[This Transcript is Unedited]

Department of Health and Human Services

National Committee on Vital and Health Statistics

Subcommittee on Population Health
Subcommittee on Privacy, Confidentiality and Security

February 9, 2011

St. Regis Hotel
923 16th Street, N.W.
Washington, D.C. 20006

Proceedings by:
CASET Associates, Ltd.
Fairfax, Virginia 22030
(703) 266 8402


DR. FRANCIS: The joint Subcommittee meeting of Populations and Privacy, Confidentiality and Security and we are here to start following up on the wonderful day yesterday. But before we do that, what we need to do is to go around the table in state of spirit and mention whether there are any conflicts. I will just start out by saying I am Leslie Francis from the University of Utah and I do not have any conflicts.

DR. OVERHAGE: Marc Overage, Regenstrief Institute no conflicts.

MS. BERNSTEIN: I am Maya Bernstein. I work in the Office of the Assistant Secretary for Planning and Evaluation. I am the lead staff to the Privacy, Confidentiality and Security Subcommittee.

DR. SUAREZ: I am Walter Suarez with Kaiser Permanente and a member of the committee, a member of the Subcommittee on Populations as well as a member of the Subcommittee on Privacy and Security and I do not have any conflict.

DR. EDWARDS: Brenda Edwards, liaison from NIH working cancer surveillance research.

MS. JAMISON: Missy Jamison NCVHS staff.

DR. McANDREW: Sue McAndrew, privacy liaison from the Office for Civil Rights.

DR. MAYS: Vicky Mays, University of California, Los Angeles. I am a visitor today but will be joining NCVHS if all goes through.

DR. FRANCIS: Let me say this is your first unofficial, being here and welcome.

DR. HORNBROOK: Mark Hornbrook, Kaiser Permanente, member of the Subcommittee.

DR. TANG: Paul Tang, Palo Alto Medical Foundation, member of the Privacy Subcommittee.

DR. LAND: Garland Land, member of the Population Committee.

MS. MILAM: Sallie Milam, West Virginia Health Care Authority, no conflict, member of the Population and the Privacy Subcommittees.

DR. GREEN: Larry Green, University of Colorado, member of the committee and Population Subcommittee, no conflicts.

MS. JACKSON: Debbie Jackson, National Center for Health Statistics, committee staff.

MS. GREENBERG: Marjorie Greenberg, National Center for Health Statistics, NVC and executive secretary for the committee.

MS. KANAAN: Susan Kanaan, I am writer for the committee.

DR. FRANCIS: And I believe there may be people on the phone and I know there are people in the audience. Let’s go to the phones first. Anyone on the phone?

DR. BREEN: Hello, Nancy Breen. I am on the phone. I am sorry I could not make it in person but I am happy to be here with you so please speak loudly so I can hear you.

MS. HORLICK: This is Gail Horlick, CDC and I am staff to the Subcommittee on Privacy, Confidentiality and Security.

DR. KHAN: This is Hetty Khan, NCHS staff to the Subcommittee on Privacy, Confidentiality and Security.

MS. CHAPPER: Amy Chapper, CMS, office staff to Privacy, Confidentiality and Security Subcommittee.

DR. PAISANO: Edna Paisano, Indian Health Service, staff to the Subcommittee on Population.

(Intros around room)

DR. FRANCIS: I think we have a couple of goals for this meeting.

MS. JACKSON: Just a quick clarification of the timing and use of the room, this is a combined subcommittee discussion from 2:30 until 4:00. The original agenda you might have in front of you is in error. At 4:00 then the Quality Subcommittee will be coming in to work on their letter and other follow up issues. There is just so much overlap between Populations and Quality a lot of us will be at the table but for those on the phone and in the audience I want to make it clear with that line of demarcation is at 4:00. Thank you.

DR. FRANCIS: So we heard a summary from Susan of what happened yesterday. As I see it we have three goals for the next little while and maybe we can divide up the time on these three goals. The first one would be more detailed reflection to the extent we wish on what we heard. The second would be next steps whether it is another roundtable with populations or a roundtable where privacy is in the lead considering the question perhaps of best practices or the use of both aggregate and in particular qualitative and smaller area data in public health. And the third would be how we want to be ready for June 9th, being very task oriented. So I will open it to others who were there yesterday who want to chime in with respect to themes for Susan that you want to make sure get underlined that didn’t get underlined in the discussion in full committee.

DR. BREEN: This is Nancy at NCI. Since I wasn’t there in person, did Susan email her remarks or present them earlier or what was the summary that you mentioned?

DR. FRANCIS: Well, Susan had begun sketching out major themes even before yesterday because she had been engaged in lengthy discussions with each of the workshop participants so she had a pretty good idea of the themes they were going to bring up and she sketched out.

DR. BREEN: So are you describing the summary, I mean the sort of outline for the report that she had sent us?

DR. FRANCIS: That was the start.

DR. BREEN: Okay.

DR. FRANCIS: And then she added in more and I think would be willing to send those around in a bit.

MS. KANAAN: She probably has that and actually Debbie is passing out hard copies. I think Nancy probably it is what I sent out.

DR. BREEN: Okay great, thank you.

MS. KANAAN: I did add a few but I think they will emerge as we talk today.

DR. BREEN: Thanks for clarifying.

DR. FRANCIS: You are welcome. Sorry about that. So maybe the question would be next steps and what we think we need to hear next. That should be both I think from the population side and from the privacy confidentiality side. Go ahead, Sallie.

MS. MILAM: Just thinking about some of the opportunities we have in front of us, I like the idea of the primmer. I also like the idea of gathering more information around the issue that Mark identified during our discussion during the full committee meeting around best practices for de-identification and the scenario not governed by HIPAA. How do you apply statistical disclosure limitation techniques given your data and particularly given the small area of data analysis that these communities are engaging in.

MS. GREENBERG: What we need to agree on I think today is whether the subcommittees do need to hold another hearing or workshop or at least a webinar, not a webinars it would have to be two-way. I guess you can do those two-way in order to gather, explore some additional areas and gather this information for this June 9th assignment and I think that the planning committee and I know that I just talked briefly to Todd Park’s assistant who is here, and I know Ed Sondik is on the planning committee, but they have just really getting started planning that meeting and so I think we need to. What I would suggest is that we make some kind of proposal as to what we feel we could do, or be prepared to do and then see if that goes with what they want as opposed to just saying what do you want because it could be things that we are not in a position to really do in this time period.

I think it should be a give and take obviously but I think now based on the hearing yesterday or the workshop yesterday and whether you decide you are going to do another one we really need to decide about that so that we can initiate or continue the dialogue with Todd Park’s office so that we are prepared to deliver. I did I know Todd generously invited everybody to come to the meeting but of course we know that state and local people don’t really have resources for the most part to travel to Washington and impact. That was the only way we were able to hold the workshop yesterday was to sponsor most of their travel.

MS. BERNSTEIN: Have you already to be bold about it asked, I mean invited I assume the members. We are having another meeting of NCVHS the following week.


MS. BERNSTEIN: It is not your intent to travel anybody for that meeting that Todd is holding is it?

MS. GREENBERG: I wouldn’t think that all of the members would want or need probably we would be able to bring in all of the members but we certainly would bring in the people who are going to be on this panel or this presentation.

MS. BERNSTEIN: I assume he is going to have a lot of technological background and so forth for that meeting and so it would be quite accessible to people.

MS. GREENBERG: Yes, it would help. That is true and so I think people could participate virtually but I think we would want to be in person who, however many people are going to be presenting.


DR. GREEN: In the spirit of wanting to move this along, I think we have three deliverables out on the table we need to settle in the meeting if we can. One is just the basic nature of the report we want to do. Is it going to be a primmer? Is it going to be a report? Are we satisfied that the outlines that we have seen before and that Susan is working on are not and so that is one thing it seems to me. That is a key next step. We want to get a result out of what we heard yesterday and we want to get it I think passed. I think we want to just move right along and get that as a product.

Secondly, we need to decide here I think what we want to say we are prepared to do at the June 9th hearing. We need to know if it is reasonable for us to be able to have a product that could be shared at that event that is based on this, et cetera.

Then for me the most important one is I think we need to decide about the nature of a follow on workshop. I think our hypotheses were verified. There are some big issues that need to be dissected and exposed about privacy and confidentiality. I am inclined to want to call that workshop I wish I didn’t know now what I didn’t know then. We will come up with a good name for it or something like that. I really, I think as a strong man, I think we should just decide that we yearn and need and want to start putting together another workshop with an emphasis on privacy, but we can also put another thing or two that is related to it into that exploration.

DR. FRANCIS: We heard a whole list of things about both barriers and some very tantalizing innovative solutions but we didn’t focus on those.

DR. LAND: I would just pick up on what I mentioned yesterday that I thought one of the illuminating things for me was the fact that these were stellar community programs and stellar networks that there is no linkage going on that we could demonstrate yet. Yet we have had so much discussion on this several meetings about we see this is the future for population data is pulling in these systems. So I really think if you all still think that that is the future for population data that we need to be trying to punt somehow and encouraging a couple things I guess. One is maybe there are some things already going on at CDC and other places that we are not aware of that we need to be aware of that they are already thinking about the type of the data that could be pulled out for public health purposes.

The other thing is maybe we should be looking at the work that is going to be demonstrated at the meeting on the 9th is federal data. Should we be talking about the possibility of some of this clinical data going into that whole process and defining what type of data could be coming out of these RHIOs that could be made available just as federal data is being made available? Could we be pushing the RHIO data be coming out and being made available for all of the technology exciting things that are coming on? It is really pushing the envelope but unless we start priming the pump we are going to be a long ways down the line before any of this linkage is occurring.

DR. FRANCIS: I would just say that that is a different path than the privacy path because I don’t think we heard, maybe the RHIO data we did. On a community data, the barriers there are resources, communication linkages, and people just not knowing one another. Those are the kinds of barriers. We have a fair amount. You were talking about that energy of the communities. A number of those communities despite their energy are really feeling the resource.

DR. LAND: Are they what?

DR. FRANCIS: They are really feeling the resource concern at least in the sense that they perceive it as a potential threat to them. They are very worried about their workloads and their status.

DR. LAND: That is why I started thinking that maybe we should be, if we are moving in the direction of having these creative systems, low base systems being developed to make data more accessible that solves some of the resource issues. That doesn’t completely solve it but it solves some. But it only solves it for some datasets. Maybe we should be looking at making more data accessible through this approach besides just the federal data systems. There is a lot of other data out there that we ought to be looking at that could be brought into the picture.

DR. BREEN: One thing that NCI was going to take the lead on but we are not anymore was to kind of hold up some of these statewide and local health surveys. You heard about a couple of them yesterday: the California Health Interview Survey and the Ohio Family Health Survey. Those surveys are very useful for community researchers to take and they are not the be all end all because often times they will want to add more data in order to get information on a neighborhood or a more granular level as they were saying yesterday. But at least as Matt Stefanak said, it brought to his attention that the intervention they have been doing had worked by comparing what was going on with Medicaid and his county with other counties in Ohio and he knew that other counties had done the same interventions.

So that kind of surveillance which is going on in an ad hoc way in different states where leaders have just talked up it has no infrastructure or mechanism for nurturing it or harmonizing the data. NCI has done a study in which eleven of those types of surveys were identified. Three of them were actually the state BRFSS where researchers had gone ahead and done a sample frame of parents and then administered a survey to the parents on their children. We counted those among the statewide local health surveys that were independent from BRFSS. But in any case there are some of those and that is part of the infrastructure that needs nurturing. So that may be something that this committee would like to consider or some other entities would want to consider going forward as something that the committee would want to support at least in a letter.

Then I also wanted to raise a question to what Leslie had said about nurturing these datasets and privacy issues. I am not so sure they are so different. One of the things that Larry talked about a lot is something that he has coined a name for. I love the way Larry coins names for things but a new epidemiology which as I understand it is linking clinical care data with data on the community resources data on the social determinants of health. Doing that kind of linkage, which might be matching in some cases, might be linking in some cases I think we do not know whether there are privacy concerns for that or not but I think that these privacy issues often times they are barriers rather than facilitators and they are clearly they need to be addressed with the clinical data but I think we do not really know in terms of community data but we need that community data because even with health services some people are outside the health care system and they are not getting services but we want to count them and know that they are there. They are part of the denominator and in many cases these surveys help articulate the barriers that people are experiencing; the reason for which they are not going into the system.

So I think that best practices, linking the clinical care data on a social determinants of health, and if we do believe that local data is the future of population health data that we should definitely support it first in our report that we do and any letters that we send to DHHS and also I would say we need to continue to consider these things maybe a little more abstractly in a next set of workshops that we have in order to explore how we could create that into structure. How we could support all of this wonderful work that is going on at the community level which is pretty invisible to the feds unless we really look for it.

DR. FRANCIS: Maya, did you want to say something?

MS. BERNSTEIN: I have been involved in several projects regarding the integration of various kinds of information: social services data, health data, whatever. Pretty much we know that there are privacy issues. That is not really a question. There are privacy issues that are difficult when you marry multiple sources of data from different places you get a much more full idea of an individual and it is much harder to identify them.

That data can also be matched with public sources of data and it is not just the identification of your particular dataset but what people can do once it is out there to match with the other public sources. Those are sort of hot topics in privacy right now. So I think these are challenges. I really bristle I guess a little bit when I hear that privacy is a barrier to this because there is two sides to it. One is yes, some privacy rules make it more difficult for researchers to get data or for policy makers to get the data but if you don’t have the proper privacy controls in place people just aren’t going to answer your questions. You are not going to have the data in the first place to even think about these kinds of things. I think we have to think about both sides of this.

DR. BREEN: I absolutely agree.

DR. FRANCIS: I just can’t resist the comment that it was very interesting dialogue between the HIE and the Denver folks, thinking about what was the community that was involved in looking at how data get used. The woman from the Bronx, Nancy I can’t think of her last name. To comb a very different set of thoughts about how she was going to get the community involved on the board of the RHIO, actually. It was a fascinating set of, I mean it was a real oh, let’s see. You are thinking about what it is to have community research very differently from how we are thinking about what it is to have a community representative on our HIE board. It was just an anyway, that could be a starting point for working on some of these questions. Marjorie?

MS. GREENBERG: I mentioned this I guess before the break a bit, but everybody cannot have all of the data that they want or need and some of the data that people want and think they need, is actually not probably that. First of all it may not have scientific validity. So they like having it but it may be based on such small numbers and such a poor response rate and everything else that, or the way it is collected or the way it isn’t collected. Population health or public health is never going to get the whole electronic health records and they shouldn’t. I have to question that although apparently in New York City they can but I don’t know. That needs a little more explanation and we know that when they did the study with hemoglobin a1c that raised a lot of issues and certainly they have gotten away with it there but I think that certainly is not the national model I don’t think that people are ready to embrace.

I come back to what broad neck what Larry was saying in relationship to PCAST report. Wouldn’t it be nice if we could just exchange some data? Then I think I mean we are so far from even just exchanging on a regular basis being able to exchange some useful data. I was quite struck by the 350 certified EHRs that we heard about this morning. I, not easily, asked Marc Overhage, I said, are they interoperable and he also choked on his coffee when I asked him. So like are you kidding? Then it also relates to I think Garland’s suggestion of really identifying some high priority population health indicators that if you just have those for all of these communities they would begin to help people.

That is where I am sort of going. I think we need to explore these issues but I mean the only issue with small area data is not privacy issues and there are many of them and you can’t deny them. I don’t think it is a question whether there are any. We know that is what happens when you get more and more granular and combine more and more data. You have people here providing testimony to us. We have had them out of the meeting in Utah and all of that. But there are other problems with those data; as I said how they are collected and the scientific validity of them. It seems like getting more and maybe it is like the quality discussion too. Everything starts coming together from this committee in my mind but everybody is collecting different things and they are not defined the same way and they are marginally useful sometimes and we have really got limited resources and it is going to be more limited.

So what ability do we have to contribute to that set of issues as to some key things from the electronic health record if you could just exchange those would be of use beyond the current meaningful use. Some key population indicators and at what level do you need them. Those are the things that I think need to be addressed along with these other privacy issues.

DR. FRANCIS: Paul and then I think Larry.

DR. TANG: I am just refining to Marjorie. The thing about the limited resource is they are already spending the resources. So ironically I think if we had some way to push out some national platform for them to gather more common standardized data elements everybody would be better off at the same total society across and I think they recognize that. I mean they would just as soon not spend that money doing their own individual thing so to the extent you can certainly make that argument which goes a little bit back to what Larry said about this if we could only have a common quote public health practitioner data infrastructure it would just be out there. It really wouldn’t take the top down strategy I think which may be something that any suggestion should get to. It is what is it called, information free circle report from ten years ago. If we had an update of that; it would be updated. It wouldn’t be a restatement. It would be updated in our new world in our new world with the age with initiative with meaningful use with public health being far more.

I mean just 10 years ago, it had none of the impact it does today in today’s world. So it would be more than an update. It would be really recognizing what has changed and what now it takes to have this top down architecture and standards that also now each of us as communities also understand that we can’t just it is not even good to do it our own way individually. Do you see what I am saying? I think the world has changed in multiple ways to say people would like this. I will say it just for illustration but not to go there but just like community health identifier. The world has changed and in fact I even that related but there is now a reason and all of the participants believe in that reason including the individually locally affected folks I think.

DR. BREEN: I am sorry but I missed what you were suggesting we update.

DR. TANG: The information for health, something like we did 10 years ago, or something like that.

MS. GREENBERG: The National Health Information Infrastructure but I have always heard and I know Don Detmer talked about it when it came back, I think transformed from England where they have a more communitarian view of the things. It was heartening just to see that our communitarians out there even if we don’t hear about them on the news yesterday but other countries deal with socio economic status and determinates and we just focus to the extent that we do on disparities, on racial and ethnic which are very important but those have as David Williams and Vicky Mays and other members, have recently with Lisa, and we have some great people on this committee tell us those are very interwoven with socio economic factors. There are communities out there who are more than aware of that.

As I said yesterday, this focus on grocery stores and we have never talked about grocery stores when we talked about data in the past. So there is a lot out there to build on but there is a real disconnect between that and what official sort of policy is and what seems to be what everyone is focusing on but in fact if we don’t address those things I don’t think we will ever address the cost equation which is the major focus.

DR. HORNBROOK: I think what I heard from yesterday was really very amazing and exciting on so many different levels. At the same time though not that I have slept on it in some ways match ups I think are extremely scary. It is kind of like the uninformed student who has learned how to use SAS and makes a gigantic progression equation which explains everything but has not biological plausibility whatsoever. We need to have public health plausibility in these match ups and in these analyses so we don’t go chasing our tails in kind of a usual way with statistics run wild. So I think listening to you Marjorie, I thought you know in a way the same principles that we have used over the years for clinical trial standardization and I am not at all expert in what it is standardization already occurs in public health. I am imaging a set of community health vital signs. If we had the same idea that these ten things were going to be reported reliably, matched up fairly and safely and appropriately then potentially from every community across the country, that might be a good way to start: very incremental; very progressive or not getting in over our head before we really know what these match ups mean. So I would propose that somehow we think about it in sort of community health vital signs.

MS. GREENBERG: Well you mentioned vital. I mean really if you think about a person’s health it starts with their birth and it goes right to their death, right? Some people may think it goes before or after. It does. It starts before their birth actually. I don’t know whether it goes afterwards. That is Larry who is our visionary. But we were trying and there are no limited resources to get standards. We have somewhat standardized birth and death records, but they are not in the 21st Century point of view of the digital world. There is work going on in that but it is grossly under- funded, et cetera.

If you could just get that data as part of – even forgetting as to whether exactly what the source of it is going to be or in the short term or whatever but the basic demographics are in that data. Certainly the cause of death information has a lot of information in it.

Now it does not include a lot of the functioning and other stuff, but I think you can look at some of the key datasets and start identifying what some of the areas that have been standardized or need to be further standardized, and I do think Nancy raises an important things about there are about nine or eleven of these state health surveys, but we do not really even know whether they all.

We know that the California health interview survey is modeled after the HIS, and the HIS I think, is the gold standard for population based data, household data not necessarily behavioral, not risk factor necessarily. But we don’t know about these other ones and there is not even any – I guess there was a proposal for a project but it hasn’t right now, it doesn’t even have a home to try to look at that.

So if you want to tell other communities or if there were funds for limited state surveys or even community surveys. People shouldn’t be making up these elements and definitions on their own. There is years of survey research and contracted research that has been done to understand how is the best way to ask these questions, what are the response categories, all of that. So I feel there is an enormous amount of information knowledge out there that isn’t really being harnessed.

DR. FRANCIS: Maya, oh sorry. I wasn’t sure who I was seeing way down there – then Maya.

PARTICIPANT: I keep feeling we are kind of losing sight of sort of a major thesis. To me a major thesis is about the fact that the federal datasets, everything that is coming out in health data dot gov (healthdata.gov) is at best at the county level. So people are going to come to June 9. I mean it is nice that we have got it all out there and we have got it in an organized form and that there will be app developers who are going to try to make this sort of graphically and intellectually more understandable. But I think what we heard is that the demand is that action needs to take place at the sub county level, and whether you define it as a and what we heard yesterday was people define things as neighborhoods of health service areas, catchments areas. So the question seems to me, the overarching question is how do we help these communities that are organized to do stuff but don’t have the data to do the stuff at the local level, and what is the role of the federal government in making that data available in a useful way to these people; available or helping them collect the data themselves.

I just want to go back to what I see as the three types of data we heard about and one is sort of surveys basically whether it is based on the HIS or whether it is based on BRFSS or it is developed by the community. You get certain kind of data from survey data. You get attitudes, preferences, risk behaviors. That is what surveys are particularly good for.

The other type of data is all of the program data based on all of the services being provided in a community. Just the way you complain about lack of continuity of care and the reason why we have accountable care organization ideas is there is also this notion of the services that are provided in a community by different agencies but are often the same people and the most vulnerable. They are siloed information. They are not organized well. You can’t get it out and it limits what people can do with the data. I think we need to make a distinction between public health and local government action and some other things that go on because government can do a whole lot directly with individuals and with buildings or whatever that sort of voluntary organization can’t and they have a responsibility to make the most of the resources that are expending in their service programs by integrating data and from that data one can sort of learn a whole lot about the causes of health problems, the responses to health problems.

The new piece of this is the electronic health data and the question has been where are we now. Now that we have the possibility of lots of clinical data, electronic health records and the potential of derogating and sharing that data how does that put us ahead like a HRSA advantage. We always tried to make community health centers a community oriented primary approach which means knowing more about what is going in your community then the people you are serving yourself; people that are just in your own EHR.

We are now at the point where we need health center control networks. We are beginning to aggregate that data and we have a greater chance at sort of community epidemiology, community oriented primary care by sort of putting together the data that is in electronic clinical records. The question is how do you do that in a way that protects privacy and that is in a useful efficient way. I think what the focus of the subcommittee ought to be is figuring out how we make progress in each of these three areas. There may be more than three areas while sort of worrying about all of the things we have to do to make this possible in terms of privacy and security and who has access to the data and how does this data get released. I mean health departments can do a lot where they do not have to quote unquote worry about privacy but there are good practices and things they should do and shouldn’t do and there is a lot of confusion and there is a lot of variability across different countries and what people do.

DR. FRANCIS: It strikes me that one of the issues that is most talked about is the HIPAA de-identification data problem and whether data can be de-identified and what do you do if data can’t. But I don’t think anything has been talked about let’s just start from the beginning assumption that maybe de-identification is not there and how do we think about it instead of the question about anyone.

DR. SUAREZ: I want to follow up on my point. I think those three were the core data, the survey data, the public health program data, the clinical data. I think also there was a mentioning of a fourth category which was sort of environmental factor data that can be brought down to the community level whatever that community. I really like the concept I mean all of this is pushing for a locus of the public health agencies to be much more granular. I mean just like we have individual clinical care and physicians and clinics delivering care to individuals and one by one, person by person delivering care. I think public health has traditionally been seen more of a population level but now the population level is becoming more granular and sort of bringing together a closure that clinical individual care level that physicians and the delivery system have with the care if you will of the public health group. It is just sort of now it is public health is delivering if you will care but it is a community level care and so they need data at that level to better deliver that care and to improve the health of populations at a higher level. So I think that is an interesting new kind of direction in which public health is going.

I think the question that I have always had in really as public health begins to have the opportunity to have more granular data what are the preconditions of having that granular data and there is preconditions around policy issues like privacy is one of the big ones. But I heard a lot of other things about the challenge of not having the right preconditions filled in order to achieve that level of granular data like the right public health information for structure, the right set of resources in terms of people being trained to handle all of these new amounts of information that would be coming in. The transition of all of this new, well the way data has been collected and maintained with the standards from vocabulary and performance to privacy and security standards to all of these new standards that are seeing in the environment and the new semantic and vocabulary and terminology standards and new transaction standards. All of these and how public health now needs to begin to get ready in order to not only get more granular data but get it in a new format with a new vocabulary and electronically and in an efficient way and in a more timely way as well.

All of that implies a lot of a new needs and new requirements before we can get there. So I think losing sight of the elements that are required this public health information and risk factors the concept that we talked about yesterday of the risk of digital device and things like that. I think those are going to be very critical to not lose sight of.

DR. BERNSTEIN: So as I am listening to the conversation, particularly Mike’s layout of the different kind of data we collect it struck me that one of the things that I tell other groups that I work with on integrated data is to think of it in particular that second category of program data. Anybody might answer a survey. Everybody touches the health care system one way or another but program data tends to be about vulnerable populations: poor people, disadvantaged people need more services and use more services so we have more data about them. Even though we all have good intentions about how we are going to use that data, we want to make better public health policy choices and we want to improve the quality of their care and the quality of services and so forth, we are using data on I tend to think it is another way that that population is made vulnerable by the use of their data. So I just want to sort in the back of our minds to keep that in mind.

One of the things we heard what I struck by yesterday we in talking about the public health data infrastructure is someone, it wasn’t me by the way, said that there is no common set of privacy rules for public health like we have now in HIPAA for clinical health data. So we heard from one community that they need individual consent every time they want to share certain kind and I think we heard from another community I think it was actually Nancy that with constant public health data we can just give it over to public health really because public health data is being used for public health purpose.

There is a lot of variety in local, community or state level rules about what data can be used and one of the ways that we sum the privacy part of this committee that might be useful is to think about as a part of that public health data, public health infrastructure is to start thinking about a common set of what are fair privacy rules for public health that could be adopted sort of as a standard across the board that would help us to both share information and protect information in a way that would be useful moving forward.

DR. BREEN: Point of clarification. Are you speaking of program data, administrative data in terms of needing some privacy rules?

MS. BERNSTEIN: I am just saying that among the states and counties and so forth there are different rules about how their data can be used. Each state, each county, whatever can have its own rules and they are not standardized across the country. So there is no standard way of using data or of making data available or certain access to different kinds of data and we could start thinking about that specific project and I have some hesitation because of the kinds of different opinions on the committee and difficulties people have and disagreements people have about what is the right thing to do and what is there but I do think if we could come up with some kind of recommendation it would apparently be very helpful in getting more data shared and also getting proper protections on data that were universal and sort of setting some kind of.

DR. FRANCIS: I think actually that is what Sallie and I were talking about in terms of their practices or at least some. I know Marc wants to say something as co-chair so I will let him say it and then I am going to go to Larry and to Vicky and to Mark.

DR. OVERHAGE: Actually this is just kind of a brief response why and the comment you made earlier, Leslie which gets back to this problem what a lot of people are looking for is a framework that they can be comfortable about and confident in that all of the parties are about. In some ways what you are trying to do though is write a manual for trust and Leslie talked a little bit about having to build trust. You can’t legislate it. You can’t buy it. It is interesting. How far down the road can you go especially with the consumer level of trust if there is sort of this big thing in the sky that gets handed down but I think there is that tension in a lot of these things.

DR. FRANCIS: One view thought might be a process view like making public what the uses are. So that is a different and that is not a substance. Yes, the framework could involve all sorts of different kinds of things. Larry, and then Vicky.

DR. GREEN: Michael it is time. We have seventeen minutes left and Justine gave us three charges and Nancy over here, I want to deliver. So in that spirit, one of the things she said was think in a three year envelope. Don’t think about just one meeting from now but think a longer view about where the committee wants to go and I believe there is a solution there for us in this conversation. If we take a longer view, if we take this three year view, I have heard in the last twenty-five minutes seven substantial juicy, needy areas that are begging for attention from just what we already know. So again I do not want to overdraw this but we need to label this something and as you know I like the notion that just three years of work we could call it population subcommittee point of view but what it could really be about, what the galvanizing focus could be is that we believe we are spotting the emergence of a new epidemiology that can be very, very powerful to improve the health of the population. We want to learn about how to let that epidemiology evolve and emerge but we want to do that in a way that is really helpful, that is affordable and achievable, practical, that sort of stuff and also that is safe and fair and respectful.

We need, I think we need to sort of prioritize some of the areas we could attack and at the risk of annoying everyone, I want to say what I think we ought to do next.


DR. GREEN: I am convinced that I want to build off of what Marc just said. There is a nearly desperate need of the creation of some sort of framework of trust. We heard this repeatedly yesterday that these communities that are in motion and that are working on the local health issues, they trade on trust. They trade on being believable as respected folks. It is okay to talk to. You want to fill out their surveys. It is okay if your program information goes to them. They may even be someone I would let have my medical records. I think we should go there next. I believe this is begging for more attention. I would go back to something Garland said I think about a half an hour ago about you know you have got to have some sort of notion of what the data are that we are going to link. I think you said something like we were going to look at how we could create this new and better epidemiology by linking data and particularly data from multiple sources that now we are in position to include service records that might be an issue.

Why don’t we decide that as soon as we can organize it and get the questions framed that are ready to be addressed that we do a follow up to the workshop we did yesterday with the Privacy co-chairs in the lead organizing that and we see how quickly we can organize that and see if it is possible to know something from that before the June IOM event. Is that ridiculous?

PARTICIPANT: No, I don’t think it is ridiculous at all.

DR. GREEN: You may want to shoot me for saying that out loud because it just.

DR. FRANCIS: I don’t think it is ridiculous at all. I think it is.

DR. GREENBERG: Not just to having the workshop.

DR. FRANCIS: And to learn some things from it. A formal report would be ridiculous.

MS. GREENBERG: What we talked to Todd about was not having a final product to tie up with a bow but to have something that we could present as almost like a trial balloon but you know with our best thinking.

DR. MAYS: I wasn’t at the workshop but I did get at least some insight in terms of hearing what Susan said, and then Susan and I were chatting. I want to comment on some of the suggestions that Marjorie made because there is really some work underway that you know one might want to think of in terms of building on it.

First of all this issue of like social indicators and thinking what it is the community wants. There is actually a fair amount of work that is being funded by the foundations. I mean part of why you are able to have some of the California people you have is because they are being funded by the California Endowment.

There is another group that is at the funding is at the University of Wisconsin, David Kindig’s(?) group, that is actually doing MATCH which is like the Healthy Cities and the Healthy Cities are actually starting to come up with report cards. What they are doing is standardizing across and Healthy Cities is across the U.S. and they are standardizing across in terms of what the community has said is important. That is why you get into food and food and security and all of these other kinds of things but again it is the foundations that are really I think taking the lead.

The other thing that is going on is there is an Institute of Medicine committee that you may want to look at. I am on it, but all I can do is tell you to look at the charge but it is on HIV/AIDS, in which what they are looking at is what are the data that we need in order to be able to evaluate HIV/AIDS care? Again the issue of data that comes from community sources, data that is a linkage between clinical and public health. All of those things are potential ways that one might think about in terms of trying to answer that charge whether they will or won’t that is a different story but I just think that there is underway quite a bit of what you are seeking that you really could use as your foundation.

DR. FRANCIS: Mark and then Walter.

DR. BREEN: I think that the group brings up a good point because just for example is funded by California State. The California Endowment as she said as well as federal agencies and a lot of local municipalities and it really is a public private foundation co-funded project and I think David Kindig’s is as well and I think most of the surveys that we identified in the NCI study or the NIH study actually were that way too. They were funded that way.

But I also and that may be part of the framework of trust that Larry was talking about. I would like to hear more about that. I think maybe you were thinking about on the ground but this coalition building that we heard about yesterday was so important to the success of obtaining data and releasing data and probably the example that is the most difficult in the United States is the example that is the most piercing would be that of the American Indian population but for every community they really want to know that they can trust the people they are sharing information with and that that information is going to be used in their best interest. That is the bottom line for getting this granular data analyzing the granular data and using it appropriately. It needs to be used in a way that is going to help the community and that they are confident about that.

I am not sure how we would go to investigate how you build a framework of trust. We obviously saw examples yesterday but are there some guidelines?

DR. FRANCIS: So there are two separate questions: how do you get the guidelines implemented and what should be the starting ones. I am hearing a lot of suggestion that we need to do next step. I am going to make a proposal and then if anyone has follow up thoughts on this but I think that this is following up on Larry what we should be doing on the privacy perspective is planning another workshop probably in May, late April to early May at least a month. We need at least a month and a half to get it going and this is February so March, April, beginning of May.

PARTICIPANT: I am just thinking of the June 9th date.

DR. FRANCIS: Right, well that would give us a month before the June 9th date, the goal of which would be to explore what the core elements are of a framework as in particularly clinical information becomes part of the mix.

PARTICIPANT: Focused on clinical information?

DR. FRANCIS: There is a question. Should it be focused on clinical information?

PARTICIPANT: No, I think it needs to be more broad.


DR. GREENBERG: We have been using a clinical model.

DR. FRANCIS: Broad framework.

DR. GREENBERG: One at a time. I take that back.

DR. FRANCIS: Okay so what are the core elements of a framework of protecting privacy and confidentiality in such a way that we generate trust on the public health side? Core elements do not have to be substantive. They could be procedural and could include things like what is in the 1974 or ’73 government privacy.

PARTICIPANT: Their information.

MS. GREENBERG: We have got model laws out there like we have one for. We are not starting from scratch.

DR. FRANCIS: People thought about the identification but that was obviously not where the action is. Matt sorry.

MR. QUINN: Something that comes to mind with this in the whole trust framework. This is very much a bottom up process. In the past few years with the emergence of Web 2.0 and online communities of people voluntarily creating communities around everything from learning out there, communicating with other people with a particular disease or other people into model airplanes. We found that people who have shared interests are both able to aggregate in ways that were not possible in the past, and to do it in a rate and breadth and depth that has not been possible in the past. With the pervasive use of smart phones, lots and lots of people have access to everything from Facebook to whatever.

Absolutely this should be part of the framework and as we move from thinking about how do we aggregate and display information at the county level or below this is the way that that takes shape in actionable things. So going from we have a childhood, a shared childhood immunization registry to how do we use that to better coordinate the immunization buss that comes to schools.

DR. FRANCIS: So question whether the 2.0 model is a model or not given that it is a voluntary model. Other quick questions? Mark? Walter?

DR. HORNBROOK: I am just impressed at how much data choices we have and you really have to then ask the question what are do you want to accomplish? What is your goal? In some sense we have to then go back and say what do we mean by public health? Are we talking about infectious diseases and toxins and sort of the notion of communicable disease and free riders? Are we talking about life style issues that we need to teach people how to live better or are we talking about social injustice and your trust being a whole other concept.

I think it would be easier for us to move forward if we had a vision of where we want to go.

DR. FRANCIS: Larry and then Walter.

DR. GREEN: We have varying levels of understanding of what is an adequate statement of our goals but what is sufficient for me to have a notion about what NCVHS could be doing over the next three years is this notion that the world is changing. It is in motion. We don’t understand it yet. It is an emergent event that is going on but that we believe that we have spotted really an unprecedented opportunity to create this new science of what is a bone to people. What upsets the people? What do they care about? How are they doing that physicians, technologies and things that are emerging right now to create knowledge that can really help virtually everyone but that that is not going to come to maturity automatically? It has some things that could enable it; some things that could impede it. But the overall goal is to wind up in a few years having this country covered with communities that are a learning system about themselves and their health and what it takes for that to be a healthy community and for the people living there to be healthy and to know whether they are healthy or not. So I find that is sufficient to sort of guide our discovering our way through this and this privacy trust framework is just begging for immediate attention.

DR. FRANCIS: Core piece of it. We will do it. Walter had a comment and then I think Marjorie and then we have to cut it off for Quality, it is their time.

DR. SUAREZ: Okay so I am going to talk to two quick comments. The first one is on this priority. Larry, I think that is great to begin to think about one priority. I was hoping that at least there would be two because I think the privacy group can work on the framework of trust but I think we need to again not lose sight of the public health information infrastructure part of it. I was hoping that the population work group or subcommittee could focus on perhaps that as a priority or another a priority that can really come back to emphasizing the significance and the importance of advancing or at least moving a little bit at the public health information infrastructure because otherwise we would have a beautiful fabric that is not applicable to anything because the information perspective now days. That is one comment.

The other one is about the fabric of trust and the concept of trust. I could not agree more with Maya about the priority of there is a variety of perspectives. I don’t know that there is really any state that has whether it is at the state or the local level a very open policy of you know you can get any data from here. Actually if there is any area where there is a lot of restrictions on who can access and who can release this data it is in the public health arena so I don’t know. I mean there are more stringent states that have a lot more controls over what public health agencies can do with the data and how they can release it but across the board and I want to be on the record to state that public health agencies have some very strong protections of the health information and how they discloses it.

I lived in Minnesota for almost twenty years and lived through and within the Minnesota government Data Practices Act, one of the first national frameworks on data practices for government data including public health data so public health is really in the hands of government. It is really part of government data just like Commerce, Transportation and other things. So that fabric begins to extend to when you think about assessment of what are the protections, what are the restrictions that states have around the country of public health data when it comes into the hands of public health agencies. I think we are going to see a variety of perspectives but one of the commonalities I think is going to be this is government data and there is a government regulation on the data the government has in its hands. And with respect to the clinical data, public health agencies across the board have already for many years collected individually identifiable health information including some clinical data and so you know the ability to now access electronic health records or perhaps collecting data from them I don’t think is new. It is probably more in the volume issue rather than in the – this is the first time we see clinical data in our hands and now we have created something new. I think the framework from the perspective of a recommendation is really we have to continue what are the approaches the states have used to protect government data and how they handle it.

DR. FRANCIS: I think that will be an important part of how we put this together. We do need to let Quality. Marjorie two sentences.

MS. GREENBERG: Okay. I have heard some long term and some short term activities all of which are going to involve Larry and Sallie as well as Marc and Leslie and others as well. But I think we need a plan here. Now I heard two short term and when I say short term I mean over the next few months. You said does this workshop.

DR. FRANCIS: We are doing the workshop.

MS. GREENBERG: And so we are going to have to organize teleconferences and I am assuming Maya you are the lead staff here.

MS. BERNSTEIN: That is what we are assuming.

MS. GREENBERG: But then you have to identify what other resources you need, et cetera.

MS. BERNSTEIN: Address and I am grateful for the help of the rest of the committee and staff.

MS. GREENBERG: Right but let’s put that on hold right now because the other thing I heard that we should get done in the next few months is a report out of yesterday’s hearing which certainly does relate to the public health infrastructure issue. It would only tantalize on the privacy, probably, issues because it would talk about the trust and all of that but because it is going to be a –

DR. FRANCIS: Next step.

MS. GREENBERG: Can we have both of those going on concurrently and with Marc and Leslie taking the lead on the privacy and Maya taking the lead on the privacy hearing or workshop and Larry and Sallie and probably continuing obviously continuing to work with Susan.

DR. OVERHAGE: Sallie and I quit if we cannot have Susan.

MS. GREENBERG: Yes, I know, I know. I wouldn’t take Susan away from you except for the fact that as long as she does the minutes also, anyway those two need to be going on concurrently. That means we are going to need now if there is some Populations people who are not involved with Quality we can go next door if you want to talk about this outline that Susan developed. I mean it won’t be taped but that is fine. This is a work session. So that option is open to you so those are two short term and then I really like the idea of fitting this into a three year plan al la our chair here because there is more than enough to be done over three years and I think over between now and June you could have some productive calls also and maybe at that workshop. I think we would want the population health subcommittee at this workshop.

DR. FRANCIS: Absolutely. We assumed it is a boat going this way with the initial tug was populations the next tug is privacy. The tug after that is population.

MS. GREENBERG: Okay so we could have both of these things some output before the June 9th meeting it sounds like but it is going to require getting to it. So am I right on that? Okay. So if you want to go next door some of you, you can or.

DR. BREEN: Are you going next door and if you are is there a way to call in or otherwise I will just hang up.

MS. GREENBERG: No phone.

PARTICIPANT: Put your cell phone on speaker.

MS. GREENBERG: We can put you on our cell. We will email you.

(Whereupon, the subcommittee adjourned.)