, I don't think we are there yet. I just don't think we know how to do that.
I am not quite sure whether actually that would really be a useful thing to do.
I actually looked at the table but, in all honesty, I never took the diagram and tried to figure it out. I am almost positive that is us out there.
DR. STARFIELD: Actually, there are other countries that spend more. One of them is Luxembourg.
DR. SONDIK: Really, per capita? How interesting.
DR. COHN: One final question, Paul, and then we will wrap up on this.
DR. NEWACHECK: Thanks. This has been a really interesting presentation. I want to continue with this point, though.
As Jim pointed out, there are obviously methodological problems here. It does look like there are diminishing returns to further investment, at least on the DAILs end of the coin.
If you look at your other chart, which was the total overall performance, you don't see those same diminishing returns.
I am wondering, does that imply, then, that you get to a certain point -- or would this be an interpretation -- you get to a certain point in investment where you can improve health status. Then, beyond that -- I am sorry, you get to a certain point where you reach maximum gains in improving health status.
Beyond that, further investment leads to improvements in the other two domains, the responsiveness and the financial contributions and that is why you see this continued upward improvement in overall performance?
DR. SONDIK: The way this report was done, you can make all sorts of conjectures. It is a black box approach, in effect, as I mentioned before, sort of on the survey side.
There is no internal model here. The only modeling that has gone on has to do with this frontier, which says that the best you can do is a function of expenditures and education.
So, I think -- that is why I think sort of the value from this is it should stimulate this kind of discussion.
Then one would want to get in and look more deeply, for example, at those countries that are out here on the right-hand side and say where are the differences in this measure. Let's take the measures apart.
Unfortunately, we have got these fundamental data issues here, that Dan brought up. This should stimulate this type of research.
DR. COHN: Ed, thank you. I agree with Paul, that this is a fascinating presentation. I suspect that the subcommittee on populations may have some discussions about next steps related to all of this stuff.
I think this is an area that we will want to have as a full committee.
DR. SONDIK: We would be very interested in getting -- obviously, the department, the Data Council, would be very interested in getting any of your thoughts on the directions that we should take.
I hope, in fact, that this does represent a vector pursuing this issue of more analysis of the health care system, and allowing us to make these kinds of meaningful comparisons between countries in the future.
DR. COHN: It definitely gets the discussion on the table.
DR. SONDIK: Absolutely.
DR. COHN: Ed, thank you so much. Now, for the full committee we have a couple of items we need to take a look at.
One of the advantages of flying in from California is you have a chance to read all the material.
Now, what I think we want to do is just sort of go through a couple of the action items. With one exception, they are not action items for today. They will be action items for today, generally. If we can get our chance, we will talk a little bit about the NCVHS 50-year history, if we can, before lunch also.
Now, the first document we wanted to review is in tab 4, which is the draft fourth annual report to Congress on the implementation of the administrative provisions of HIPAA.
Agenda Item: Review Draft Fourth Annual Report to Congress on HIPAA Implementation.
DR. COHN: I think as you all know this is a mandated report that we prepare each year for Congress and the Secretary of Health and Human Services on sort of what has been happening, our views of what should be happening in relationship to the implementation of HIPAA.
This is actually in its second draft right now. This was drafted, and I want to thank Jim Scanlon for his work on that.
It was reviewed by the Subcommittee on Standards and Security and then revised on the basis of that.
What we would do is to ask for input from the whole committee, assuming you have read it, with the idea that we hope we can bring it back for a vote tomorrow at the meeting, with whatever word smithing that needs to occur.
We would ask that tomorrow it be approved with the ability of the executive committee and John Lumpkin to make whatever word smithing needs to occur around all this.
I guess the question I would have for all is comments or otherwise about this, especially from people who are not members of the subcommittee on standards and security.
DR. STARFIELD: I have just a couple of questions more than comments. I guess I really am not an expert on HCPCS or whatever they are, the procedure codes and the level one, two and three.
I found the discussion on local codes confusing. I mean, HCPCS is not a local code. It is a national code?
DR. COHN: The Health Care Procedural Coding System.
DR. STARFIELD: That is it. The common procedure coding system, level three, are not local codes, and yet they are discussed in the context of a discussion about local codes. I just got confused about it, page 10.
That is just one thing. I found that confusing.
DR. COHN: I think that sentence actually misrepresented I think the intent anyway. There is actually a paragraph there.
DR. STARFIELD: Let me just mention the other thing, because Kepa has got something to say about that, and that is on page 21, and that is the unique health identifier.
I was confused about actually what Congress said. I guess it seems to me to be circular reasoning, that HHS can't use any of its appropriated funds to promulgate or adopt any final standard until legislation is enacted specifically approving the standard.
I guess I don't understand what, for example, this committee is prohibited from doing. We certainly can discuss it, I assume. We just can't propose a standard; is that it?
DR. COHN: Yes, we could, I think, actually do whatever we want, but obviously they can't be taken anywhere.
DR. STARFIELD: But we have basically interpreted congressional action from stopping attention to it in this committee. At some point we should discuss it.
DR. COHN: Jim, do you have a comment on it?
MR. SCANLON: I think it is directed at HHS. We are not to use any funds to adopt -- this is the exact language from the law. It has been in there for the past two years.
I think the intent was to stop HHS from adopting or promulgating any standard in this area and leaving it up to Congress.
Whether anyone would really want to pursue this is another issue, other than HHS. I don't know anyone who is pursuing it.
DR. STARFIELD: HHS can discuss it. We can discuss it; right? We are not prohibited from doing that? We are not prohibited from using funds --
MR. SCANLON: We could not promulgate or adopt.
DR. COHN: This is actually the language that has been in previous annual reports, although you do bring up the point that we can discuss, whatever.
I think the reason the committee was working hard on this area had to do more with the fact that the Secretary was being required to promulgate standards in this area. It certainly is open for discussion. We could bring it back on the agenda. Other comments? Barbara, did you have anything else?
DR. STARFIELD: No, but Kepa had something.
DR. ZUBELDIA: Barbara, let me address this issue on the local codes on page 10. HCPCS level three are local codes.
There are also national codes. Those are in level two of HCPCS. Level three are all local codes. There are other local codes that are not HCPCS. Some payers have other kinds of local codes.
DR. COHN: I think what Barbara was saying is that this paragraph is confusing and if she doesn't understand it without having to be an expert in the differentiation, then I think it needs to be -- the idea is the right one. It just needs to be stated in a way where it is obvious to her.
MS. FYFFE: Is there any rule against using footnotes to explain things? What is a local code? That would help the reader.
DR. FITZMAURICE: Simon, I would point out that on page 29 there is a more elaborate explanation of local codes and level one, two and three codes, that might help in understanding.
Page 10 refers just to the executive summary, but in the body of the report there is a larger paragraph.
MR. SCANLON: For almost everything in the executive summary here there is often more detail in the report itself. For the most part, it is meant to be factual.
This paragraph was taken from the NCVHS letter to the department. Maybe in the executive summary it would be better just to say that there were a number of issues related to the local codes, and these are discussed -- this was even boiled down from the letter itself, which went on and on.
DR. COHN: Okay, other comments? Let me just make a comment or two about the tone and the substance, just so that you are all well aware, because there were some decisions made by the subcommittee in terms of how to characterize things and just to make sure that everybody is comfortable with this.
We had changed from the previous report to this report how we are characterizing privacy. So, there is now a final rule that is including paper as well as electronic media.
On the basis of that, whereas in previous reports we have always said the privacy rule is not enough, you need to have a comprehensive law, we elected, based on the fact that this regulation seems to be more comprehensive, to omit the references to once again calling for a law.
That is something everybody just needs to think about. I would just draw it to your attention because that is a real substantive change from previous positions we have taken and people just need to be comfortable with that.
MR. SCANLON: There is some commentary on page 4 in the executive summary after the report that the privacy rule actually was issued.
In the last paragraph, at the top of page 4, is sort of the committee's --
DR. COHN: Why don't you read it.
MR. SCANLON: Yes, let me read it out for everyone. After describing that the privacy rule was issued, the paragraph reads, the NCVHS supports and applauds this progress on national health information privacy standards.
The committee views health information privacy protection as a foundation for the full complement of health care administrative simplification standards, and for progress on additional health information technology applications.
MS. GREENBERG: Although as was heard this morning, it is for the committee to decide, although paper is now covered, there are obviously a number of groups that are not covered by the regulation, such as employers being a significant one.
I can't remember, I have read it a few times and I actually think the report looks very good now, but whether there is something here about the fact that the committee will look at areas that aren't perhaps covered, it is going to be examining the regulations and look at whether there are areas or gaps that need to be covered.
If the committee is not going to be mentioning anything about the need for comprehensive privacy legislation, then that may not be the position that it currently wants to take.
It will still continue to look at where there are gaps and make recommendations on ways to address them to recognize that there are gaps, such as not everyone being included.
MR. SCANLON: Actually, in many of these areas, when the committee reports on what occurred during the year, it is sort of a backward look and then a forward look.
We have often said that the committee plans to perhaps assess implementation or any obstacles and so on and make recommendations.
Maybe we could use a sentence like that relating to privacy that -- again, this has to be what the committee wants to do.
We don't want to promise we are going to do something if we are not really going to.
You don't even have to mention comprehensive privacy legislation. You could just say that we would look at the potential impact and so on, and make recommendations regarding areas of improvement or expansion and so on, something like that.
DR. COHN: I think this is an area we probably also want the input from the Subcommittee on Privacy and Confidentiality.
Unfortunately, our acting chair is not here right now, but I presume he will be here in the morning. Hopefully at the subcommittee meeting they can at least look at this and provide any input that seems appropriate.
MS. GREENBERG: Will you be at that meeting?
DR. COHN: I am a member of that subcommittee, as are a number of other people at the table. We can just reflect on that.
Anyway, we will be taking any comments and happy to hear from any of you as you review it. Once again, the intent here would be that we pass it, or that we bring it forward for action tomorrow.
I would have you all sort of look at it. Those of you who haven't reviewed it hopefully will be able to review it tonight and hopefully we will be able to act on it tomorrow.
The next item is a letter that you don't have, which is on NDC issues. Hopefully the subcommittee itself will see it this afternoon and we will discuss it and bring it back for action.
The next item after that is something called the NCVHS update to the department.
MS. GREENBERG: Tab 5.
Agenda Item: NCVHS Update to the Department.
DR. COHN: Tab 5, thank you. It is a short document that, at least as currently written, is mostly meant to highlight new reports in the past year coming from the NCVHS.
John Lumpkin and I have had some discussions. I think that actually, with a little bit of edition, it actually provides a nice sort of executive summary for the Secretary of sort of the full breadth of committee activities in a digestible form.
It is not that our annual reports or other reports aren't immediately digestible, but there is something to be said for something that is only a couple of pages long.
Having said that, it occurs to me that there are a couple of areas that need to be augmented a little bit. I am not talking about whole pages, but sentences or paragraphs.
One area is privacy and confidentiality, that is mentioned but probably needs a little bit of augmentation here.
There probably needs to be a little more augmented around standards and security, maybe primarily just referencing the annual report -- see the annual report for more information.
Then, very importantly, the issue of the subcommittee on populations, which is sort of mentioned but needs to look at this area and see if there is something that needs to be referenced in relationship to that work.
Once again, I am not looking to turn this into a 10-page document. It sort of defeats the purpose. If there are some ways to put in pertinent sentences or paragraphs, I know that is something that John is looking toward and so am I.
Maybe in the subcommittee deliberations we can sort of review that. Does that make sense? Do other people have comments about this document as you see it? The good news is that it is short. I think we should consider that to be a feature. Comments about this one?
DR. ZUBELDIA: Simon, should we attach to this document the report to Congress?
DR. COHN: I think that is something we should talk about in the subcommittee. I was thinking more that we would want to reference it anyway.
I am not sure that attaching a 30 or 40-page document --
DR. ZUBELDIA: There is a new Secretary and this is a report to the Secretary.
MR. SCANLON: We do submit the annual report to the Secretary as well. It may be more a matter of timing, depending on which gets in first.
While we transmit the annual report to Congress to the appropriate committees, we can typically send it at the same time to the Secretary.
It may, Simon, just depend on the timing of each one of them or which one of them might be first.
DR. COHN: Other comments about that?
We are now actually at 12:15, which is when we are supposed to adjourn for lunch. Why don't we adjourn and we will come back at 1:00 o'clock. At that point, we will be back on schedule and start the afternoon.
[Whereupon, at 12:15 p.m., the meeting was recessed, to reconvene at 1:00 p.m., that same day.]
DR. COHN: We are now going to hear from Ann Hardy about work she has been doing with the NHIS and sentinel health indicators.
DR. HARDY: I am here today to talk about some work that a group of us have been doing at NCHS, trying to evaluate the use of the National Health Interview Survey, or the NHIS for what we are calling the surveillance of sentinel health indicators.
I know you have a busy two-day schedule, so I appreciate being given some time to talk about this project, which we are very excited about.
It is still at a fairly early stage, so I am really anxious for your reactions and your feedback as well.
Okay, what do we mean by surveillance system? Basically, we are evaluating the feasibility of releasing less than annual estimates for a small set of key health indicators from the National Health Interview Survey very shortly after data collection.
This is a project that we received funding through CDC's evaluation project money. So, it is an evaluation project. It is not clear it is going to be a go, but it is looking very promising at this point.
What I want to do this afternoon -- and I am hoping to only take about half the time I have been allotted, again, to allow time for comments and questions -- I am going to give a little bit of background about the health interview survey focusing primarily -- I am sure almost everybody or everybody here knows about it, but focusing on some of the aspects that I think are relevant to this project.
I am going to talk a little about the projects and background, how we got into it, how it was conceived, and the progress we have made to date. Again, I am very interested in feedback.
Okay, the health interview survey has been around for a number of years. It has been in the field every year, I think, since the mid or late 1950s.
It is an annual survey. The target population in the U.S. household population.
It has two main purposes. One is to monitor trends in basic measures of health status and health care utilization.
It also can address issues of current public health concern as they come up. The HIS does this by having a set of questions that we call the core questionnaire that stays the same from year to year. We have updated it periodically, but try to leave it alone, again, to monitor trends.
Then we can add onto that special supplemental surveys on a variety of topics, again, to allow us to get at new things as they come up.
In terms of the sample, again, just summarizing some things that I think are more key to this particular project, we get information from a nationally representative sample of about 40,000 households a year across the country.
We are in the field throughout the year. Actually, the sample design is such that each week is actually a small representative subsample.
Now, we do this primarily so that if we have to cut the sample at any point during the year for budgetary reasons -- and this has happened -- we don't lose our ability to make estimates for the target population, for the U.S. population.
It also does allow us to then look at less than annual data. We don't have to wait a whole year in order to get a representative sample.
There is some precedent for this. We have done this on a few occasions, I know, back in the late 1980s when we put together an AIDS knowledge and attitude supplement to go along with the Surgeon General's activities.
We started looking at that data on a monthly basis and we have also done it for childhood immunizations.
So, we have done it. It is difficult. It often meant that we had to pull those parts kind of off line, apart from our regular processing and production systems, but it can be done.
We also, in this current design, over-sample both the black and Hispanic populations.
Now, as I did mention, we have a core questionnaire that generally stays the same, but we do make changes to it.
In 1997, we introduced a fairly dramatic redesign of the questionnaire content and format. I am not going to go into a lot of detail and there is more information on our web site.
We did feel that it was really time to rethink our approach to some things, to put some things into the core that hadn't been in the core and so on.
We also, at that time, changed our mode of data collection from what had been a traditional paper and pencil interview to one where the interviewers enter the data into a computer-assisted personal interview mode or CAPI mode.
I think we were trying to increase the relevance of the data again. There were topics, like health insurance, that wasn't in our core questionnaire before 1997.
We also wanted to improve data quality. We did a lot of investigation and changed, in some cases, the way we were asking questions on certain topics, to get more accurate data from people.
Also, the timeliness, the CAPI does eliminate the need to code and key your data after collection. So, we thought we would save some time.
The topics that are currently included -- and many of these have been on the HIS core questionnaire in the past -- we are now calling the core questionnaire the basic module, but these are the topics that we get at least some information on every year.
Again, the HIS is meant to be a broad health survey, so we try to touch on a number of aspects of health, some measures to get at health status, such as limitation of activity, chronic conditions and some injuries, get some information now on health behaviors, another topic we didn't get in the core questionnaire in the past.
We get information from adults on smoking, alcohol use, physical activity, height and weight, and then health care access and utilization, health insurance.
We continue to get a large amount of detailed sociodemographic information.
I don't want everybody's eyes to glaze over at our processing. This is not one the most exciting topics, but I just want to make a point that the data, CAPI does give us the data fairly quickly. It is in better shape.
We still have to, in order to release -- I think we are releasing about five or six analytic data files complete with documentation of both the questions and the data themselves -- that there are a number of steps that we still have to go through.
I just want to focus on the initial phase, where we really try to determine if we have gotten all the sample households that we were supposed to, try to determine who the responders and non-responders are, get everyone in the right families and households, and the data associated with the right respondents.
After that, we do initial editing to do the weights, and then our final editing, recoding, documentation and so forth.
At this point, our lag time between the end of data collection for a data year and the release of our annual files is about 18 months.
We are hoping this year that, at the end of the year -- this summer we will be releasing 1999 data. We are on track, however, to start to catch up a bit and actually improve on that and, at the end of the year, hope to release 2000 data.
So, at least we are back to releasing data about 12 months after data collection. Again, I think it would be nice if we could release at least some estimates in a little more timely fashion.
We get some information from everyone in the family. Then we pick a sample adult and a sample child and get more detailed information from them.
We have to sort of get everybody all associated back to the right place. That is all.
Now, I want to turn to talking about the surveillance project. Again, what we are trying to do is, can we pull off some estimates more quickly in the process.
We got thinking about this for a variety of reasons. Obviously, each week is a representative sample. We have done this in the past, knew we could do it.
One thing from our CAPI system, we do get data very regularly, about a month after data collection.
In the past, with paper, we had to wait for coding and keying. It wasn't always as predictable, how quickly we would have data to look at.
Our data are cleaner, if the CAPI instrument is functioning correctly, and it took us a little while to get to that point.
If it is, it handles a number of things that one could have had problems with in the past with paper. So, the data are cleaner and some of the items really require little or no editing after they come out of the CAPI system.
In terms of steps for the project, the first thing we have been grappling with has been what indicators.
I haven't gone into a lot of detail on the items, but there are literally hundreds of possibilities from the health interview survey on which items.
Then we have been trying to evaluate. We have come up with a set of candidate indicators, and I will talk in a minute about how we arrived at those.
We tried to look at both the operational and the statistical aspects of producing less than annual estimates. Can we go down to a month, should we stick to a quarter, semi-annual, how do things vary, month to month or quarter to quarter, things like that.
We have also been thinking about how we are going to disseminate the data. Again, we have a plan for at least an initial release, and I will talk about that.
Then we are going to, once we start to release data, really have to take a look at how useful the data is, how useful our dissemination mechanism is, and the indicators that we have chosen.
In terms of selecting the indicators, we wanted to get input from folks outside NCHS. We held a meeting in November, and actually Dr. Newacheck was there. I should actually thank him because he is one of the reasons that I am not here, I think.
We included people from a number of other federal agencies, throughout the department, as well as outside experts on health policy, representatives from the Council of State and Territorial Epidemiologists was also there.
Really, we just sat people in a room, looked at all the possible items we could and really kind of forced the group to pick some.
We also looked at other -- there are obviously other lists. We are not the first people to think about this. There are a number of other sources -- and I will mention those in a moment -- of leading health indicators.
We have spent some time now thinking about the kinds of statistical considerations of the ones we have selected.
In terms of other lists, this is just an example of some of them. The 2010 has the 10 leading health indicators.
President Clinton's initiative to eliminate health disparities has certain focus areas.
CST has done a lot of work coming up with indicators for chronic disease surveillance.
Actually, NCHS has also done some of this, primarily related to health care utilization.
At this point, with the input that we have gotten, these are our proposed set of indicators that we are going to at least release initial information on.
It includes influenza vaccination of folks over 65, smoking, regular source of medical care, unmet need for medical care, heavy alcohol use, HIV testing, and self-assessed health status.
I will show some examples of what we are planning to release in a moment.
Things that we are still working on that we would like to release but we are not there yet for a variety of reasons, I want to talk specifically about health insurance.
The meeting in November was very useful to us, a lot of discussion, not always a lot of complete agreement except on probably this one particular indicator, and that is really seeing some information on health insurance coverage, which everybody -- health policy folks and others -- felt would be very important.
Our problem with that -- so I am going to focus just a minute on that, and we agree this would be a key indicator to get out from the HIS and get out quickly -- our problem is, right now, as the data is reported, there is a fair amount of misclassification and we have to do some hand editing to get rid of that.
The biggest problem, the biggest source of error are people who have a single service plan but report that as comprehensive coverage.
We have made some changes in the questionnaire to try to avoid that. It doesn't seem to be working as well as we would like. It has eliminated a bit of the problem but not enough of it.
What happens if we just take the data as it comes out of CAPI? We underestimate the uninsured population by about three percent.
The feeling was that was just too much error in that particular estimate that people are really keyed on.
What we are trying to do now is figure out if we can, in a fairly automated and quick way, correct this, at least get people correctly into one of two groups, insured and uninsured, so that at least we can put out an estimate of the uninsured population.
We are just starting work on this. I don't know if we will be successful. We are hopeful that we will be. So, at least we can do that, and then ultimately include health insurance in this system, at least a measure of the proportion of people without insurance.
So, we have our set of candidates. Again, these are candidate, even though we are going to release some data on these. They are still candidate.
Again, I would be very interested in your feedback now, later, on this.
What we have done to kind of evaluate some of the statistical and operational aspects, we went back to 1997 and 1998 data, which we already have final annual data released.
We compared estimates at different phases of data processing, compared the reliability of estimates for different time periods, looked at preliminary estimates.
We had to come up with a preliminary weighting scheme with, again, final annual estimates.
I am not going to show numbers specifically, but just in terms of what we have come up with, we feel that we can produce reliable semi-annual or quarterly estimates for most of the indicators, if not all the ones that we are working on right now.
I think we are going to probably not go to any smaller time period than quarterly at this point.
The data, we can actually, at least for the indicators we are looking at, get it after the first processing phase. So, these are estimates that don't need a lot of editing. They come out pretty cleanly out of the CAPI system.
We have been able to develop a preliminary weighting scheme that looks very close to what we ultimately end up with. Again, the estimates we end up with should be very close to what one would get if they had the final data.
In terms of dissemination, at this point, we are going to put the information up on our web site. We are looking at a series -- and I will show this in a little more detail -- of PDF files that will have primarily graphs.
We are going to try to keep it pretty graphical, some tables with enough explanatory text that people understand where the data is coming from, some other caveats but not to kind overwhelm it.
We are going to start in a few months and release some semi-annual information from the first half of 2000.
Then we are going to be taking a look at -- we hope to get some feedback from this and take a look at how useful people think this information is, the way we are doing it, what we are doing, and then move to try to release some of the 2001 data, hopefully later on in the summer.
Things we are still grappling with. How to show trends. I think we have decided, at least for the initial information we are going to put up for the first half of 2000, that we are going to show some data going back just to 1997.
Because of the major changes that were made in 1997, for some estimates we really made a break in the trend line and it is really hard to compare to previous HIS estimates.
For some things, the questions didn't change much. The estimate are pretty stable. Rather than trying to bounce around, I think that is what we are going to do now.
Again, I think once we start producing estimates regularly on a quarterly basis, I think we are a little concerned that we could get more estimates out in the course of a year, but not focus on the same ones. Again, we are not sure whether that is what we are going to do or not.
I was going to show some examples of what I mean. On our web page, again, we would have, for each indicator, a PDF file that the person could click on. I will show two examples of what we are thinking about with smoking and influenza vaccination.
We will also have a file that will describe the project, give a little more information about the source of the data and some of the other statistical considerations.
For smoking, for example, right now we are envisioning that if you click on this, you will get to a series of charts.
Again, what I think we are going to do, the initial chart will compare the most recent time period, and we are talking about the first half of 2000 with the final annual estimates for the previous years. We are going to show the confidence intervals for people can kind of eyeball it.
A little bit of information about the source of the data, there will be more detail in the description of the project, and then just some bulleted kind of highlights of what the sort of take home messages from each of these are.
For example, the prevalence of smoking, this is a significant although slight decline over time in the proportion of adults that smoke.
We will also be able to show some major demographic breaks, although not a tremendous amount. For example, age and gender. Again, this shows current smoking for three age groups, males and females.
For influenza, one is a vaccination. Again, the initial graph will be similar, with the most recent data showing some of the previous years.
There is some seasonality in this reporting, even though we are asking for people, whether they have gotten vaccination over the last year.
So, we decided to show the same time period for the earlier years rather than the full year.
In addition to age and gender breakouts for some of the estimates, we are also going to show the four major race and ethnic groups. Again, we can show some differences.
That is all I had planned to show for today. Again, I am very interested in what you think about this, what you think about the indicators or other suggestions about ways to get input from others on this, and what we are planning to do. I am happy to answer any questions.
DR. COHN: Okay, Dan, Kathleen and then Vickie.
DR. FRIEDMAN: I think it is really terrific that NCHS is going to be doing this and releasing the data more rapidly and on a sub-year basis.
At the same time, I am somewhat concerned about the impact that this might have on the use of the leading health indicators.
I think that it may give the impression that CDC is not really supportive of the leading health indicators.
I realize that there are some good reasons for not using them.
DR. HARDY: I think some of the ones we have chosen are, and some of the ones we don't have.
DR. FRIEDMAN: I think that some of the ones you have chosen are, and I think it would be really helpful to basically put the leading health indicators up against the ones you have chosen and, for all the leading health indicators, just indicate why or why not those are not included in the sentinel indicators.
It is not that I am that crazy about the leading health indicators in and of themselves, but the concept, I think, is an extremely important one.
DR. HARDY: Right, and I think it is important that we don't do this kind of in the dark or in the blind, not being sensitive to other things.
I think that is easy enough to do. We did look at that and there are some -- it did help us in selecting some of these, and there are some that, frankly, we don't have data on in the HIS, but I think that would be useful to tell people up front.
DR. STARFIELD: This is corollary question to Dan's. The whole notion of sentinel health indicators, what is it meant to convey?
Is it meant to convey something that is responsive to health system change? Is it meant to convey something that is responsive to social change? What underlies the theory of sentinel?
The second thing, which is also related to that is, how much was 2010 used to inform the choice?
For example, I could see saying, you know, if we are really interested in disparities, we would take the 10 leading things that have the greatest disparities and pick them. How much did that enter into these decisions?
DR. HARDY: I think by sentinel, we were trying to get at the notion that there are some important indicators that people would like to have from the HIS on a more rapid turnaround than we have been able to produce.
Again, we did look at the 10 leading health indicators from other sources of information on leading health indicators, and we are open to suggestions here, although we are constrained somewhat by what we have in the health interview survey.
I think our hope was that people would like to get some of our data more quickly and that that would be useful.
We had a lot of discussion about this at the meeting, about how much flexibility we would have to look at something new, if something changed or something came up, and I think we want to have that flexibility.
Were we going to try to select our indicators based on what was going on currently.
I think someone suggested that if the unemployment rate suddenly starts to go back up, are there some things that we would want to jump on and look at and see if it makes any difference, if we can note any difference in insurance coverage or something like that.
I think we are grappling with trying to balance what we can do from our data, how far we can push it at this early stage of taking the data basically right after it has been collected, and what would be useful to people on the outside, and we are looking for some suggestions.
Again, when we release some things, I think we are going to be making -- we are not wed to those for that time after and we will be looking for some input.
DR. STARFIELD: Just a couple of quick observations. There is not a lot on children. It is very adult focused.
I guess I didn't realize that it was recommended that everybody get an AIDS test. That is the implication in this, the higher this figure is, the better it is. I didn't understand before, that that was a goal, to get the whole population AIDS tested.
DR. HARDY: I don't know that it is, but it is an indicator. One of the problems has been knowing how many people -- there are lots of different ways to get tested, people going to public sites, and CDC, that funds a lot of those, has some fairly good data on people that go there.
People go to their doctor, people get tested in a lot of different ways now, and the HIS, I think, is the only source that kind of pulls it all together to kind of come up with an estimate of people who have been tested in a variety of means and look at the different places people have been tested.
It has been an estimate in the past that people have wanted to know, have followed for that reason, not that we are trying to get to 100 percent, I don't think.
DR. MAYS: I think that my colleagues and I are kind of in kind of the same line of questioning. I think that anything that gets data out faster, I think I am very supportive of.
I guess I have three things based on what has been said. I am still confused as to the overall goal that this is going to serve.
There is a lot going on in terms of the leading health indicators. Then this seems -- I don't know if it is attached to it, I don't know if it is like supporting it. I guess I don't understand it.
I think if data people don't understand, then the public gets even a little more confused about that. That is number one.
Number two, it is like the data, at least if I take just some of the examples, it is somewhat limited, in a sense.
Let's just say that we do big age groups or we do only some of the ethnic groups. Then we have this other initiative out in which we are really finding out that we ought to try and do the data in ways in which it gives us more information.
Then you say, well, do we want to use those results, or when should we use those results, if the data that is coming out now tells us that we should have a broader age range, and be more specific, or it tells us that we shouldn't put the Asians in the other category or something like that.
Then the third thing is the issue of, is it possible to really focus on things where we don't often get, like rural health.
I mean, is it possible to kind of do these releases in ways in which it really augments other things that sometimes we don't actually get a lot of data from, that might possibly help people to kind of get people to get things going in that area, such as rural health or focus on gender.
It is like in the advance reports, sometimes when you take very specific things, and then it helps other people to know how to use the data.
DR. HARDY: I am going to work backwards. I think the last one may be easier for me to answer. I think there are a lot of different ways to use the HIS.
I think this is just one way, if you want to look at some data very early on for a small time period. You are going to be limited in the amount of detail you can do.
You can take the data we have already released, you can combine several years worth of data, which we have done, and then drill down into some of the smaller population subgroups, some of the less frequent health events and things like that.
This is not meant to just be the only release we will do from the HIS or from NCHS. It is just meant to, if you want data early and quickly for a small time, for a fairly recent smaller time period, then you are not going to get the detail.
If you want the detail, you are going to have to wait, but that is possible. We are certainly trying to push forward our own internal analytic program.
We are also continuing to support outside users of the data to do those kinds of analyses, and we have produced some of those.
I know we have done a report on Asian subgroups, on Hispanic subgroups. We are coming out with one, hopefully in the next few months, looking at foreign and U.S.-born blacks and whites.
We are continuing to try to do those, make those uses of the data. So, this is only one small part of a variety of things that one can do with the health interview survey.
So, no, with this we can't focus on a lot of detail. That is one of the down sides and, frankly, if people don't end up thinking that is useful, okay, or if it is only useful for certain things.
Again, it is going to have an impact on what we ultimately do, because we don't want to just crank out a lot of numbers that are so broad that they aren't really very useful for people.
In terms of our overall goal and how this fits in, I mean, we are aware of what is going on with the 2010 health objectives and those leading health indicators, the disparities, eliminating disparities activities, and things like that.
We have tried to incorporate some of that. I agree with the first comment that we have to make real clear how this fits into some of the other things that are going on in the department or people are going to be confused.
I don't know that I had particularly given that a lot of detailed thought, but I think that is something we will definitely need to do.
Obviously, we have used some of that information to guide some of the selection of some of the estimates that we are going to look at, or ones that we hope to look at in the future.
It is limited, again, to what the health interview survey can do. Whether some of the other data systems at NCHS can do some of this and fill in, there are some of the other data systems that are providing information for some of the other objectives.
Again, HIS doesn't do it all. I don't know. Maybe we can try to be a bit more coordinated. Certainly I am willing to take any suggestions on what folks would think about kind of packaging this in a way that people understand what we are doing and know how it fits into the larger scheme of things.
DR. NEWACHECK: I think this is a very important undertaking and one that can bring a lot of visibility to the center.
One of the problems, I think, over the years with the health interview survey is that there is a fairly lengthy turnaround time for it.
If we take, for example, health insurance, it has been the case that, for years, people have used data from the current population survey when they produced national estimates of the uninsured population, even though the questions used in that survey are terrible by comparison, methodologically, to what is in the health interview survey.
Because there is an additional year's lag, typically, in the release of the health interview survey, people just don't use it.
This kind of effort to bring this information out quickly and even scoop the CPS, for example, on health insurance potentially, I think is very important, could bring some very positive visibility to the National Center for Health Statistics.
I think it is also going to be very important to think about the dissemination strategy very carefully.
I think just putting it on the web site, that is a good idea. Going beyond that to some kind of release that gets out to the media, press releases and maybe selected mailings to policy makers with a standardized format of glossy pamphlet or something like that, would also be very helpful.
Then with regarding to picking the indicators, it might be worthwhile to go to the Data Council with this and present it, to see what the different members of the Data Council think about which would be the most important indicators to include in the first set of releases.
DR. HARDY: I think, again, at least one of the messages I took from the meeting we had in November was try to get something -- not to not give it thought but get something out quickly, go ahead and make a start and then kind of evaluate from there. This is our initial thought.
We had also discussed at the meeting about whether we actually wanted to release more like little reports more quickly, something more than just some graphics and a few bullets with this, that I think could be more media ready or provide a little bit more meat for other folks who want to use the data. I think we need to think about that as well.
DR LENGERICH: I agree with a lot of the sentiment and comments that have already been expressed. I think it is important for HIS to get the data out more quickly, and I think that is to be applauded.
Being involved with another indicator project, I think we do have to be very careful about the adjectives that we apply to those and to think very carefully through that, whether it is sentinel, key, core, leading, whatever it is.
They have similar but dissimilar connotations and it is important to be consistent.
Those are my comments. The question that I have is for the unresolved issues. I am just wondering if you have thought about a process to address some of those.
Are you going to establish particular criteria or an advisory group or the Data Council or other processes to address some of those issues, which are the indicators, how do you distinguish important trends, meaningful trends, things like that. The question is, have you thought about a process for that.
DR. HARDY: Yes and no. I think we will not have the funding to do anything organized like we did last fall. I think we are going to have to rely on something perhaps a little less formal unless we can secure additional funding to continue to have a group, or get a group together and have it meet on a regular basis.
I think we are going to have to keep on looking for places to go to talk about this and get feedback. Again, I would welcome any other suggestions for that.
DR. LENGERICH: One might be just to, if you have money to pull some people together, just to establish some of the criteria that you would then apply in a repetitive fashion, to help answer some of those questions.
MR. SCANLON: A couple of things, Ann. Actually, in general, I think this is a good direction in terms of focused analysis and faster turnaround, but there were a couple of issues.
I guess one of the last things we need is to have more estimates about things over which there is already confusion.
I think we will have to have a certain amount of message discipline. I think this is a point that others have raised.
If we have good estimates or we have several estimates, do we really want to introduce another one, particularly at an intervening sort of point.
There is the question of criteria. What are we selecting and why.
The other thing I thought where this could be quite useful, although again it is kind of pushing a household based survey a bit, many of the measures that you pointed out there are candidates are health status measures, largely -- not all, but many.
I think we have similar needs for understanding changes in the health system besides health insurance, although that is an important one, perhaps about usual source of care or changes in the arrangements or visits -- how visits may be changing, that sort of thing.
I realize that the respondent may necessarily not be the best source of information on some of these distinctions, but in many cases they are.
If we are looking for sort of early warning of changes, that could be investigated further, that we could perhaps push the envelope a little bit here, to see if there aren't health systems trends emerging that we would want to monitor.
Then something I have always thought about for the health interview survey -- let me use this occasion.
I have always thought that the next step would be to reserve, on the health interview survey, some time for experimental or quick response sorts of issues.
This is not exactly that. What it would require would be reserving time, fairly quick turnaround time, for people to come perhaps even during the year to try to get estimates.
Admittedly, you would not necessarily publish these are official estimates, although you might, but it would be sort of a quick response kind of a mechanism where you reserve time on the HIS for things that you could accommodate very quickly.
Then I think you would have put together many of the pieces that one needs for immediate turnaround, short term turnaround, then longer term analytic and in-depth kind of work.
Finally, in terms of Vickie's point, the Census Bureau -- and I know NCHS has done this, too -- often publishes statistical briefs that are fairly short, they have got all the press context and everything else, and hey are fairly analytic, and they only try to get one or two points across.
They pick an area or they pick a population or they pick an age group or whatever, and they try to provide information that you are not going to see otherwise, and certainly you would wait a long time for otherwise.
These are just some ideas. I think we would be happy to offer the Data Council as a sounding board for the general direction and for possibly opening up the indicators themselves.
I don't even think I would refer to these as indicators. I would just refer to this as just quicker types of data.
Some of these, I wouldn't necessarily promote them to the class of an indicator. It is just good data to have.
DR. FRIEDMAN: I do think that Jim's last point and Gene's point is really important. Most people think in terms of sentinel indicators as essentially the encephalitis chickens and the west Nile virus crows, and if that chicken keels over, you know that you are going to have a problem in your city or town.
Having said that, I think it is very important to distinguish between the goal of getting the data out more rapidly and the goal of having sub-year estimates.
If you are really thinking of these as sentinel indicators -- i.e., something bad is going to happen -- sentinel indicators are usually time sensitive with a fair amount of time variation.
If it is possible to do that within HIS, and if there are some things, whether it is within the context of leading health indicators or health system indicators that can be done in terms of true sentinel indicators on a sub-year basis, I think that would be terrific.
I think if it is going to be more getting the data out more rapidly, but not really sentinel indicators, if you are not going to be having chickens there, I think you would be better off not calling it sentinel indicators.
DR. HARDY: I think there was some debate about it. We actually had some discussion about that during the meeting in November. It also related to the discussion about whether to do the same ones every month and we would put people to sleep after a while.
Some things don't change much or don't change enough that, woo, all of a sudden it is a big weighty problem, whether the NIH is really that sensitive to some things.
I think we were encouraged to look for some things where we could actually see some change as a result of something else going.
The unemployment rate and its impact on health care was one that happened to be mentioned. There may be others.
I think we are not sure. At least I don't feel that I am sure whether we do, in the HIS, have things that would really sound the alarm when you looked at it. I don't know.
Given that, we probably do need to clarify the terminology so that it is not, again, sounding like more than it is.
DR. COHN: We have a question. Could you introduce yourself?
DR. HANDLER: Hi, I am Aaron Handler with the Indian Health Service. I am a staff person to the subcommittee on populations.
I have a simple question. Will the health interview survey be modified to ask a question on multiple race, the way the Census Bureau has identified?
DR. HARDY: Yes, it already has been.
DR. HANDLER: How are you going to go back in time?
DR. HARDY: What do you mean, go back in time?
DR. HANDLER: How are you going to show trends for 1987 on? That is the question.
DR. HARDY: I thought it was an easy question.
MR. SCANLON: Actually, Aaron, if you really want to comment on this, there is an OMB notice current out that you can send in comments.
This whole issue of bridging from the old classification to the new is not just an HIS problem.
DR. HARDY: Actually, the HIS, we did collect data, unlike a lot of systems, we did collect data on more than one race in the past.
We then had a question asking people, if they did report more than one race, could they identify a main race, and we have kept that question on now since we changed the question wording to be a little bit, I think, more in line with what the new OMB guidelines are regarding that.
I think actually we may be in better shape to bridge than some other surveys.
MR. SCANLON: Ann, are you asking the new standards question format now or is it next calendar year?
DR. HARDY: Yes.
MR. SCANLON: It is this calendar year?
DR. HARDY: We did some of it in 1997. I am not sure we did it all. Then I think either in 1998 or 1999, we have the new groupings as well as telling people they can report more than one race, if they want.
DR. COHN: I think Vickie has one final question. I am going to try to wrap up after that.
DR. MAYS: I think part of this is, once the issue of what we are talking about in terms of the indicator gets resolved, some of these other issues might also get resolved.
One of the things is, for instance, when I looked at indicators that you are working on and some that are coming, there are issues around some of them.
I am hoping that what we want to do is try and address kind of future trends and not necessarily rely on kind of what we have done in the past.
For example, the question which is very simple -- self-assess health status -- we already know that there are real ethnic differences in that.
I guess what I am hoping is that, when those kinds of things are put out there, they are put out there also within the context of pushing us toward things that we already know and trying to advance that.
Even your AIDS questions where you ask about blood tests, well, we have some other tests that we use now. Again, it is like, if we look at it -- I hope what we are doing is thinking about where we want to go, rather than documenting where we have been, especially because it is just a short-term look.
DR. COHN: Ann, I want to thank you for coming. I am sort of struck, as we are talking about all this, I am hearing the committee is very much in favor of anything that can move forward time lines and make data accessible more quickly.
Obviously, there is the usual, the devil is in the detail, in terms of exactly what you are going to be measuring and what you are going to be publishing.
As you were talking -- and talking about really sentinel indicators -- I was sort of struck that tomorrow Denise Koo is going to be talking about the national electronic disease surveillance system.
Obviously, I am sort of beginning to see sort of a spectrum of things that are going to be coming out from CDC, various branches, HHS, period, the data dealing with various periods in time.
You may very well want to listen in just to sort of see how this all fits together.
For myself, I am beginning to reflect on the comment that Jim Scanlon made this morning, where he was talking about a strategy group in the Data Council that has a strategy. I sort of went, oh, yeah, a strategy.
What I am beginning to see is a whole bunch of systems with all sorts of different kinds of time lines, all different sorts of names, perhaps beginning to come together into what may actually be a strategy for at least some aspects of health data.
It may be -- I am not saying we will -- but it may be that after we hear Denise tomorrow, that we may decide that it may the time to ask HHS for a more formal discussion of their strategy, in terms of dealing with health data in the United States, recognizing that there is need for both acute and chronic data, data that is reported from events that happened yesterday and events that happened a number of years ago.
Now, with that, I have the happy job to adjourn us. We are going to break into subcommittees. We will be here for the standards and security subcommittee. Then I believe on the fourth floor, for the populations subcommittee.
You all know that the privacy and confidentiality committee comes back together at 8:00 a.m. tomorrow morning and 9:00 o'clock starts the NHII work group.
Then at 10:00 o'clock, we reconvene for tomorrow's activity and John Lumpkin will be back. Ed Sondik, final comments?
DR. SONDIK: I just wanted to sort of reinforce what you said, and I didn't hear Jim this morning, but the strategy point, I think, is really very important. I think that also fits with the 21st health statistics very much.
My view of this particular topic is that we have got something that, in the past, has not seemed, I think, all that dynamic.
In point of fact, I think it is pretty dynamic -- i terms of the HIS, I am speaking now. I think it is important to try to do as much with it as we can. I think that for a variety of reasons.
I do think sort of the notion of an ongoing advisory -- bad word here -- but sort of an editorial body, giving advice on what to do when, looking ahead at what the critical issues might be, would be very useful to us.
I have spoken here before about the need for continuing advice. This is an interesting point because this could be something that would be on a more dynamic basis, allowing us to make more immediate use of the data that we are actually getting.
That is true with HANES as well. We have changed HANES so that HANES is now done on an annual basis.
Data from HANES came out this year within six months after the end of the year. So, we may be able to use these large surveys that, in the past, have seemed like battleships, in terms of being able to turn them.
I think we may be able to use them in a more flexible way than we have in the past. When Ann starts bringing in the chickens from the field, then we know we have really made it.
DR. COHN: Barbara, we will let you make the last comment before we adjourn.
DR. STARFIELD: I heard a report from the Casey Foundation yesterday and the birth outcomes are worse in this country over this last decade, 1990 to 2000, low birth weight in particular, prematurity.
Yet, all the things that we think of as going into it have gotten better, smoking, prenatal care and teenage births. So, so much for the relationship between things we measure and the outcomes.
DR. SONDIK: We brought that up in the Healthy People discussions, that there is a great gap here between what appears to be the input and what we think the output should be.
Agenda Item: NCVHS 50-year History
DR. COHN: A final comment before we adjourn, on another topic. Paul, you notice on the agenda there was an action item around the 50-year NCVHS history.
Actually, at this moment, we have lost a quorum for the vote. So, we will have that vote tomorrow when we have regained our quorum.
Okay, we are adjourned to our subcommittee meetings.
[Whereupon, at 2:05 p.m., the meeting was recessed, to reconvene the following day, February 22, 2001.]