DEPARTMENT OF HEALTH AND HUMAN SERVICES

The Wyndham City Center Hotel
Potomac Room
1143 New Hampshire Avenue, N.W.
Washington, DC 20037

Proceedings By:
CASET Associates, Ltd.
10201 Lee Highway, 160
Fairfax, VA 22030
(703) 352-0091

P R O C E E D I N G S

DR. FRIEDMAN: What I would suggest we do is review the comments from this morning, as quickly as we can. If there are any that people have get reactions to the comments and, then, decide how we want to deal with them tomorrow. Whether or not we want to suggest that we have - there are a variety of options.

One option is we could defer asking people to report on this type. Another option is we could ask for approval of the report, pending final review by the executive subcommittee some changes. A third option is that Ted, Jim and I could be sent back to rewrite the report. At which point, I think that

DR. FRIEDMAN: I do not either. I have divided the suggestions and comments on the report into two categories, one of which is chapter 1 to 4. The second deals with the recommendations. I have listed four comments under each.

For chapters 1 to 4, there was one comment about developing a graphic illustrating the relationship of those statistics to the NHII. In fact, we, really, already have 95 percent of that graphic in the NHII report. It can be easily adapted. I started to adapt it a couple of weeks ago for another presentation, anyway. That would be a very good addition. We should do that.

A second is the relationship what I heard was the relationship of health statistics to NEDSS, which is the National Electronic Disease Surveillance System. We mentioned that in the report. NEDSS is, basically, architecture, at this point, for infectious disease reporting and, ultimately, hopefully, for other disease reporting systems. I do not see it, frankly, as being related.

DR. LUMPKIN: I snuck in. I was thinking that, maybe, one of the ways to do that would be to have a foreword from me, as the Chair. In that foreword, I would describe the interrelationship with this project, the NHII, NEDSS and EMRI. Or, to give the overview as the foreword.

MR. SCANLON: Yes, but then again, it not so much NEDSS as the whole belief for that statistical system. It does not matter whether it is the NEDSS framework or not, that is what we would have to set a broad, foundry force.

Say that it relates to this, or it encompasses this, or whatever. It is not so much NEDSS being as it was whatever we have now, it is a national reporting system for disease and other things.

DR. FRIEDMAN: We tried and, I hope I do not sound pretentious on this we tried to directly deal with that issue in the first chapter of the report, where we differentiated surveillance from public systems. We can make that differentiation more prominent in the first chapter if people feel that it is necessary.

DR. MAYS: I just think what John is talking about is echuan(?). It was like the day Washington said those things. I, suddenly, had a different head set.

DR. FRIEDMAN: I agree with what John is saying. I am just trying to make it legitimate.

DR. MAYS: I am saying he is thinking a little different than you are. He is thinking of a broader spectrum, which is helping people, using NEDSS, - an example of "as opposed" to the "same as." I think to start that way, assuming they have it at start time, gives a sense of why the committee is doing this and what other relationships that are important to think about as you read it.

DR. MAYS: I do not think they need to redo what you are saying. What he is doing is very different, but, very important.

MR. SCANLON: You, ultimately, will have to decide where this will fit in. Are you encompassing the re-reporting systems in your organizational proposal here, or are you in something different?

MR. HUNTER: that compares one to the other, on page seven. There is a clear, interactive relationship. It is not one or the other. You do not have to do a then diagram or they do not overlap. One is a mechanism for collecting, the other is the newest.

MR. SCANLON: The issue is in the foreword, to just try to lay these out, big tectonic plate things that help people understand where they are coming from. Maybe, we could all look at this again, If it does not, on second look if it does not with surveillance and the right context I do not like the term surveillance but, the identification of cases is different than numerical population-based statistics. Beyond that, everything else can be very highly related.

You use surveyors to do surveillance. In some cases, you use surveillance systems to aggregate statistics. I am not sure that I would not want to see us try to divide the world. We are looking more towards ways that we can integrate some of these concepts so that you can use whatever systems you have to aggregate the fiscal information for the purposes described here. If it does not work, then, you could go back and do it. The foreword helps place those things. I would not dwell on it.

SPEAKER: It is boundary issue. It always helps to do that. It is indicative of both ways.

SPEAKER: Do you think that you can point this agency data from wherever it came?

SPEAKER: We are not talking about statistical agencies. We are talking about the concept of what are health statistics.

DR. LUMPKIN: We will get to the recommendations. That is what the clarification needs to be, but, the differentiation between surveillance is, really, all about individuals. Where it overlaps is if I aggregate in surveillance data and you enter the health statistics.

MR. HUNTER: If we are not talking about who collects the surveillance information, I do not see that there is a boundary.

DR. FRIEDMAN: A third suggestion relating to chapters 1 through 4 was to change the title to, for example, Population Health Statistics.

DR. NEWACHECK: I disagree with that because it is much broader than that. You have got some health utilization with expenditures and all that, as well as population.

DR. FRIEDMAN: A fourth issue is related to when to sample and when not to sample. For me, this is a daily issue. I am not sure there needs to be

DR. FRIEDMAN: You are not that far off of an important point. The vehicle I am not sure that needs to be done. I do not think it needs to be done. If people think that is important, then certainly, add a sentence or two relating to that.

DR. LUMPKIN: It is an important issue, but not one that needs to be included in the report. That is all part of deciding. You have got resources, you have got a goal. What is the method to go to the goal? That is more methodological than we want to get into.

DR. FRIEDMAN: Let me move on to recommendations. I want to lead with what seems to be the most troublesome collapse. There were four comments I had written down relating to the recommendations.

Vickie, one was your comment relating to the second recommendation on training, which was specific to people who are currently in the enterprise. I have no difficulty making that as broad as the earlier one. It is what we were trying to address in that second one that was right now, the mind set of many people who deal day-to-day with getting the data in, getting the data out, analyzing it. It is not population health, but, very much a particular disease, a particular data site. That is what that was addressed to.

DR. LUMPKIN: How about if we were to reference the earlier recommendation, saying, in addition to the steps recommended there, there is a specific need to do what is in this one and, then, it links those two together. That way, you will not have to do as much rewriting.

Another point related to the paragraph on longitudinal and, I think that can easily modified. It was not exactly sloppy, but, I understand why it needs to be stated more carefully. I would suggest that we rewrite that to state it more carefully, to refer back to the privacy recommendation, to talk about it on an as-needed basis, et cetera, et cetera, et cetera. That is what I would suggest. We are not talking about a single constructed data site that would be constructed and stay there forever.

DR. LUMPKIN: Here is my suggestion. If someone could come up with language tonight, I would consider that one to be a showstopper, just judging from the response in the committee. Everything that we have described are editorial changes, but, that one sounds to be substantiative. If there were some alternative language, the chances of getting this through the committee would be better.

DR. NEWACHECK: The first of a few things that we wanted to say we will fund if it has to be, we will create one, longitudinally, as opposed to the lead data site. The second one was that we would provide privacy and a longitudinally provision, and make a privacy impact statement.

DR. LUMPKIN: Right. Something along the lines that when the what you said in the privacy impact statement, and so forth. Once that process has been completed, if it is determined that the benefits of longitudinal studies outweigh the risks, then, that should be able to occur. Something along lines like that, then, launch into the rest of that recommendation. But, hearing that language will be important to a few of the members who responded reticent.

DR. FRIEDMAN: What I would suggest is that I can try to draft something. Then, if anybody is available at 9:30 or so, maybe we could go over it or, at a quarter to ten, whatever.

The resource statement Vickie, you had raised a point about resources. John, maybe, that is a point that could be addressed in the foreword, as well. If the foreword will be an NCVHS foreword, maybe you could address the point that Vickie raised, which is a perfectly good one, about resources and the potential impact.

DR. FRIEDMAN: He can address it, I mean, just basically. It is not as if we are making recommendations for specific expenditures of resources. John could put some statements in there saying that this, for example, could not occur at the cost of harming existing core data sites, it could not occur at the cost of putting even less attention to existing over samples, et cetera. That is what I would suggest, in terms of the way to do it.

DR. NEWACHECK: We want to admit that this is going to cost more money to do this, right?

DR. NEWACHECK: It is not like we can do this within the existing budget, which is your concern. We try to squeeze it all in. What we could say is something to the effect that improving and strengthening the health statistic system will require additional resources.

There will be some savings, like greater efficiencies, better use, reductions of overlap, blah, blah, blah, blah. However, additional resources will be required to accomplish this major new initiative.

MR. HUNTER: It could go one step further, too, and address part of the question that it might have been Clem who asked about the point of the report. Who is the audience? We could add a few things that state that this is, as a committee, broad framework for the way multiple agencies that are party to this system ought to be thinking about this and that.

Each of them will have work to do to pick this up and implement it. It is not just that the department has to do it, or somebody else. That it is a long term thing. Putting the resource issue in the context of many parties that will have to pick up, hopefully, within some organized conceptual framework and go forward together, mindful of trade offs, or not.

The things that Dan was saying, not making trade offs for things that are important today, you need to be calling more people to action than just the committee saying that the department ought to do things because there are many the states, the private sector, are also a big part of this, as well. They have to engage in the process.

DR. LUMPKIN: Shouldn't that be in the introduction to the recommendations, rather than in the foreword?

The other thing that the foreword will have to explain is why the foreword will also I was just thinking about what I have to say about the first chapters adopted by the data council and NCHS, but, the document, as a whole the recommendations remain a document wholly of the committee, or something like that. We have to give the plausible deniability to federal agencies that are not in a position to make the kind of recommendations that we are making without going through a significant I mean, that is not saying they would disagree. It is just that

MR. SCANLON: We just cannot make in fact, when Dan presented these to the data council, our budget office said that we cannot publish anything unless it conforms with the budget. If you are not planning this this is not an empty fishhook. It is an NCVHS recommendation.

DR. MAYS: In the beginning, though, you are saying something about calling all people to the table, or something. Then, prepare it later. It does come across as at least, from a budget standpoint that many of the things you are calling for are within NCHS, or related to NCHS. What you are saying is that it is broader than that. Anything you can do, for example, to find IOM recommendations that you can say, this spirit is presented based upon several recommendations that we have seen in recent years. That gives other people the opportunity to see themselves in it and give money. Otherwise, it comes across as NCHS. Then, it is to let them solve it. Let them do their one percent evaluation money. Let's leave it alone.

For example, there is a recommendation that is made it is on page 67, where you talk about research agenda causal relationships. The federal government should support research and have health institutes, if they can be used more effectively, dada, dada, dada

DR. MAYS: that, actually, do that. I thought oh, you are going to offend NIH. They are going to say, we do that. Actually, it is more about not to support research and show how, but, to support research into the best collection of health statistics, or the best practice of health statistics that can do that. You have the NIH saying that they already do.

DR. FRIEDMAN: I know. We tried to, wherever possible, reference parallel and related recommendations. We could accord them a more prominent place. As a matter of fact, we also noted that the NIH report, basically, gave birth to the current round of RFAs.

DR. MAYS: I am just saying intellectual recommendations that you are saying, again, I think if John takes the task on of making more people see themselves, then, the reception will be better. The budget will float better.

MR. SCANLON: The other thing that is presented that has significant reverb with new patients for all state and private sectors, you cannot have it both ways. You cannot have it national and NCHS.

DR. FRIEDMAN: We are talking about adding something, both to the recommendations chapter and to the foreword on this. Is that what we are

DR. MAYS: It is just a flowery way of saying he knows what I am saying. It is not adding it this is his Ford that he is talking about.

DR. NEWACHECK: Ideally, like with some kind of, maybe, a call to action of sorts. This is a challenge.

DR. LUMPKIN: Just let me say one thing about this. When I say I am going to sign it, it is going to be collaborative.

DR. MAYS: No. We thought you were taking this when you made that announcement. I thought this was so nice. I thought, how nice.

DR. FRIEDMAN: Recommendation 1.1, pages 54 to 56, the reconstituted NCHS in fact, our language on this was inconsistent. That has been part of the problem. If we look at page 55, what we talk about is lines 6 through 14. Basically, we talk about national leadership for integrating, coordinating an integral part of the hub, interpretation of data on population health, influence on population health from a wide variety DHHS and others then, federal, agencies, et cetera, et cetera.

That language is somewhat inconsistent with what we said. I am trying to look at where the other place what we said, above where we talked. That is on page 54, where we talked about centralizing responsibility for conducting DHHS health statistics planning data collection. This is the point that Mike Fitzmaurice was responding to, analysis and translation functions.

DR. FRIEDMAN: Lines 32 through 34 on page 54, as opposed to line 6 through I do not know 19, say, on page 55.

DR. LUMPKIN: Could I ask a question? I cannot call up the whole document, but, could we use a less politically charged word than reconstituted?

DR. LUMPKIN: What it implies is that stuff is going to be taken from somewhere else, then, put into this new thing.

MR. SCANLON: It, actually, reads that way. I think we do have to be clear. You are saying more like reinvigorated?

DR. LUMPKIN: No, I am just tossing it out, that what we are, really, describing is the leadership role on all of the issues.

MR. SCANLON: We are not on policy. Why do you have one agency what exactly let me go back. What, exactly, do you want this agency to do?

DR. LUMPKIN: I am going to look at Paul because, in fact, there were a couple of spirited discussions. In fact, we may not have achieved full consensus and clarity on it. Paul, why don't you start? Then, we can pitch in.

DR. FITZMAURICE: This idea of having a reconstituted NCHS that has greater authority and a greater role in the data planning, collection and analysis process came from this notion that, right now, we have a tremendous amount of overlap and duplication. The worst part is competitiveness among different agencies.

We collect everything in a different way across different agencies. ARC collects health insurance one way, NCHS collects it another way. It is collected yet another way at CMS, et cetera. We have no comparability of data, right now very limited comparability. We collect similar statistics, different surveys. We can generate different estimates, numbers and all that.

It, basically, leads to confusion. It is the silo effect. Everybody has to have their own data collection, their own analysis capacity, their own reports on that. Then, it is competitive as to who gets it out first. Some of that is good, but a lot of it is destructive and leads to duplication and waste.

Let me ask Ed Hunter, who Chaired a team that went through all the surveys and had, as its goal, reducing costs and avoiding unnecessary duplication. I know that for our survey, we now use the MEPS. We use NHIS as the framework for the survey, the sampling panel. Ed, was that a successful measure? Is there more work to be done? The answer has got to be yes to both of them, right?

MR. HUNTER: I should, probably, say yes to both. I do not think it for us as federal people who are in the middle of that competitiveness or turf feud, to get into what the national committee wants to say they do it organizationally or politically. What is a clear consensus, even within people in the department, is that there is a lot more work to be done on integrating and coordinating.

Where the consensus breaks down within the department is how you go about doing that. Is it an institutional, organizational thing? Is it a standards-based thing? Is it a resource things? There is a whole bunch of reasons why.

DR. FITZMAURICE: It should be all of them. It is, actually, all of those things you mentioned. They are major issues.

MR. HUNTER: Right. So, the issue becomes how does one solve them. There are a lot of different ways to tackle that.

DR. NEWACHECK: Looking at it from the outside, it does not seem like this data integration effort has been very successful. The example of using the NHIS as a platform for MEPS is fine, but, it is a minor thing and the cost of equivalent things.

MR. HUNTER: Also, it ought to be broader than that because we work closely with the Bureau of the Census on data collection. It is not just a single department thing. We, probably, have to focus on what we can do as a department. Then, focus on leadership for the nation and public health statistics.

MR. SCANLON: I guess I would phrase, and again, this is up to the committee to do, but there are and without there are issues of duplication. Then, overlap and things like that. In fact, within NCHS, there are issues like that. My worry is that it is a centralized organization. It is not necessarily a 21^st century answer to the healthcare system today.

Integration is standards are leadership and coordination, but, I am not sure the idea of putting them all with organization is necessarily I do not think anyone is supposing centralization these days, but again, is that is what you really want to say.

The other thing is, I would be careful about the kinds of things you throw into health statistics. There are things that are, basically, programmatic. As a byproduct, there are statistics that, honestly, you probably do not want to think about here because then you would lose them.

DR. LUMPKIN: Again, we do not want to think about them as being consolidated, but, we probably want to think about them as being part of the standardization, coordination and integration.

MR. SCANLON: That is the result. Integration and standardization is different than centralization.

DR. MAYS: I think that is what is bothering me. To me, this is a deal breaker. If all of this goes for a vote, I am uncomfortable with it. This one piece is what bothers me. I do not think that the benefit it is like doing benefit and harm. I do not think the benefit outweighs the harm, here, in terms of the centralization.

It is like there is not the resources. There is not a lot of stuff there to make sure that this is going to be better. I think that some very strong recommendations about them tuning the process has been going on.

I agree with exactly what you are saying, that there are silos there. There are times where it is harmful because of the competitiveness. There are other times when having that difference has been beneficial. The department has been going through it has, mainly, been resource-driven with some of the integrations. It has been to save money, it has been because sample sizes are getting to be more difficult. It has not been conceptual.

It has not been in terms of a 21^st century perspective, where do we want to be? Then, taking a look at what kinds of differences that we want to have and what is okay to duplicate because of some very specific reason. Or, using technology, even at its best to be able to exchange things among agencies in surveys, and what have you. That has yet to be if all of that had been done and it was still failing, I would say maybe. But, they are are still going through the process and the leadership.

The issue is leadership. The leadership may need to be a different leadership than we have been suggesting.

MR. SCANLON: Another thing is, much of this is a reflection of resources. If you had a well-supported, general-purpose statistics agency, a lot of these other things would go away. Unfortunately, it is not. The statistic agent has to compete with all of the other programmatic activity.

DR. FRIEDMAN: I am not going to listen to that, really. The league willed to support a central statistics agency. One which is focused on a set of discreet data sites. I do not think that would do the trick.

DR. LUMPKIN: Let me toss this into the mix. Operationally, if CMS wants to do a survey that costs $15 million to $20 million, it would be much easier for them to find the money.

MR. FITZMAURICE: Easier, maybe, than NCHS, but it is difficult. They cannot tap into the trust fund and things like that. They have got to fight for it within budget in the department the base is a lot bigger than DR. LUMPKIN: The base is bigger, and the second is that is this a survey. Is this an operational tool that has survey implications? They use operational tools that have survey implications, we believe, that ought to be consistent and integrated with the overall structure for health statistics.

DR. LUMPKIN: The question is, if we see that the NCHS is the keeper of the flame, the guider of the integration the leader of what we are talking about that does not necessarily mean that they would be the ones who would implement that survey or fund it.

DR. LUMPKIN: There is a difference between what they need and how it should be done. It is a question of the data being comparable.

MR. HUNTER: I would go back to the things that are described in chapter 3 in the hub discussion. The point of the discussion there and, I have not heard, despite years of review of this I have not heard real controversy over the things we said there. We are talking about a very virtual system. We are talking about public, private, multiple levels of government, non-profit, a variety of things, where the point is to create some multi-organizational center of gravity that pulls these things together with leadership, with research, with standards development. Very much like the SDO kind of framework, where people can come to a consensus on things.

The issue in that discussion is never who collects all of the information. It is obvious that it is coming from almost everywhere. The question is, who is the keeper of the flame of the connectivity and the architecture of it, the geopolitical issues of who does what and how do they fit together.

The same thing extends here in much of the discussion, but probably, not all of the discussion. On the other side of that, there is - that kind of suggests that this be describing more virtual terms, that it be described more in leadership terms, on the top of page 55, in the direct operational terms on the bottom of page 54.

On the flip side of that, people from the outside and, we have heard this a lot in the hearings people from the outside look at this. They did it even after 1995, when we integrated those two surveys. They look at us and say, you people must be nuts. They say, you are one department, how can you have three surveys that do health insurance. Or, three surveys that do substance abuse or teen risk behaviors.

There is a counterpoint here. I am not sure you could take all three of those we talked about these at length on the data council you cannot take all those teen risk behavior surveys and just smoosh them together because they you go with different things, go with different places.

There, really, is not a part of this hub that is evolved enough to say, gee, why do they have to remain incompatible and conflicting in their results? Why can't we standardize the questionnaires? Why can't we do something a little bit different? That is what the outside folks were saying in these hearings. I am not sure they went as far as to say the data collection stuff, it is more that

MR. FITZMAURICE: There is a counterpoint to that. That is, sometimes the policy question cannot be answered with a standard question that asks a particular thing that other surveys are asking. You have to change your question to address the policy need at that particular time period.

MR. HUNTER: I do not think anybody says otherwise.

MR. FITZMAURICE: But, then, that means that the policy expertise, as well as the statistical expertise, is deemed to reside in the same spot.

MR. SCANLON: I do not think that anyone has said that. They have to reside in the same file.

SPEAKER: Though, in practice, the

MR. SCANLON: There is one distinction we could make that might help. You may or may not agree. It makes the distinction that needs to be made a little clearer. It is the general purpose versus the distinction. If you said and, you do not have to say NCHS, you could have just said that HHS should create an entity, whatever, that pulls together the general purpose statistics activities of the various agencies. That forms a virtual hub for the other data related things. You would have less of a problem.

Not doing that, you are saying a national center for statistics. If you added the words, in terms of the responsibilities to general purpose, you would get less resistance because people are willing it is not that they will stop their survey, but, they may be willing to say, yes, general purpose health statistics are the purview of and even our I could, probably, say this for the purview of NCHS.

DR. MAYS: That is the general purpose.

SPEAKER: Yes, it really is.

DR. FRIEDMAN: They say it is research.

MR. FITZMAURICE: Maybe, advice on general purpose surveys. I do not think we would say it did make it in general purpose surveys if you did. It is you have all the surveys out there.

MR. SCANLON: Yes, but, it has already been designated as general purpose health statistics ideas.

DR. FRIEDMAN: General purpose is a term that means something within the federal structure. I have never heard anybody else use that. I do not know - not that I am being argumentative I do not know what it means.

MR. SCANLON: It is a term that L and B has used. If that is your specialty, then, designate certain statistics to agencies as a first among equals, or having the central responsibility.

SPEAKER: I have no problem with that function as you described it.

MR. SCANLON: If you say health statistics, you are talking about an awful lot. If you say general purpose health statistics, I think that we speak with federal but again, it is up to you. I am getting the sense that you have a 21^st century framework and an early 1990s or 1970s sort of a recommendation. It elects to centralize everything. That is not the way healthcare works in the US. It is not the same as it was.

DR. FRIEDMAN: No, no, no. What you are hearing are somewhat different thoughts. Personally, I would say that we need an agency to coordinate the plan. We need an agency to work with the rest of the enterprise on taking a look, overall, at what in God's name is being collected and what is not being collected. Where are the gaps, which in some places, are really huge? We are just off base. We are not focusing attention on it, to work on coordinating analysis.

We are not, necessarily, doing all the analyses. But, to make sure that appropriate analyses get done and that there is some et cetera, et cetera - not necessarily a large collection. That is what I would go for. I do think it makes sense for NCHS.

I agree with what Paul said several months ago. I do not think it makes sense to create a new entity. We do have an NCHS.

DR. NEWACHECK: NCHS has a tradition of being a public service agency, too. That is, it is not a research group or an agency with an ax to grind. They have that history.

DR. LUMPKIN: But, also, within the federal government, I think, maybe, the fourth step is and, this is, probably, the wrong language but, to create a form for adjudication of these differences. I will go back to one that now, I cannot remember because Paul, earlier, had mentioned the youth surveys. There may be a legitimate reason for each one of those youth surveys, but, I am telling you that the schools are now saying, sorry, we are over-surveyed, we are not going to do it anymore. Whatever the legitimate reason is, it goes out of the way of the fact that if we do not have a combined survey at schools, that is the reason we are not going to have a school survey. Whoever is going to be the last in the door is the survey that is not going to be completed in a statistically sound fashion.

MR. SCANLON: Who decides it?

DR. LUMPKIN: Who decides which is the one? I do not think anybody, internally, is going to be comfortable with saying, okay, we just turn that over to NCHS.

MR. SCANLON: Of course not.

DR. LUMPKIN: But, people might be willing to say that it might be an NCHS job to create the mechanism to resolve that. Then, present it to the data council, or to the higher-ups or some sort of mechanism, so that there is a forum for these issues to be resolved and worked out.

My guess is, as we are looking at it internally, to try to create three different masses where we are trying to create a single survey. There are some reasons for the question to be different than other areas. The reasons do not seem to hold water.

SPEAKER: They are not even federal surveys.

MR. SCANLON: Again, I do not understand this. And, again, this is your report, you make the recommendations. Why do you think putting one agency in charge over other agencies and, the general purpose statistics agency, at that is a better way than some others? Not that I know the other way, but, you have got a very specific solution to what is a much more complicated system problem.

MS. GREENBERG: I know I came in late, I started in the other meeting. Going back to what Ed said on page 53, it says in order to realize the vision, NCVHS believes that priority needs should be replaced with strategies for building a stronger integrating and coordinating hub for the health statistics enterprise. Is there disagreement around the room on that?

MR. FITZMAURICE: It said what are the functions of the hub.

MS. GREENBERG: If you look back at this, you can see

DR. FITZMAURICE: I am making head at the bullets as the functions on that.

DR. NEWACHECK: The real issue is, if we get into operational issues like data collection, analysis and all that, then, we are going to run into resistance. It seems like that is very clear. Maybe, we need to be a little more general and just say leadership. Leave it a little vague, now.

MR. SCANLON: [OFF MIC] for creating common analytic resources, or something like that, because then, I do not think that they will resist us much.

MS. GREENBERG: What I was going to say was that this first bullet could be revised to say assignment.

SPEAKER: Which bullet are you referring to? What page?

MS. GREENBERG: I am on page 53. I think this first bullet could say, instead of assign overall responsibility for health statistics, as has been suggested, it sounds like that means all of the collection, all the analysis, everything. You could say but, maybe this is not going to satisfy everyone assignment of overall responsibility for integrating and coordinating health statistics activities. That picks up from here that that is what you are looking for more integration and coordination, which I do not think is a 1970s kind of thing.

I thought that the state centers for health statistics the division that they would not collect everything or do everything either. They, ideally, would serve as this kind of coordinating, integrating hub. If that is a reasonable vision at the state level, it seems that, at least, it is a reasonable vision asset at the federal level. Now, this is not going to be easy. Certainly, the data council has had, I would say, very little success at doing that.

MR. SCANLON: Coordination is hard without authority.

MS. GREENBERG: But, then, it has not had any resources from the point of view of any program, or any of the surveys, or anything.

MR. SCANLON: Hold on a minute. Who said that it is a state? You think that the state helps the center idea from the 1970s was the right way? It is not the folks in the state, necessarily.

DR. FRIEDMAN: You know something, Jim? One of the frustrations -

SPEAKER: I am talking about the 1990s view of the first of them.

MR. SCANLON: Maybe, there is a 2000 version of what that is.

DR. FRIEDMAN: I think one of the difficulties I am going to say this and let's get back one of the difficulties now is that, in fact, operationally, you have got more coordination, more integration and more centralized responsibility at the state level than at the federal level.

Essentially, you have the states holding the bag for the lack of integration at the federal level. For me, where I have 20 different data sets all reporting to different parts of CDC, it is a real problem. We have to, then, go back to CDC and say, would you guys coordinate your own

MR. SCANLON: That is programmatic data. That is somewhat different.

DR. FRIEDMAN: No, it is not.

MR. SCANLON: Are you going to have them reporting to NCHS?

DR. FRIEDMAN: What I would let's -

MR. SCANLON: Could I make a

DR. FRIEDMAN: Yes.

MR. HUNTER: I do not think the general purpose thing is meaningful. I do not think it means anything, even to people within the field. Many of the things Dan is talking about could be described as general purpose statistics once they are taken out of their initial collection context and used for some other purpose. I do not think that solves the problem.

What I would suggest doing there is, probably, more of a consensus on the need to make a statement in these recommendations that the status quo is not working effectively. There is a lot of examples of

MR. HUNTER: I think there is a lot of consensus that the committee from what I am hearing and, certainly, from the hearings and all the commentary we have had on the report to date, if someone needs to do something about the organization I do not want to use that word the lack of integration and coordination in leadership, consistency and standardization from within the federal government and the department, in particular, in a committee, at a minimum, seems to want to say that.

Maybe, there is a way to change the very first thing to say that it is a very clear indictment that the committee recognizes from what people have said, that it is an indictment of the existing status quo in the department.

The next step is that the department needs to establish some effective mechanism for doing this. Maybe, there are some numbers of these functions that are there to be built upon in existing institutions, like NCHS. Then, there is a continuum up to organizational issues, centralization, things like that.

I would not say I think building up from the bottom and seeing where the consensus ends is, probably, a useful way to do this. I do not want to lose the clarity of what everyone said over three years that the federal government needs to figure out some way to do this more effectively or the rest of the field cannot follow because there is no one to follow.

I do not think that we could ever be as specific, in any kind of a committee or collaborative process, on what the solution is. There are too many I mean, Mike and I are never going to agree, although, probably, we agree more often than not on this because the organizational issue is the sum total. We cannot say things like NCHS, or anyone, should be responsible for coordinating and integrating without being given any additional responsibilities, empowerment or line item control. It is the same reason why the data council has not been effective. There are too many competing power centers, there are too many reasons why you cannot, just by force of will, say let's all get together and solve this problem.

Somewhere in these recommendations, what the committee is trying to say is that the department needs to empower some process, some place, to just think this through and do better. Maybe, it is study. Maybe, it might be the secretary shall, but then, delays things. We are clouding the named organization with the issue that the committee is trying say it is broken, and it is not being fixed. The committee wants the department to step up.

MR. SCANLON: Even in then, in the more say that the Secretary should you could say, have a look at alternatives, or develop, but, you could say an organizational framework within HHS that brings together the integration and the coordination that is envisioned under these points.

SPEAKER: That sounds like the data council.

SPEAKER: It does. To me, that is a committee.

SPEAKER: It sounds exactly like the data council.

DR. NEWACHECK: A reconstituted data council.

MR. SCANLON: A committee cannot, just as NCHS cannot. There is no one who can say no to another agency. A committee, certainly, cannot. A committee can do a lot of good, but ultimately it can re-pave things.

DR. LUMPKIN: Here is the difficulty I have. First of all, unless we have a lot of good, sound reasoning, which I do not think that we I will not say good, sound reasoning. That is not what I want to say. Unless we have reasoning that is immediately apparent to the reader, they will get to recommendation one, meaning the federal readers, and, all of a sudden, their turfism will come out.

SPEAKER: The report will go to the side.

DR. LUMPKIN: And, it will go to the side. Just as a side comment, I am reminded what Phil Lee once said, that he always knew that data was a four letter word. He just never knew it was spelled T-U-R-F. I am not sure that our recommendation, although we would like to get around that, all the good stuff in the report ought to be held hostage to trying to overcome what may not be overcomeable.

The other piece that I am concerned about is that what we want to say is that there ought to be a focus at this level of the Secretary, who has responsibilities. Perhaps, it was a mistake to move NCHS out of the Secretary's office. Given that

MS. GREENBERG: It was within the Office of the Assistant Secretary. It was never in the Office of the Secretary.

DR. LUMPKIN: Or, the Office of the Assistant Secretary. But, that was when we, actually, had an assistant secretary who was an assistant secretary. If we look at the sum total of the recommendations from this committee, we have already put our stake in the ground about this new secretarial level of responsibility for NHII.

We are going to start looking a little bit foolish if we keep on, every time we have a problem, our recommendation is a new secretarial level entity. I do not know what the answer is, but, that puts some constraints on what we do as a recommendation. It may be that what we are talking about we talked about the things in the middle of that diagram

SPEAKER: The hub.

DR. LUMPKIN: The hub, yes. That was what our recommendation is, that either a new or existing or that there be a virtual hub created within the organizational structure of the department to carry out the tasks of coordination and all of the things that we have talked about and, also, given the authority to implement those decisions.

WINZIMER: Just as an observer, Dan and Paul

look like the ones who are most frustrated. They are the only two people who are on the committee, who have been working through this process for the last four years. It seems like we are going completely backwards.

You want to say something strong and specific. Here we are, all telling you, you better not do that for all these reasons. Why would you listen to us? I do not understand that part. If it made sense, you would not need any of this.

DR. LUMPKIN: They do not vote it out. The problem is that we took it to the committee

WINZIMER: We are listening to everybody, but these two guys are unhappy. We are doing all the talking.

DR. LUMPKIN: We took it to the full committee this afternoon. There was some resistance to this recommendation.

DR. NEWACHECK: That resistance came about the data collection part of it. That was the primary resistance that I heard.

MR. SCANLON: It was more than that. Once you tell one agency they are in charge of all the other agencies in an area, you have to be very you are talking about FDA and

DR. NEWACHECK: That seems like a rather central part of this whole 21^st century vision is to have some kind of leadership.

MR. SCANLON: I do not think that is a 21^st century way to approach it.

DR. MAYS: Then, what is?

MR. SCANLON: I am saying this document.

DR. NEWACHECK: And, work on integrating it.

MR. SCANLON: Leadership, distribution and integration.

SPEAKER: Centralization is

DR. LUMPKIN: Given the hour and, I do not want to because Dan is Chairing this committee I do have an operational suggestion. If you believe, first of all if the desire is to get this out tomorrow, what I would like to do is two things. First of all, separate the first four chapters with the foreword and the recommendation. Then ask for a vote on the first four chapters and with the foreword because I do not want that to come back to the committee again. If it comes back again, we will find other problems with it. Every time we have it come back, there will be other problems. Then, we make the suggestion that if you want to try to run it through the committee with the leading data collection as it is and see what the full committee's response is, give it a try. The alternative would be, then, if that fails, we will have to bring it back and bring the recommendations to the meeting in September.

DR. FRIEDMAN: Let me make a suggestion. Then, I want to hear Paul. My suggestion is that we go for chapters 1 through 4 and the recommendations, except for 1.1 and the longitudinal. Frankly, I do not particularly want to I am tired. I do not particularly want to hold up the rest. I, also, do not want to screw up rewriting it and have to deal with that. I would rather try to do a good job.

In terms of the other, initially, Paul, I, Barbara and Vickie need to revisit it. I do think that this, probably, is genuine disagreement. My guess is that Paul and I do not agree, with all due respect, with what we are hearing from our federal colleagues. Hey, Paul. What is your

DR. NEWACHECK: I do think this, in my view, is a real opportunity to give a boost to an important agency that has gotten neglected on the vine and that has the potential to be a leader in this area. I would like to see that happen.

DR. LUMPKIN: What do you think about Dan's suggestion?

DR. NEWACHECK: I think it is final, though, I do not know how we are still going to resolve it. I like the idea of taking out the specific words of data collection planning, blah, blah, blah, blah, and substituting leadership and leaving it vague.

DR. FRIEDMAN: I would like to take out data collection, but, put in integrating, coordinating and planning. There has to be something about temper setting because somebody has got to set an agenda.

DR. LUMPKIN: My suggestion would be that if we do that first of all, if we take out those two recommendations, we pass those. At the meeting in September, we come back with those two recommendations fixed up. I would ask that you come back with a whole back pocket alternative for number one and vote your favorite.

If it gets voted down, come back with an alternative so we can get this baby to bed, that we recognize that we have to give the full committee all the recommendations in September. I do not think it is fair to take a recommendation out of context and, then, say that it is somewhat different and pass it. But, we will remind them that all of these were passed at the last meeting, unless there was something that really jumped out and said it needed to be changed. In light of these new recommendations, we are not going to take another vote on it.

MR. HUNTER: Can I note, just from familiarity with the structure of the rest of them, there are some later recommendations that refer back to who is going to do what. It does matter what you say in the first one. It would be very helpful to the process of getting the summary written and getting the thing produced. Then, at the last I mean, we are not going to be there until around September, anyhow. To be able to stick in a couple of things and, then, know where the changes are later would be very helpful. Leaving the whole recommendation thing until September will be very problematic.

MS. GREENBERG: I thought there was a real misunderstanding on this longitudinal data set issue and the discussion. It was almost the strongest point that was made in that National Academy meeting in well, two things in 1999. One, with all of this fragmentation, I guess, we can say too bad. The other was the issue of the last the data means you cannot look at these influences.

DR. LUMPKIN: Actually, our earlier discussion was not that we had a problem with that, but, we thought that we had a wording change that we thought we could deal with it. Dan's point is that since we are not going to get all the recommendations through tomorrow anyway, rather than trying to cobble together something tonight that would talk about referring to the section on privacy and doing the privacy statement and impact, it would just be to buff it, not because anybody wants to pull back from that one.

MS. GREENBERG: I would hate to see withdrawing from it.

DR. FRIEDMAN: No, no, just to clarify it. Then, Paul, I and whoever else is interested in this will spend a little bit of time on the phone. I do not think it is going to be that much, frankly. I will insist the people cannot come unless they read the thing.

MS. GREENBERG: What did you hear when you sent it out for comment?

DR. FRIEDMAN: Excuse me?

MS. GREENBERG: When you sent out for comments?

DR. FRIEDMAN: Come on, you have been around here longer than I have Marjorie. What have we heard? What have we heard? We have received not a single comment. I meant from inside the committee, we have received not a single comment.

MS. GREENBERG: No. I meant from the outside world.

DR. FRIEDMAN: From outside, the most substantive and helpful comment I received was from the Canadian Institute for Health Information. I mean that seriously. It, really, rates the comments.

MR. HUNTER: A point of order, though, is the idea for tomorrow. Obviously, we go back just to make sure I heard this right, so I have an idea what the agenda is go back and sort of try to move the 1 through 4 and the mention of the new prep

DR. FRIEDMAN: As modified.

MR. HUNTER: as modified, with these other agendas that Dan talked about. Then, test whether the removal of the word data collection would satisfy the issue among the committee members on the foster number one's [OFF MIC]

DR. FRIEDMAN: I think we should bring it up, Ed, but, frankly, my experience has been that unless we have got the wording there, even if they say yes, that will do it, it is not

MR. HUNTER: The reason I am asking is, rather than going to the trouble of asking that question, is it better to just say we will come back on that one in September rather than trying those words?

DR. LUMPKIN: I would not try the words tomorrow. You are better at this issue than I am.

DR. NEWACHECK: It is more than data collection, much more.

DR. LUMPKIN: Otherwise, we will spend all the time talking about this one. We will not reach any resolution. People will not focus in on the other, do the others. Say that we are going to bring it back, these are the issues that we are discussing. The way the committees look -

[SESSION ADJOURNED]