[This Transcript is Unedited]
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
SUBCOMMITTEE ON POPULATIONS
September 25, 2007
Crown Plaza Hotel
8777 Georgia Avenue
Silver Spring, Maryland
CASET Associates, Ltd.
10201 Lee Highway, Suite 180
Fairfax, Virginia 22030
List of Participants:
- Donald M. Steinwachs, Ph.D
- Larry Green, M.D.
- Garland Land, MPH
- William J. Scanlon, Ph.D
- C. Eugene Steuerle, Ph.D
- Kevin C. Vigilante, M.D.
- Virginia Cain, Ph.D
- Debbie Jackson
- Dr. Boenning
- Michael O’Grady
- Edward Sondik, M.D.
- Nancy Breen, Ph.D
- Miryam Granthon
TABLE OF CONTENTS
- Welcome and Introductions
- Report on NCHS Board of Scientific Counselors
- Surge Capacity; Followup
- Data Linkages to Improve Health Outcomes
- Summarize Discussion and Determine Next Steps
P R O C E E D I N G S (4:40 p.m.)
DR. STEINWACHS: You have an agenda in front of you. Let me ask you to take a look at it for a moment and see if there is anything that ought to be added.
Bill because of his car also needs to leave early, so I told Bill we would move up some of those discussion items involving him. I also wanted to welcome Michael O’Grady, who is representing the BSC, so that is one of the early discussion items, too.
Why don’t we just dive in? Why don’t we start with the NCHS Board of Scientific Advisors? There was a meeting recently, I guess.
Agenda Item: NCHS Board of Scientific Counselors: Highlights and Potential Areas of Collaboration
DR. SCANLON: A week ago. I attended the first day. I actually think of it as a meeting of the new NCHS Board of Scientific Counselors. You have got nine members or eight members?
DR. CAIN: We had five new members but a turnover of the chair as well.
DR. SCANLON: Right, so it was like a new board and the first day did involve a lot of discussion about potential modifications of how the Board works. I’m going to let Mike talk some more about that.
The new chair is Irma Elow, who was our former liaison, so it is their gain and our loss, because she was a very good liaison and is very interested in the work of the Populations Subcommittee. I think she will be a very good chair to the Board.
PARTICIPANT: Who is it?
DR. SCANLON: Irma Elow. Essentially my role there was to only report upon the letter that we did linkages and the fact that we are going to be talking about continue work in the area of linkages, and to talk about the work we have been doing this summer on secondary uses, because I think there is a strong overlap between the interests of the Board and in some respects the interests the Board represents in terms of research and secondary uses. We haven’t fully come to grips in our discussion this afternoon with this boundary between research and operations and the different rules that apply to each, et cetera, but that is something that certainly will be a factor that impacts both of our groups and the interests of both of our groups.
I think the most important thing is to hear from Mike in terms of the discussion that they had in terms of coming to an agreement on a future agenda. In some ways they were reviewing what they have been doing. They have been admirable in terms of how ambitious they have been in trying to review the NCHS programs on a relatively tight time schedule.
I said at one point when I was observing them do the first one, which was the mortality statistic that I was very envious and felt dysfunctional on our part because they have done everything so quickly and done so well. But in this meeting they surfaced some of the pain that they have gone through to make those things happen, and questioned themselves as to whether or not they should be running at that pace for the future. So it turns out that they are human too, and that was a reassuring message.
I think that we have talked before about the shared interest in vital statistics, the shared interest in research data and in access to research data. There was a real issue, and this should be our discussion to discuss, the potential best way for us to work together in terms of pushing forward those interests. So Mike, maybe you can talk about the Board, especially the second day, since I heard none of that.
DR. O’GRADY: Sure. There are a few things that we did, and I want to start with the one that we specifically thought we wanted to bring up with you guys rather than just general discussion.
We were going through in the presentations — there were a number of presentations because we had all these new Board members and stuff like that from different parts of the organization and the agency. One of the things that came up in the vital statistics discussion had to do with what appeared to be not necessarily a problem, but it is time to reach out to you guys and start talking a little bit about the harmonization and some of the vital statistics.
There is a fair amount of work going on in terms of the electronic death registry. It has got WHO, trying to make sure that it is consistent with that, trying to make sure it is consistent with the states, which is roughly 56 different agencies, because they are the ones actually doing it. But at the same time, thinking about some of the stuff that you folks have been doing over the years in standardization, developing standards for medical records and thinking about how you bring that together to have some sort of an industry standard.
It struck us as at least a concern to make sure there wasn’t a slippage there, and making sure that some of these records were harmonized with what was going on with medical records as much as possible. Some of the stuff had to do with cause of death. It seemed kind of advantageous to both communities in terms of knowing certain things about what was going on with the medical records that would help the people who work, especially death records, to have a much clearer picture of what the real cause of death was, both in terms of the straight research and in some of the things having to do with the homeland security and some of those concerns, what was going on out there, what was the actual cause that was going on.
At the same time, with some of the research community in terms of tracking people through the medical records and doing the research and not knowing what happened to old Mr. Jones or whatever, and it turns out he died in Minnesota, and not where his birth certificate was.
So we wanted to at least open up to your guys the idea of thinking about comparing notes enough to make sure that we weren’t heading in two totally different directions. Now, hopefully agency staff have been talking a little bit on this, but just to make sure that that didn’t slip through the cracks, was one of our main concerns that we wanted to make sure of. In terms of my marching orders from the Board, that was the main one to bring up with your guys, and make sure that like I say it didn’t slip through.
The other thing that came up, as Bill mentioned, there is this notion of doing reviews, trying to give the agency some help in terms of thinking about different things that might be — they are under tremendous budget pressures at this point, that don’t look like they are going to let up any time soon. Therefore, some of this is an attempt to look at what is going on, with an open helpful sort of notion, are there things that could be done that could make it a little bit more efficient and a little faster as some of these things go through. So that has gone through.
As Bill said, some of that is a little bit of second guessing now in terms of, it was a big push to get it done quickly. It does involve outside reviewers. So there are a number of different discussions, to ask outside reviewers to come in and examine either a survey or data analysis and dissemination functions, those kinds of things.
But there is no real budget for it. So when you think about getting a second opinion, getting that outside review or however you want to think about that, I think we all know the level of effort we put in for a journal article that no one is paying you, it is sort of in the overhead, versus some of the things we had done when I was at ASPE, where we wanted people to evaluate. So we bought two weeks of a senior actuary’s time and made that their fulltime job for a period of time.
So there was some of that discussion, the idea of, did we go too fast on the idea that some of the things that came back. There wasn’t quite the time to get a draft, discuss it, give the staff a chance. A number of the things were just factual errors. So if you are rushing a little too much, there wasn’t a chance for the staff to go, that is not really how we do it, and then rethink it and reflect it in the final product. So a little bit of that, we are going to try and open up the discussion a little bit more and have a few more draft back and forth, peoples’ chance to comment on things before we go to anything that people are putting in final.
One of the other issues that we were approached about is that there is a new OPM directive that is in comment period now. That is an interesting one. I personally found it interesting because it is an attempt to make sure that the statistical community is protected from the ill will of political back and forth.
It is pretty clear when the statistic comes out. Mostly it is the economic indicator sort of thing, play with the crop report, that sort of stuff. So it is that idea of —
DR. STEINWACHS: Of real population significance.
DR. O’GRADY: There you go, that’s right. So that notion of the firewall between the civil servants and the politicians, so that they would probably get a heads up, nobody is looking to surprise, but at the same time it is not like, if you don’t like what the poverty statistics are going to say you can hold it for a few months and wait for a quiet time, or however you want to think about that.
That is in comment period right now. We were preparing to have a letter from the Board of Scientific Counselors that turns on advice from the agency. We basically realized that that is a little bit beyond our charter, so it is something that probably individual members will be weighing in with very similar letters from a number of different people. But certainly we can comment as individuals.
But that is something to bring to your attention on that. Candidly, as someone who had been a political appointee, there was nothing problematic in here. You want that trust level, that when somebody is coming out with the uninsured numbers or the unemployment, however those things are, that everybody can — even if they are hotly debated political debates, everybody can agree the numbers have not been played with.
DR. HITCHCOCK: Can I say something about that? There are a couple of provisions in there that are kind of interesting. One of them is that, if you have got something like the Labor Department’s annual release or something, the release is not supposed to include any policy statements. Policy statements should be in something else that can come out later to reflect upon what the agency has released.
But there is something else in there, an odd statement that gives the policy makers a chance to review the release and check to see that there are no policy statements in there.
DR. O’GRADY: Right, and in terms of whoever wants to comment, if you want to comment on that in the group. To a certain degree, I am putting words in your mouth, but you strike that balance between, you don’t surprise the Board. Nobody who runs an agency is going to want to read that within their department, read about it in the New York Times, is the first time they see this statistic come out that has been released by their own organization.
So there is this attempt to try to be responsible, the same way you would embargo something and you would allow reporters to see it the day before so that they can prepare their stories, and they are not slapdash stories; they have a chance to do their homework. It is allowing the agency heads outside the statistical agency, the Secretary or the White House, to at least have a heads up that it is coming. So they can’t really touch the release, but can they have something prepared to release simultaneously.
DR. HITCHCOCK: Because they could sit on it.
DR. O’GRADY: Certainly in terms of the OMB, they didn’t give the impression that they could sit on it.
DR. HITCHCOCK: An opportunity to review.
DR. O’GRADY: An opportunity to review. So if you give somebody three days to review before — but it is different. When you think about things like the way the Medicare trustees report is handled, that is a date certain. In those cases, the Secretaries of those agencies are the public trustee, they are trustees themselves, but they have no opportunity to say, let’s release it next month instead of this month because we are down in the polls.
DR. HITCHCOCK: My thinking of it was, back in my SAMHSA days and the War on Drugs, the Office of National Drug Control Policy wanted to comment on all the findings from the SAMHSA to put a spin on it. It is different than what you were saying, where by law something has to come out.
DR. O’GRADY: Right, and the more practical is not as nefarious. We have seen from all kinds of Administrations in the past where they will hold a statistic until 5 o’clock the day before the hearing in the Senate on whatever. So it is a tool that has been used that this is an attempt to protect, like I say, a firewall.
DR. CAIN: It is also just things like — perhaps not so nefarious as that, but for example at CDC they may have a program that they want to promote, and they think this will be a good opportunity, we will hold the statistic until we are ready to put out this program or to publish the results from something and they can be compatible. I think the statistical agencies, the directive is to sever those things. Yes, you can be prepared, but it is not going to be something that can be used to promote an individual program. I was going to say, Don, I can send you the letter that was drafted.
DR. STEINWACHS: I would appreciate it. Then we can circulate it to the whole committee.
DR. SCANLON: In terms of individual members, submitting comments was the motivation to add weight to the — there isn’t a sense that this is going to be vigorously opposed.
DR. O’GRADY: I didn’t get that impression. The people who oppose it would not necessarily telegraph that.
DR. SCANLON: Was there any sense in the discussion that people thought —
DR. CAIN: I think it is hard to tell what the various departments will say because it does take some authority away from them in terms of controlling the release. I could imagine there would be some opposition, could be some opposition, but the comments don’t come in until the last day usually, anyway.
DR. GREEN: What is the agenda item?
DR. STEINWACHS: I’m sorry, we jumped down to 515. I think it is two things combined together. One is giving a report back on what the BSC did, and the second is to identify some areas where we might work together.
DR. GREEN: I’m concerned about Bill having to go. I am interested in what he thinks that might be.
DR. SCANLON: I would like to think about exploring some more of the issue of vital statistics, because I think there is a question in my mind of what you are talking about in terms of future potential of electronic health records and the vital statistics system.
We haven’t really gotten into electronic health records in that detail yet. We deal with the principles at a high level that hopefully do not interfere with useful applications like this. We don’t actually talk about the details on the application.
That doesn’t mean that we shouldn’t be thinking about a next stage or starting to do that. This is something that during this discussion of secondary uses of data, I think there has been in everybody’s mind this vision of a big picture of how having electronic data can facilitate data flows of one sort or another, create some capacity to do this task or that task.
Since we are in the 21st century we can’t use that anymore, it is like a 22nd century vision at this point. Your electronic health record, people are sitting there with pencil and paper copying down data in order to create statistics. We have got a long way to go from this idea of how we turn something into useful electronics, and that is not where we have been yet.
So I think if this is an area we want to think of for common work, it would be a new task for us, but potentially an interesting one. We have talked about vital statistics before in a different context, more of a budgetary thing, and worrying about some of the tradeoffs in terms of, if the budget is tight, where is the least harm caused, I think that is maybe the way to characterize it, and not to be thinking the way an economist hopefully should, but not to think about tradeoffs in a narrow scope but more broadly.
If the vital statistics system had marginal value that is so much greater than this other use over here, then we should be drawing money from that use to do it. I think that is another area that we talk about.
DR. O’GRADY: Two things to consider. One is that we have already seen that erosion go out. We no longer collect marriage data and abortions.
DR. CAIN: Divorce.
DR. O’GRADY: Divorce, some of those things that are more social. This is an agency that is dying the death of a thousand cuts, a freeze this year, a minus one this year. It is important you get down to the bone. So that is one of the concerns.
The other thing, in terms of the way I conceptualize these things, and I don’t want to speak for you guys, but the development of electronic medical records and the way that community operates, vendors interacting with insurers and providers, this is a different kind of paradigm, a different structure. This is state agencies collecting their way of doing things, the traditional CDC, being in Atlanta and deal with the states.
There is also this World Health Organization sitting down with Health Canada, are we consistent as we go state to state, nation to nation. It is easy to see how things would slip, given that kind of structure, versus a more market oriented, commercial structure that seems to be evolving in electronic health records.
DR. STEINWACHS: Can I push the two of you a little further? I think this is definitely an area we should be considering. I am thinking of the turnover in people in our committee, people like Garland coming on board. But can you be more specific in terms of what is our value added? Is it that we can hold hearings on this issue? How would we proceed, other than to — there would have to be some level of generality beyond what we discuss in here. What would be do that we would add to the process that is not being done either somewhere else in HHS or by NCHS?
MR. LAND: Maybe I can respond a little bit. One of the things that — this committee just the way it is structure reports to the Secretary, where the Board of Scientific Counselors reports to the director of NCHS. So I think people in NCHS know the problems, but I’m not sure the Secretary knows the problems. I think it gives a whole other forum for illuminating what is going on that I don’t think it is recognized at CDC right now, or given the attention.
There are several different issues. The reference to World Health Organization, a lot of us are saying, if we are going to go to electronic death systems, shouldn’t we take advantage of electronic systems to be collecting cause of death in different ways than we have always done it before, which we know is very poor. There is a lot of resistance to that right now.
So having some hearings on how are electronic medical records adding value to the process of collecting death certificates I think would be a valuable thing. It has not ever been done before, in terms of looking at that from that perspective.
DR. STEINWACHS: We could hold the hearings and get the letter to the Secretary.
DR. SCANLON: That was the linkage I was thinking about, between these state agencies and the private sector, the issue of saying there are these needs and the ability to interface with the private sector and to be able to usefully obtain data in an efficient way. Part of it is potentially of value to the private side, too, knowing that there is a value to designing your product in a particular way because it is going to meet the needs over here on the private sector side, which is not potentially known in the private side today because it hasn’t been discussed in these terms.
MR. LAND: The other thing that I think are two issues right now, the electronic birth and death systems have been going on now for four or five years, but they are moving at such a slow place because of the lack of state and federal funding. It is going to take over a decade for all the states to have them.
So right now, there is not going to be a national comparable data set on births for a decade, because some of the states are collecting them under the old standards and some collecting under the new standards, and you can’t put the two together. So we are going to have a whole decade of birth data for some variables, not all variables, but for some variables, that you are going to have to say, these states said it this way and those states said it that way, and you are left with that dilemma of how do you analyze the data.
DR. O’GRADY: And you will have to live with it, because these guys are going to make it hopefully 80 years or 90 years.
DR. HITCHCOCK: I think one of those variables is race ethnicity.
MR. LAND: That’s right. Education is another one. All of the important ones that you would think about. Smoking is another one.
DR. HITCHCOCK: Another area that some are addressing in the Department now is, how does all of this adoption of health IT, and this is a good example, disproportionately affect different populations. I think the fact that we are not collecting the same sorts of race ethnic data on the old method versus the new method is an area that we —
MR. LAND: Another example of that that is really paramount right now is, the CDC is saying that death data is a fundamental surveillance system for pandemic flu. It should be, obviously, but you have got to have timely data to do that.
Electronic death systems are a way to do that, but somebody has got to put some money into these if we are going to have all the states having electronic death systems.
DR. STEINWACHS: The other thing is, I don’t think it is inconsistent with where we started off with the linked data sets. It is not just the issue of just trying to make this system is redeveloped or sustained in some way or another, but the question of what do we do with the data, who do we send it to and what do they link it to. So the data sent to CDC, that offers all sorts of options for different types of linkages.
MR. LAND: Absolutely.
DR. STEINWACHS: A lot of people misunderstood, when we did the original set of hearings on linked data set, a lot of the emphasis was on the fact that people were ignoring what you could get out of administrative data. So it wasn’t just a question of linkages, but a lot of people were emphasizing going out and doing more surveys. But there is all this administrative data out there properly linked to something else to really inform us.
So I think this is a very practical example of some of the things we were trying to touch on.
MR. LAND: Let me give you just one more example. That is, Congress passed back in 2005 what is called the Intelligence Reform Act. The National Center was supposed to put out regulations within two years of that. They kept to their time frame, but it has been held up in OMB, it has been held up in the Department, at CDC, within the Department, and now it is going to get held up in OMB until its regulations are out.
Those regulations not only relate to the confidentiality and security of birth and death data, but also potentially it relates to the statistical aspects. We anticipate there could be some talk about linking birth and death data in there, which would really have great opportunity.
But if this whole thing gets politicized, as it already has been, who knows — and of course there needs to be some funding behind it and so forth, but even taking that aside, there is real concern with what has been going on with these regulations, why it has taken so long for them to get out for public comment.
So I think there is a whole series of things that you can probably take up for a long time to be chipping away at, in terms of different problem areas and about specific systems right now.
DR. BREEN: What would be your suggestion to move the public comment part of the process forward? What would be the suggestion? It sounds like they are hibernating right now.
MR. LAND: Yes. I think they are going to have to figure out which ones of these, what time line you want to take and so forth.
DR. BREEN: Was there a time line associated with it?
MR. LAND: I don’t think the Board of Scientific Counselors — the reviews that they did on the birth and death systems I think were pretty cursory. I think for the most part, the National Center wrote a report and they kind of agreed or disagreed with it, and that was it. There really wasn’t a hearing like you all have hearings, where people come in from the outside and testify. I know our organization was not invited to do that.
I’m not sure we want to go back and replay all that again because they have done that. But there are some fundamental issues with the quality and time line right now.
DR. SCANLON: I am wondering if it needs to be suggested that we have one more meeting, but I am wondering if maybe a recommendation should be made to the full committee at the next meeting toward holding a set of hearings on this issue, perhaps just stating it broadly enough that you have got a subcommittee that works on defining exactly what those hearings would be, who would be ideal to invite, what precise questions would be worth asking. It wouldn’t necessarily be repetitive of other information that had already been gathered. I wonder if that is a reasonable goal.
DR. STEINWACHS: This may be the item to move on, but I was looking for items that we wanted to put for this year. So this makes a lot of sense to me, but I wanted us to have a chance to go through other things that may be on that agenda. If this is it, then it seems to me that the next step would be just as you are saying, to talk about planning a set of hearings and identify staff who can work with Mike and others who want to be involved in that working group.
DR. BREEN: It kind of, if you want to do it, links back to one of the first things that Jim brought up on harmonization of data, medical records and death data. It would make a lot of sense to do that, and it also might bring to the fore the fact that these other kinds of data are not collected. I’m sure a lot of people don’t realize that, and they are going to be quite upset when they do.
DR. O’GRADY: One other example, when you think about FDA and their attempts to postmarket surveillance, especially with the mobility of the population. You have a record from Illinois at some point, but they die in Florida, some of that stuff of how you track people, how you get that outcome, especially if you all of a sudden after the fact realize you have got a problem with a medication or device, how are you going to link that together. If you don’t have reliable death records, it is going to be tough. That is all new policy, pretty much, for them too over the last year or two.
DR. STEINWACHS: Let’s try and work through — we have about 15 minutes — to provide some updates. Kevin, surge capacity.
Agenda Item: Surge Capacity: Follow Up
DR. VIGILANTE: We had our testimony some months ago, and since then we have been trying to get the right group of small folks together, or a small group of folks together to talk through what the priorities are around data collection for what we call surge before, but I think we are expanding the notion into the preparedness enterprise in general.
We spoke to Sally Phillips over at HRQ, an epidemiologist there, and Janet Schiller over at ASTO, the Secretary for Preparedness and Response. We have a number of other folks on our list to talk to, to get their thoughts on what the federal government is collecting related to preparedness data.
DR. BREEN: That was in the paper, in the Post the other day.
DR. VIGILANTE: Then we can stop. What are they collecting, who is collecting it, where the gaps are and where the redundancies are. Interestingly, Sally Phillips seemed to feel very certain that Craig Vanderwagen did that through ASPE a year or two ago. I don’t know if that is true.
PARTICIPANT: Not ASPE, emergency preparedness.
DR. VIGILANTE: No, but ASPE that — Helga Rippin when she was there helped do that.
This may be misinformation, I don’t know.
DR. O’GRADY: What did you say Helga did?
DR. VIGILANTE: That the Secretary had mandated that HHS inventory the spectrum of preparedness data, that is, collect from various sources, whether it is from hospitals or CDC or HRSA or others, AHRQ, what are we collecting, what are the gaps, where are the overlaps, and to try to harmonize the whole process.
So we have been doing this in a retail fashion, going from agency to agency, AHRQ, CDC. It sounds as if it has been done already, and if it has been done that would be great, because we would like to get our hands on it.
DR. O’GRADY: Does this sound familiar at all to you, Dale? I can think of an interpretation of what he might be —
DR. HITCHCOCK: Yes, I can, too.
DR. O’GRADY: During Katrina.
DR. HITCHCOCK: Yes, exactly. Helga actually worked for me in her last few months there. That is why she quit, probably. But if she did something we probably could find it.
DR. O’GRADY: What went on at that point, to give you the quick story, Katrina hit, so it is a meeting all night, command center kind of — so anyway, what happened in the Department was, most of the preparedness guys are ex-military, so much data just comes in with almost no analysis.
So Deputy Secretary Alex Hafer at that point said — I thought ASPE was just going to sit and watch during Katrina. We don’t drive trucks down there or anything like that. What they asked us to do was to set up — they asked for an analytic product that would take all this information and do what ASPE does all the time, distill it down, give them the bullet points, say what the main things are, and do a fast turnaround.
A number of people did it. Certainly Jim Scanlon did a ton of it. Helga stepped in when I was afraid Jim was going to burn out and die on the floor. So you had all this stuff, and it was quite clear. You saw a massive difference in the ability to have that real time data back. Agriculture was the best. Right now they can tell you who used food stamps at three p.m.; it is very real time.
But if you are HRSA and you are giving grants to community health centers, your ability to know which ones are open, what is their bed capacity, do they have electricity, do they need the air drop pack from CDC with the drugs and the bandages. We were just trying to get people on their cell phones.
So you saw all this information come in by 5 o’clock in the afternoon, and ASPE would work through the evening. Jim and I would send it to the Secretary at 9 o’clock at night or so. So we would have two or three pages that say, here are the New Orleans hospitals open, here are the ones that are down, here is what is going on, CDC is nervous about mosquitoes and alligators, whatever it happened to be. Then he would get about a five-pager behind that to give a little more detail, and then he would get the big report from Lieutenant So-and-So in Biloxi saying X.
It was all electronic, so he could filter appropriately. If he wanted more, he could go down to the field report. That is what ASPE did during the two or three weeks when everybody was in high gear around the clock kind of stuff. But it did mean that we certainly got very strong impressions about who could give you strong data fast, who could give you crappy data fast, and everything in between. It did nudge them pretty hard. I think some of the agencies ended up with — maybe we creeped into some of their parts scores where they are being judged in terms of their ability, can you tell us what you have actually done with the taxpayers’ dollars.
DR. VIGILANTE: That actually may have value for us to review, because I think it does start to map at one level what kinds of data people do produce and how promptly they can produce it.
It doesn’t get to the whole question. The larger question is, between HRSA and AHRQ and ASPR and CDC and DHS, what are we asking states and hospital systems to report on, how burdensome is that, and how much redundancy is there, and should we be looking at strategies to harmonize this so that we reduce the burden, get better data, get it faster.
It is kind of what is going on in the quality world. If you look at the quality world, it is a real push to harmonize quality measures, because you have American College of Cardiology saying this should be the standard, and you have other people saying that should be the standard, and you have different people asking for quality measures from different — frankly there is just too much request for data going on, and there is a real move to harmonize what we are asking for.
Then there is the sense that in the preparedness environment, there is a similar problem space.
DR. O’GRADY: I think that is right. Jim Scanlon, by the time we were done, he had a Ph.D in worrying about those problems, so he is definitely the right guy.
The question is going to be, it is one thing, did everybody to go back to business as usual when it was over, or did they actually do something about this. The one thing that became quite clear was the ugly divide in HHS between the human services side and the health side. So you would be in a place like Southern Mississippi. It turned out the Head Start shop was open. They had electricity, and with the medical side talking to the welfare side of HHS to say, yes, the clinic is down but we have a Head Start center in the same town, just send the trucks there. So some of that kind of stuff, of being able to reach across those different — those guys just don’t work together very much, guys working with foster children and Head Start don’t talk to the Surgeon General very often. So it is that kind of breakout.
I don’t know if Admiral Vanderwagen told you about it, but now there is all this modeling stuff, of how you pre-position in Jackson, Mississippi, in the junior high and what needs to be there, and what do you think — it is almost like a contagion sort of thing; something happens here, you should expect 10,000 people in this town, 8,000 here.
DR. STEINWACHS: Distance modeling.
DR. O’GRADY: Yes, that kind of where they will go if.
DR. STEINWACHS: Let me ask the next question. Next steps.
DR. VIGILANTE: I think we should interview Jim Scanlon Then we have a number of other people on our list to go meet to get their view of the terrain to try and define what these data sets might be and how we can get an inventory. That is really the first step, is to develop an inventory of the kinds of data that is being collected at a high level.
DR. STEINWACHS: In terms of the inventory, I don’t know if that is fairly intensive or not. I’m trying to get a sense. I know you have Doug working day and night on this, but do we have the resources to do that?
DR. VIGILANTE: I think we have to make a decision as to what is out there, what can we collect, and then decide whether we have the resources to do it.
So when I say inventory, I don’t mean that we are going to actually go out go over different questionnaires and data sets that people have and harmonize them. I am saying, what is it at the high level that is being collected, and what would be the barriers to collecting them.
DR. STEINWACHS: You think the next time we get together, if it is November?
DR. VIGILANTE: That would be our target.
DR. STEINWACHS: Doug, anything you want to add, or say that you don’t know Kevin at all?
DR. BOENNING: We made progress in the last few months. The data around surge capacity and preparedness is really a change in landscape. The meeting at AHRQ was really mind expanding. To learn that there are people at other agencies that we don’t routinely think about, the Department of Transportation, the Department of Education, that have similar interests.
One of the things that I heard at the end of that meeting is, maybe there is enough multiple agency interest that we should form some special interest group informally. I think this summer there was a second national congress on health system readiness for pandemic influenza, but it was really more about health system readiness in general.
The best speakers in my opinion were the ones who admitted that metrics were a laudable goal, but we didn’t in many respects know what preparedness was, and that we first needed to define it before getting to the metrics, and then getting good metrics was not an instantaneous process, either. You needed that before you could do any reasonable evaluation.
So I think this is a problem that is going to be with us for a long time. This is not going to be, write a report and give it to the Secretary and we are done. I think this will be an ongoing concern of the subcommittee.
DR. STEUERLE: I would suggest again that the more narrow issue is what particular role can this subcommittee play. It sounds like there is a lot more we need to know and learn, because HHS has a lot more things it is going to want to do and discover what to do over time. The question is, what value added does this subcommittee provide that won’t be done necessarily internally.
Again, maybe a convening conference, which we seem to be very good at, but I’m not sure. I think that is part of what —
DR. VIGILANTE: It may actually be fairly limited in scope around making recommendations regarding HHS and probably DHS seeking across agencies with the kinds of data they are collecting, the ways in which they are collecting it, to both decrease the burden of collection, also to maximize the utility of it, to identify the gaps. Not that we would do that work, but I think that it is becoming clearer that that is an issue.
DR. STEUERLE: My philosophy is, consultants going to an agency and finding out what people inside the agency generally already know, but then bring that to the level where some decisions are made on it. So I think what you are partly saying is that maybe we can serve the role of emphasizing to the Secretary that something needs to be done.
DR. VIGILANTE: We just need a little more concrete evidence base to base those assertions on.
DR. O’GRADY: There was a concrete suggestion made by Janet Schiller at ASPER. That was to look at OMB rules around surveying. It is not adequate anymore to wait nine months to get approval to do a survey about preparedness. We need to be a little more nimble. She thought maybe the contribution of this committee would be to have special dispensation for urgent assessments relative to preparedness.
DR. HITCHCOCK: That pretty much already exists. We just have to know which switch to switch on and off.
DR. O’GRADY: Right, or her Assistant Secretary has to call OMB.
DR. HITCHCOCK: They can play around with notices, or do what CDC does in Atlanta, which is to get the clearance that will allow them to go out, if there is a food outbreak or a poison outbreak or something. They will go out and run a survey and then inform OMB afterwards as to what they did.
DR. VIGILANTE: I can tell you that in a survey that we did for OMB, probably the largest preparedness survey ever done, 29 hospitals, they wouldn’t let us use the data because it was more of a convenience sample.
But I would submit to you that the response rate in a certain substrate of hospitals of a certain size would reach levels of statistical significance that would be publishable in any peer reviewed journal. The fact that we can’t use that is kind of disturbing.
DR. HITCHCOCK: I have dealt with OMB for God knows, probably 30 years now. I know their ways of getting things done. I wouldn’t waste time when dealing with OMB.
DR. STEINWACHS: Okay, on to the next item.
MS. JACKSON: Before we do, I would just follow up on Gene was saying about trying to get something as concrete as possible even for the next meeting. We got some great information we got from the surge capacity section of our materials. We don’t want that to get cold.
I know things are progressing and changing and the landscape continues to change, but considering what you do in other agencies, it would be great if we can get something in, a statement of what needs to be done or just the investments to make.
Agenda Item: Data Linkages to Improve Health Outcomes — Cross Cutting Followup Issues
DR. STEINWACHS: The third item was data linkages. We did an excellent workshop on data linkages. We produced one letter to the Secretary. The question was, is there more to do on this that is worth doing.
Joan Turek had wanted to be here, but she lost a crown. She went to the dentist. Then she was heading home, then she had her garden club this evening or something. It was complicated.
I did ask Joan from her point of view what suggestions she had for the committee. One idea was to convene a group and meet with the centers that make data available for NCHS and I guess those are mostly also consensus centers, except for the one, about what they are doing to try and make access easier, what they are trying to do to improve access, would be one next step.
In the letter we sent, we identified the access concerns. We never got down to potentially the next level of detail about what are concrete things that might be done.
DR. HITCHCOCK: We said we were going to monitor them, in the very last sentence of the Secretary’s letter, as I recall.
DR. BREEN: Do we have response from the Secretary?
DR. STEINWACHS: Not yet.
DR. STEUERLE: It is one thing when you raise the issue, but somehow or another eventually if something is not being done, you have to change slightly the incentives within the system for doing something about it. In some cases it is just having somebody responsible, like creating a national coordinator for information or whatever.
Here, it still seems to me that the ASPE people, for whom this just resonates, I’m sure, because in their hearts half of them are researchers, and they understand the need for better research. But I don’t sense that there is anybody in the system whose grade for what he or she is succeeding at doing is related to this. So ideal merges and opportunities go by the wayside.
DR. VIGILANTE: An information sharing czar.
DR. STEUERLE: I don’t want a czar. I keep struggling with how to actually affect the incentives with this system. I think most of us feel, and certainly the hearing supported this, that there is an enormous amount of information out there. It is paid for, in the sense that we have paid 90 percent of the cost of getting the information, but then you don’t have it available in usable form for researchers to use or whatnot. So there is always this dilemma.
A lot of it ends up to be very narrow targets, like Kevin has got a research project, he knows what he can if he has that data set, which is down at that level. But there is nobody around to try to make sure he can do it, even if it is an internal or external researcher. So I still struggle with this issue of, is there some other organization or aspect or incentive system that should enhance our ability to make some of these things happen.
DR. GREEN: What is your thinking about how you would do that?
DR. VIGILANTE: Can I just say one thing? I was sort of joking when I said czar, but it is the same problem in the intelligence community. They have this huge issue both within the CIA but across CIA. That was the concept of creating the DNI, is to make one person responsible for information. Obviously that is at a much higher level, and we don’t have anything like that in the Academy.
But the idea that there is somebody who is both the steward of in the sense of being responsible, but a steward of the taxpayer, in the sense that we are trying to get the maximum benefit for your investment in collecting this data by seeing that the data we have is responsibly shared or linked to produce the greatest benefit.
So I do think this concept of somebody whose responsibility it is to do that is not a bad concept.
DR. STEUERLE: And also, I don’t know if it is somebody having a fulltime job, but maybe somebody who is part of the Department who is responsible for producing a report once a year on progress being made in this area. If you have got to do a report, usually to get the report done you start to make decisions so the report looks good.
DR. VIGILANTE: Is there something from the GAO or OMB perspective, that the cost of not doing this — in other words, from an ROI perspective, what are we losing in terms of knowledge?
DR. HITCHCOCK: Actually we had a process we submitted as part of our research plan to evaluate the data centers. It was very helpful. I think we could probably put it in again next year.
DR. O’GRADY: When you think about that, Gene, when you think about what you said before about how you get the incentives and how you get this in, I don’t think the czar is the way to do it. You will see guys in HHS who have great titles, and if there is not the juice behind it, nothing happens, or they lose their seat at the table when the real decision gets made, or there are a number of things there.
The time when you think about when data really has to come out and where the momentum is behind it, things like the poverty rates. It is in legislation. We are not going to know eligibility for next year’s — that is where you get it.
So when you think about linkages, to a certain degree I think there is probably a development of a constituency for it that is in the actual day to day policy making. So when the CBO cost estimates require linked data sets, to be able to give those committees the runs they need to tell them, is this drug benefit going to be $400 million versus $500 million, I don’t want to be the guy that gets in the way of getting that linked data set, because everybody is waiting for it. It drives policy. Then you have Congress giving you a good return on investment, and they need that sort of stuff. However you want to think about how that feeds in, whether it is a CBO or a GAO or whatever, but when they feel that they need those linked data sets to do their job, that is how you get the juice behind it on an ongoing basis.
DR. STEUERLE: We do have people like at Census. Nancy Gardner has got a special office whose job it is to think about how to create these data centers to get access to data sets and do all these other things. It is built into the structure.
Whether it is entirely working fully, I don’t know, but I don’t know that there is such a thing in HHS. Is there somebody at HHS who —
DR. O’GRADY: HHS is organized within different organizations. Like, for the Medicare and Medicaid data, certainly they have their own surveys, they do their own things, and they have their own data user centers. Their model has been to farm it out to academics. So there is Chaddick, which is out at the University of Minnesota. It is a big contract. You call up and you say, I’d like to do some work in Medicare. They are supposed to be out there to say, well, we have this data set, but if you went back to 2002 you could also pick up this, this and this, and show you the ropes. You still have to have your data use agreements, you have to cross your T’s and dot your I’s, but somebody to help you along with some of that stuff.
So you will see these different things and different parts, depending on the data. But it is the privacy concerns and things like that. But I don’t know anybody other than — but that is a perfect world also for the HHS Data Council.
DR. STEUERLE: I did a study on — we started off with the Population Committee; when I first got on it we were very interested in the distribution of health by socioeconomic. I did a study on the distribution of social security benefits. I realized if I could link the social security records to the Medicare records, I could do the same thing for Medicare.
You had a former head of CMS who did a study. He didn’t have the real data to do it, so — he could use real data for this. It is an obvious example of something that could be done. I know other researchers who got their own sets of linkages, Nancy is an example, where, if I could get this data I could do this really useful study. Nobody is responsible for figuring out how we who have these ideas can get it done. Maybe you wouldn’t go with an outsider, maybe you would have to do it in social security or Medicare or something.
I just get the sense that there are a lot of these opportunities out there. It is not just a question of waiting for a CBO to demand it, but it is again one of these classic cases. We all see the opportunities or see particular opportunities, but we don’t know how to actually —
DR. O’GRADY: That probably is the Data Council to get it started. Then once you get people using it, then you have got this ongoing set of incentives to continue to do it.
DR. STEUERLE: But it is not happening.
DR. BREEN: I think you are right, there is a lot of demand out there, but that would be a whole research project unto itself in order to find that out and get the impetus going for demand for all of these kinds of data linkages.
But you talked about incentivizing, and then Mike talked about prioritizing. Both of those would need to go on. Legislation might be a good idea. It is not something we could necessarily do, but we could recommend that — or suggest ways in which the estimates would be better if they used linked data. Maybe that is the way to go, to go through some cases where we know it would be better, and just describe that.
DR. O’GRADY: We do have a track record. Saying that researchers need it is not a winner when you go to HHS budget.
DR. BREEN: Well, not researchers need it, but that the public needs it for some of these estimates that are required to be provided by legislation.
DR. SONDIK: This may be a pipe dream, but I have felt if there was more — if the research in the Department were more observable, if you will, if you view the Department like a big circle, then all around the edge of the circle there is lots of terrific stuff going on, whether it is in NCI or NCHS or wherever.
But what it doesn’t seem to have is anything really significant in the center of the department. There is all the specific issues that ASPE deals with, but it would be terrific, I have always felt, to have an analytic activity that was not completely driven by, these are the very specific things that you have to work on, but where people could get together across the Department and share problems. I don’t mean just get together and do it, but actually work on problems together across the Department.
There is not a lot of incentive to do that now. It happens every once in awhile. You get a Nancy Breen who gets out there with real motive and says we need to do this kind of thing. But there aren’t very many Nancy’s in the Department. There is not a lot of reward for that, which is kind of what you were getting at, but there could be.
One place that is really interesting where there is an emphasis on the data sources but not so much the deeper analysis is in Healthy People. Healthy People is kind of anal compulsive in the documentation of all of the data sources that go into these 637 objectives, whatever it is, 438 or whatever it is, but then it gets multiplied. Every one of those has — the data source is described in a separate volume.
The program has actually been I think reasonably successful, meaning that it actually has worked in many cases, to get new data sources established. The problem is that I don’t think there is much analysis of the information that is in Healthy People. It is around the edge of the Department, again.
I would say this even if Michael weren’t sitting over here. He knows what I was going to say. We actually started a group that was focused on insurance or uninsurance or whatever, but looking at the different estimates and that sort of thing. Got people together from across the Department, it was terrific, and outside the Department.
We need an agenda. The group was enthusiastic about doing it. So I think there is great potential, but like everything it requires resources. I think it requires resources for people who will be able to to some degree follow their muse on this.
One of the best things to think about is mathematical modeling of these things. That is what we are really talking about, an analysis. But that gives rise to understanding the need for databases, the need for linkages and how to do it. It would be wonderful to have an analytic agenda.
We have the Secretary’s priorities. It would be great to have the ASPE analytic and research priorities that were put up in front of the Department and the heads of the various agencies say, we realize why that is important, and emphasize that people should be involved in doing it.
We have got an issue right now where they did the DNA and the release of DNA data and how to do it. We got two big pieces in the Department that are going to be trying to work this thing out and try to figure out the best way to do it. It is not exactly the same problem, but it is bringing different pieces of the Department together.
So I think it would be wonderful to see some recommendations related to this. I would certainly endorse more analysis.
DR. STEINWACHS: Do you want to think about some next steps or some ideas? I am looking at my linkage people.
DR. STEUERLE: In the end, it could be the type of thing where it is a recommendation where the staff in the Department feel, this is a recommendation I would like to make happen, not that it is an outsider pushing the burden.
Maybe it is very simple, but maybe just some simple annual report that the Department should make. Maybe it should be related to data that is not being adequately developed to meet high priority health needs in the population, and the report should go to the Secretary. It should include what data sets are available. I’m not quite sure what it is, but some requirement to report our progress.
DR. SONDIK: I think the Data Council could play that kind of role.
DR. STEUERLE: But do we have a role in it?
DR. O’GRADY: You certainly can make the recommendation. One thing about the experience about like the uninsured. When I was on the Hill, I assumed the CBO health economists talk to the CMS actuaries as we headed into that drug benefit. There are like five guys on the Hill and six guys, it is not massive machines of people out there.
It turned out they did a little bit if they really trusted each other, but neither one was willing to start. So when I was on the Finance Committee I invited both — CBO was willing to host it, but you had to invite both teams to a meeting. That was very much the same experience with the uninsured. Census was happy to come. They were scratching their head about these issues. They wanted to talk about skip patterns and reweighting versus imputation and stuff like that. Everybody came in thinking their survey was the best and the other guys all made the mistakes. So there is some convening role here.
There is a new nominee for ASPE. He is an analyst, or at least, I am pretty sure he is a reasonable analyst. So to make that kind of — from a group like this to say, would you take on as a priority, focusing on the potential linkages that could go on within the Department of this wealth of data that the Department collects, I would think at least a fair hearing.
DR. GREEN: Can I ask a clarifying question between you two? Such a recommendation must be made within the constraints of the current budget situation, is that correct?
DR. SONDIK: It depends on what you are recommending. If the recommendation was to get a group to look at, you name the topic, different estimates of from various sources, for example, that doesn’t require any significant outlay of resources. You can argue this is the kind of thinking that people really should be doing. So that doesn’t require resources.
If it is to develop let’s say a linkage project that is going to do something, then it needs to be cranked in. But I don’t think that is a huge allocation of resources. If it is a brand-new survey or something, then that is another story. There were these situations. But a couple of hundred dollars one way or the other out of the ASPE budget was able to allow one team to try their questions on the other guy’s survey.
The methodological work that you would like to go on here is to test some of these out. ASPE paid for the linkage of the Medicaid data with the survey data, with the administrative data. Real candidly, that is what I thought were strategic investments that sitting in that seat I was supposed to be making, how do you leverage what is not a big investment into a big return, because other people are spending millions over here and millions over here, and where can you do that.
So I wouldn’t think that takes it off the table. It is up to whoever my successor is it finally ends up being to think about how they want to spend a limited research budget. But you can make some pretty strategic investments and get a lot of return for them.
DR. BREEN: I think Mike’s point is really well taken as a general point. Then I was going to say that right now, the discussions about Healthy People 2020 are coming up. They are forming committees. Ed, I don’t know if you want to talk about this instead of me or Virginia, but — shall I go ahead?
DR. SONDIK: Sure.
DR. BREEN: Okay. Right now they are convening a bunch of people from the federal government staff to work on a series of committees that they have got, ranging from — I don’t know if people are familiar with Healthy People 2010, but it is a whole list of goals that we should be meeting.
The two major goals — as Ed said, there are a million objectives, there are a lot of different objectives, and a lot of them are condition related, risk factor related, but this time for 2010 the two main goals were to eliminate health disparities and to improve quality of care. So there are often these very big picture things and of course there is a question about how to measure these things.
That is where I am thinking this issue of data linkage could come in. Depending on how the goals are measured or what the goals are, you may need to link data, or you may need to bring data sources together in new and creative ways. That is certainly possible. That would be the kind of incentive or priority that would be there that people would need to respond to. So that is one possibility.
Then another thing which is a huge, huge bang for the buck is mathematical modeling, where you take the data you have, you impute the data you don’t have, and you come up with a lot more suggestive information which is usually pretty accurate and usually an improvement on what you have, and certainly helps identify research gaps or gaps in data, which I think really has to be part of the enterprise.
We don’t just want to move forward saying here are redundancies and we don’t need this and we don’t need that. I don’t think we are looking for cuts right now. I think we are looking to develop and stabilize our data sources and create a situation where we can get more out of it, not less.
I think the description of what is happening with the vital statistics is scary. I didn’t realize that we were at that point. But I would say mathematical modeling and 20 objectives are things that this committee could actually think about.
There are a number of staff people here. I am still in a position where I am trying to figure out which of those committees if any I want to participate in. We are welcome to do that, and we are welcome to get input from outside. Then eventually there will be some town meetings kinds of things, where there will be input welcome from the public. Certainly the more people that have thought about these things, the better able they are to put in meaningful input that has an impact.
DR. STEINWACHS: Why don’t we have some discussions, maybe do a conference call of the committee and talk some more about what to do on data linkages. But I think there have been some good ideas here about how there might be a next step.
DR. BREEN: Can I say one more thing?
DR. STEINWACHS: Please.
DR. BREEN: I’ll be brief about this, because I can talk for a long time, because I am a big supporter.
A project that I work closely with at NCI is the California Health Interview Survey. People from the California Health Interview Survey came and talked a couple of years ago under different auspices, when we were looking at data on ethnicity and language, because it is a leader in that area.
But another thing that the California Health Interview Survey is a leader in is getting local level data, so statewide local level data. The sample size for the THIS — it is telephone survey, it is not in person, so it is cheaper per person, but it is larger than the national health interview survey.
Consequently they are able to link those data to a lot of other geographic data that are available, and those sources are pretty rich in California. So there is an enormous amount that can be done at the local level that we are not able to do at the national level.
I think looking for those kinds of models and revealing those kinds of models and exploring those kinds of models — we only really have one, the California Health Interview Survey, but there is a cancer registry throughout the state, there is the California Health Interview Survey throughout the state, there is excellent mortality data that has occupation and a number of other aspects associated with it, and birth certificates. They are still collecting divorce and marriage and those kinds of information. They have got hospital discharge information, they have got air, water and soil pollution information, which is a little spotty, but it is there. The more these data are used, the better they get, because people want them.
The other thing that has happened with CHIS, and this will be my last point, is that by putting a lot of money into dissemination and by making sure that they have got data reported at meaningful geographic levels like voting districts or counties or where state health departments are working and making decisions, or smaller areas than that in large cities, these data now are meaningful to these policy makers, to advocacy groups, to public health officials. They clamor for it. I think they would object if those data started to dry up on them.
So they have created a need for these data. It is a real model I think for the way data can be collected, disseminated, used and improved in a pretty agile way. So that is something we also might want to explore if we do have a hearing on this. We are just exploring ideas at this point, but I wanted to bring it up.
DR. SONDIK: Can I add something?
DR. STEINWACHS: Yes, please.
DR. SONDIK: I complete endorse what Nancy said, I think it is very important. Another aspect of the local data has been the idea to use the standards that are in the national health and nutrition examination survey on a local basis.
We have been talking about this for a long time, but of course it takes money to do it, and we don’t have the resources to do it. But there has been quite a bit of interest in this as of late. New York City did it. New York City is thinking about doing it again, actually. It would be once every five years or so if they do that.
Wisconsin is gearing up to do it and actually it may already be in the field. California interestingly enough also has a real interest in this. That is exactly the right kind of coupling, so you get the health interview survey. It is in a sense what we are doing at the national level. You get the health interview survey which asks lots more questions, and then you get a smaller sample that is a hard clinical sample from the NHANES.
Did I mention Washington?
DR. BREEN: No, you didn’t. You said Wisconsin.
DR. SONDIK: The State of Washington in the Seattle area and Canada actually is doing this. This is across Canada, but they are taking the same standards and approach, and they are watching their own NHANES. They call it something different, but it is the same kind of animal.
You wouldn’t do this every year. You would do this periodically. It really becomes a benchmark. I think the point you are making, I think it would be a wonderful thing for the committee to focus on. It is how you bring all these various resources together. You get the biostatistics as the foundation, then how do you build on that and then link up to the national level.
Years ago, everybody knows, everybody talked about this idea of the integrated data set, where you drilled down and so forth. Now we have the possibility as it gets cheaper to get information, although it may be more difficult to get information, we have the possibility to do that kind of thing. We certainly have the possibility to blend, meld these data sets together like we never did before. So I would certainly endorse that idea.
DR. HITCHCOCK: When NCHS was first organized, it would get the data from three different sources, the people that made the measurements and the hospitals and providers, and somehow merged them together.
DR. STEINWACHS: Let’s touch on one other area. Larry and I were talking some. Do you want to say a little bit about the medical home and what we were talking about?
DR. GREEN: I’ll try. My mind is roiling with the ideas on the table.
DR. STEINWACHS: I thought the purpose was to get all the ideas out, and then what we will do is have a conference call as soon as we can schedule one. Then I think we need to choose, but we can do a little background work too, because these also need staff support.
DR. GREEN: Let me make a go at trying to insert what I am about to say into the context of the discussion that we have already had, as opposed to a separate and discreet set of ideas.
I am not a public health person. I am just a concerned citizen physician sort of guy. From where I sit, I am most impressed by squandered opportunity to make a difference, because of reports when some important changes occur.
I am going to impose on you just for a minute a frame of reference that says the vital statistics system that we have and the public health system that we have, the mental health system that we have, the health care delivery system that we have, most of these were born in the 1800s. And we are working with in my view mature manifestations of the ideas and the methods that were available at that point in time before things like robust mathematical modeling was possible, or before a personal health record was feasible and trackable and usable, and before an EHR that was integrated, capable of cutting across home care through inpatient care to nursing home care to the school system, with nutritional information that if we wanted it to be there could be there.
So it just seems to me that we are poised not for a tweaking of these old systems that we have used for mental health, public health and personal health services, but that what is really trying to happen here is a transformative change that is largely enabled by the information revolution. We are in the richest country in history with two trillion dollars to spend, that is the way I look at it, at a moment where it is very propitious that major, major change can occur.
What kind of statistics make sense in that framework? I think using birth certificate data is nothing short of stupid. It is ridiculous. But connecting birth certificates through life to death certificates is totally — it should just be a given.
So in that context I’ll get serious here. It is interesting that virtually every Presidential wannabe in 2007, on the website and in their interviews, is talking about this idea of a medical home. These are two words that are not terribly well defined, but they are attributes already established and agreed to by the largest platform of health care delivery in the country. There is something about it that is acceptable to the politicians to where they can talk about this.
All the medical home is, is robust primary care with the right data linkages. I intuitively believe that there is a serious opportunity to be seized by getting ourselves organized to make the measures that will let us know whether or not people have one of these things, and using it as a critical point of integration.
The way the medical home — this idea has been percolating for a couple of decades, and it really got a shot in the arm for research from 1999 to 2002, but it is now getting political attention for the first time. That is the reason I am talking about it today. It might be so, if there could be progress on this, the statistics necessary to integrate care and know whether or not we got any results that mattered and are available.
This might be the sort of thing where the committee could use its convening function to begin to understand if there can be an accountable location for the integration of the care of the patient, why can’t there be a location for statistical integration of an individual, whether they be a patient or not.
To drag another thing onto the table, when you look at the NIH Road Map and you look at the maturation of community-based participatory research, you look at the successes and failures of 2010 and the promise of 2020, CPPR, geospatial analysis, the medical home, the AHR, the huge interest in getting to something that is sustainable, the turning of attention on managing and reducing risk instead of diseases. Managing risk instead of diseases is going to be here, that is the future, that is where we go if we want to get the value for health care services if we don’t capture risk in a good enough way.
If we could do something to stimulate consideration of what it will take to execute that and then try to think through what are the statistics that will be required to know whether it happens or not, and to look for the associations of those sorts of things, then I think some of the big policy issues, some of the really big compelling questions about how do you want to spend the next three trillion dollars may take on a new life.
I am disinclined to think that this committee or anyone else should try to resurrect the way we have been approaching these problems. I really think there has to be some refreshment of staying faithful to the values and the principles of population health as a measurement that we have got new toys on the table, and that we ought to grab them and see if we can’t use them to advantage.
I’m interested in being in this group because it seems to me that people like you who know so much about this, you ought to be able to figure this out.
DR. STEINWACHS: Talk about privacy just for a few minutes.
DR. GREEN: I don’t think I can stand talking about privacy. Well, okay, I will talk about it. What frustrated me all afternoon is that we sat there and allowed ourselves to pretend that privacy exists.
We just moved to Health Sciences Centers in Colorado. They know every time I enter the parking lot and they know every time I enter the building, they know the time, and they know it was either me or someone who stole my ID card. They also photograph me at every damn door, and the lights turn on when I walk into the building. Tell me that there is such a thing as privacy. Why are we trying to protect something that we know is already gone?
So I thought that the fruitful part of this afternoon was when we turned on the stewardship. I want to say, given that privacy is not possible, we better get serious about stewardship, right?
So again, forgive me for being sort of reckless in this statement, but I am unprepared to surrender to the constraints of policy issues that are not really assailable. I think that we should try to do better. These data linkage things are so potent. You have got a true believer here about the potency of data linkage. But data linkage by itself isn’t enough, either. You have got to organize this around where fresh opportunity lies.
I want to get into the discussion for the subcommittee the possibility that we might be looking at creating a reporting mechanism that monitors the evolution and establishment of an integrating focus for health care, if that could exist, and when it does exist, what are the population health implications of that and what are the tools we have that we never had before that you want to build into it now, because when those EHR standards are in conflict then you are looking at a ten-year exercise to try to rotate it.
There are going to be 80 million people using that sucker in 12 months, is what I think. I think that there should be some mechanism to take advantage of that, to make some estimates of what is going on with the health of the country. If you have got 80 million people using the same device, for Christ’s sake, can’t we do something useful with that from a population perspective, as opposed to saying, we need more money to do another survey. Although, don’t mis-hear me, I am prepared to go to war if we want to do another survey.
Forgive me for the diatribe, but I really would like for the Population Subcommittee to have a chance to try to brainstorm about some of this stuff, and bring you guys into that and tell us if you think it is stupid.
DR. BREEN: I like the idea. We tend to talk about data, and a lot of details. I think you tried to frame it a little more broadly, and particular focus on medical care and how we could look at continuity of care, which is a serious issue in cancer control, so continuity of care has loomed so large in that area.
Our medical system and medical records system are not equipped to take care of that well.
DR. STEINWACHS: From the public health point of view, we rely on personal medicine to provide so many of the services now in terms of risk assessment in addressing those. You could make the argument that everyone needs a medical home, healthy as well as ill, and therefore you could look at it as a public health strategy.
DR. GREEN: In surge capacity just for a second, if everyone had had a medical home in Katrina, we would have been relying on Walgreen’s to try to take care of the people in Houston and Denver and the places that they were, because we would have organized the statistics, the database for that. There would have been an organizing focus for that.
If you can do surge capacity off of that, I’m sure we can also do public health stuff off of that. That platform is trying to help. I wish we could figure out some way to take advantage of it.
DR. SONDIK: Can I just respond very briefly to that, and then put something else on the table?
DR. STEINWACHS: Sure.
DR. SONDIK: The committee was instrumental in coming up with a report about the health statistics system for the 21st century. I think you have done some discussion of revisiting this, but I think revisiting it with the kind of scope that you are giving this could be a very good thing to do, because when that was done, there wasn’t that much emphasis on the availability of electronic records. It is conceivable that that could help realize at least a portion of the vision that was spelled out in that.
The key in this vision was linking a variety of different pieces of information together that no one source gives you everything that you need, but it is the bringing in a variety of things, the kind of thing Nancy was laying out before, that you are really interested in at the local level, whether it is the geography, the location of services, the context in some sense. All of that goes into this. You would like to go into the analysis. This vision was about how to head toward bringing all of that together. I think that fits with what you were saying.
The thing I wanted to put on the table, and I gather you were talking a lot today about privacy and confidentiality, was, — and I just want to alert you to this. I did this with the privacy committee upstairs. Working actively to find ways to use the DNA data that is in NHANES could be tremendously valuable because NHANES has not only the DNA information, but it has got hard clinical data, it has got behavioral data and it is a representative sample of the U.S., and it has got these clinical measurements. So you put them all together, perhaps this is sort of a Rosetta Stone toward helping put all of this information together.
But the issue is, how do we do this and preserve the confidentiality of the respondents? The more information you put out, the more vulnerable they are to be identified.
What has come up recently is that there are two different approaches to this. The NIH has just started a program it calls genome-wide association studies. The idea behind that is to take studies that are population based or something approximating a population base and DNA and other information, and make those available to researchers around the country, adding safeguards to it.
Now, we have been attempting to do this, and actually have done it to a small degree with the DNA in NHANES. But the approach that we take is that we put the responsibility in terms of being sure that we don’t disclose information that is identifiable. We can put that responsibility on us.
In the NIH approach, it is placed on the investigators, and to a degree on the data bank, but it is mainly on the investigators and their institutions. These are quite different approaches. We worry about investigators inappropriately sharing the information, inappropriately trying to identify directly or indirectly, however it might happen, individuals, and what might happen to those individuals if in fact this information becomes known, maybe that it might get into the hands of people, insurance companies, or whatever it might be. You think about all kinds of terrible things that might happen.
So in the federal statistical system, we think of these things in terms of, it is the responsibility of the system to live up to the law — in our case there is a law — and the pledge that we make to the individuals to maintain their confidentiality.
In research, it is a bit the other day, where the investigators pledge that they will maintain that. So we are heading toward probably a meeting I thought was going to be in December with NIH. It is probably going to be in February or March now. I think it is important if I may say so for the committee, the overall committee, to be aware of this, and perhaps to participate, have members participate.
It is a really significant issue. The fact is, we just don’t know the best way to do this. I think we have to feel our way along.
We are very concerned, last point, that if there is a breach in confidentiality, it will hurt specifically NCHS in our ability to collect this information in the future. I’m sure NIH feels the same way, but perhaps our sense of this is more acute than NIH. I don’t want to put words in their mouth, and I shouldn’t be the one to speak for NIH. But at this meeting, if it happens the way I think it will, we will lay these things out and try to address how best to meet this. We probably can’t do it in one meeting, but there might be a series of these meetings.
I just wanted you all to be aware of that.
DR. STEINWACHS: We appreciate that.
DR. GREEN: Are you going into that meeting with the assumption that personal DNA can be de-identified?
DR. SONDIK: Well, we had a meeting in the Department recently. It was actually a terrific meeting. NIH was here, we were here, and the Department with personalized health care and so forth was also here. We were all discussing this. There wasn’t an issue on the table.
Somebody used the term de-identify and I said, whenever I hear the term, it bothers me. It gives me the willies, to use an old term. The reason is because nothing can be de-identified. It can’t be made with probability of zero. I think it is important that we know that and we state that and we get that really up front. It is really a matter of risk and what level of risk we are willing to — and when I talk to our experts, I want a level of risk. I want you to give me a probability. They throw up their hands, they can’t do probabilities with these things.
But I think that is one aspect of this. I think we need to try to do that. We need to be able to somehow get a better handle on what the risk is and what level of risk we are willing to accept.
Everybody agrees. I couldn’t be more in favor of this. We want to get the data out, we want it to get used, we want the most we can get out of it. We do not want to put it in a mayonnaise jar on Funk & Wagnall’s doorstep, as Johnny Carson used to say, and just leave it there unopened.
DR. STEUERLE: I just want to make one concluding remark. I think to some extent this subcommittee is still in part struggling with how to take each of these agenda items and add value added as opposed to reporting on them.
I will mention that this is actually my next to last meeting.
DR. STEINWACHS: Oh, no.
DR. STEUERLE: I think about two to three years ago, I’m not sure when, Don and I went into a little bit of a collusion, where we started to put the vital health statistics back into the National Committee on Vital and Health Statistics. It is not that I am opposed — my personal view is, it is not that I am opposed to spending half our time discussing things like whether we can measure privacy or not, and how many complaints come in and how many we pass through, and a lot of things we discussed at that meeting, but I think all of us feel that there is so much power in these data that if we can add more weight to them in the hearings, it integrates well with the whole emphasis on electronic health records.
So I think part of what you are seeing playing out here is how we find our role. I think we have found several niches. I think we have made some real advances, but I really encourage those of you who are new members of the subcommittee to take that ball and run with it. I’m not so sure we have gotten quite over the top of the hill, but like Larry who got so excited about some of this stuff, I really think the possibilities are extraordinary. I just want to say, I feel very good about this meeting. I think we have come a long way from where we were, thanks to you, Don, mainly.
DR. STEINWACHS: I don’t think so.
DR. GREEN: So you are going to have a phrase that says, put statistics back in vital statistics, and I’ll have a phrase, put care back into health care.
Agenda Item: Summarize Discussion and Determine Next Steps
DR. STEINWACHS: Next steps, do two things. One is, I will try and summarize briefly some of the ideas we have put on the table, and then get it out so you can correct my summarization, which I’m sure will be wrong. So that will form the agenda then of a conference call of the committee. Also during that time, some of these things will require staff support when we try and move them, and so I will try and have some conversation with them.
Thank you all.
(Whereupon, the meeting was adjourned at 6:25 p.m.)