[This Transcript is Unedited]
Department of Health and Human Services
National Center for Vital and Health Statistics
Subcommittee on Privacy, Confidentiality, and Security
“Net Steps for Community Data Use”
April 17, 2012
Double Tree Hotel
8727 Colesville Road
Silver Spring, MD
CASET Associates, Ltd.
Fairfax, Virginia 22030
- Introductions and Opening Remarks
- Leslie Francis, JD, PhD, Co-Chair
- Linda L. Kloss, MA, RHIA, FAHIMA, Co-Chair
- Panel I -Beyond Data Use Agreements: Governance Models
- Kelly Edwards, PhD, University of Washington
- Bradley Malin, PhD, eMERGE Consortium, Vanderbilt University
- Phillip Smith, MD, MPH, Chair, National Institutional Review Board, Indian Health Service
- Panel II – Protecting Small Groups
- Paul Spicer, PhD, Professor of Anthropology, Center for Applied Social
Research, University of Oklahoma
- Malin Villegas, PhD, National Congress of American Indians
- Paul Spicer, PhD, Professor of Anthropology, Center for Applied Social
- Panel III – Using Results to Improve Community Health
- Jerusha Nelson Peterman, PhD, Assistant Professor of Nutrition, School of Public Health & Health Sciences, University of Massachusetts, Amherst
- Linda Silka, PhD, Director, Margaret Chase Smith Policy Center and Professor, School of Economics, University of Maine
- Panel IV – Consumer Attitudes about Community Health
- Initiatives – Dave Kaufman, PhD, Director of Research and Statistics, Genetics & Public Policy Center, Johns Hopkins University
- Ryan Spellecy, PhD, Associate Professor, Medical College of Wisconsin
- Statements from Members of the Public
- Wrap up and set agenda for next day
DR. FRANCIS: On behalf of the National Committee on Vital and Health Statistics and in particular the Privacy, Confidentiality and Security Subcommittee of that committee, I would like to welcome you all here to our hearing on Next Steps for Community Data Use.
I need to just start by thanking Maya and Gail and others, who have worked so hard to pull this together. It is really a difficult hearing because we didn’t even quite know what questions to ask.
Just a little background and then we are going to go around and introduce everybody and deal with the conflict things. About a year ago the committee, the Population Subcommittee and the Privacy Subcommittee, held several hearings on how data can be advantageously used by communities. We published a report, The Community as a Learning System for Health, which emphasized the importance of trust. But now we are trying to think about what that actually means.
We are tremendously aware of the multiple initiatives at the federal level for the use of health information, new ways of aggregating it, getting it, new powers out of clinical records for secondary uses, and what some of the protection issues might be when information is reused or put to secondary uses for what we all know are wonderful advantages for communities.
What we are here about really is brainstorming with people who we know have been thinking very seriously about these questions, how can activities of data use for community benefit, go forth in a manner that is ethically and politically acceptable. With that just little prelude now that everybody is settled in, what we are going to do is have introductions and then we are just going to jump right here.
I am Leslie Francis. I am from the University of Utah in the law school and the philosophy department. I co-chair the Privacy, Confidentiality, and Security Subcommittee. I am a member of the full NCVHS committee and I don’t have any conflicts.
MS. KLOSS: Good morning. I am Linda Kloss. I am a consultant in health information management. I co-chair this subcommittee. I am a member of the full committee and I have no conflicts. We look forward to a very productive day.
MS. GREENBERG: Good morning. I am Marjorie Greenberg. I am at the National Center for Health Statistics, CDC, and executive secretary to the committee. I just wanted to comment that this hearing is really timely because as of the June meeting the committee will be launching a new working group on HHS data, access, and use. We appreciate that you are foreshadowing that.
MS. MILAM: Good morning. I am Sallie Milam, West Virginia’s Chief Privacy Officer. I am with the West Virginia Health Care Authority. I am a member of the full committee and this subcommittee and co-chair of Populations. I have no conflicts.
DR. TANG: Paul Tang, Palo Alto Medical Foundation, a member of this committee, full committee, and co-chair of the Quality Subcommittee and no conflicts.
DR. BURKE: Good morning. I am Jack Burke. I am the compliance officer and the privacy officer for Harvard Pilgrim Health Care in Boston. I am a member of the full committee, a member of the subcommittee, and I have no conflicts.
MS. HORLICK: Good morning. I am Gail Horlick from CDC. I am a senior legal analyst there and I am staff to the subcommittee.
DR. GONZALEZ: Good morning. I am Natalie Gonzalez. I work at CDC. I am a public health analyst there. I am not staff, but just thought I would introduce myself.
DR. SMITH: I am Phillip Smith from the Indian Health Service. I am here as a guest.
DR. MALIN: I am Brad Malin from Vanderbilt University. I guess you could also say that I am guess, but I am also here representing the electronic medical records and genomics consortium.
DR. EDWARDS: Good morning. I am Kelly Edwards, one of the speakers this morning, from University of Washington Seattle. I am an associate professor in the Department of Bioethics and Humanities in our school of medicine, also in public health genetics in our school of public health. I am here partly on behalf of our Center for Genomics and Health Care Equality.
MS. KHAN: Good morning. I am Hetty Khan. I am from the National Center for Health Statistics, CDC, and I am staff to the subcommittee.
MS. PETERS: I am Iliana Peters. I am with HHS Office for Civil Rights and sitting in for Sue McAndrew this morning.
DR. MAYS: Good morning. I am Vickie Mays. I am with the University of California Los Angeles Psychology and School of Public Health. I am a member of the full committee and a member of the subcommittee and I have no conflicts.
MS. BERNSTEIN: Good morning. I am Maya Bernstein and I am the privacy advocate of the Department of Health and Human Services. I work in the Office of the Assistant Secretary for Planning and Evaluation. I am the lead staff to this subcommittee. Welcome everyone and thank you for coming.
DR. FRANCIS: Do we have anybody who has called in? Do we know over there? No. Around the room behind me.
MS. SQUIRE: Marietta Squire, CDC, National Center for Health Statistics, staff to the committee.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics, CDC, committee staff.
DR. SPICER: Paul Spicer from the University of Oklahoma, speaker for the next panel.
MS. JONES: Katherine Jones, CDC, National Center for Health Statistics.
DR. FRANCIS: The idea of this first panel is to think about ways of protecting data through governance that go beyond the contractual model of data use agreements. Without anything more I am going to just turn it over to our wonderful speakers. We have asked them each to take a few minutes. They have submitted written testimony. They are going to hit the highlights and then we are going to jump into a discussion. Our first speaker is Kelly Edwards from the University of Washington.
DR. EDWARDS: Thank you very much, Leslie and to the committee for again inviting me today to talk with you about this important topic. I think rather than read my remarks directly as you have them in front of you, I do want to highlight just a couple of things out of them. One of the biggest things that we have come to when we have thought about use of either public health datasets or some of these existing repositories or registries of information, how do we use these resources for research purposes. And to our group this has really become one of the places where there is a disconnect between the regulatory requirements and perhaps our ethical obligations or our ethical sensibilities.
I do think that in many ways what I call the cautionary tales that came from Texas and now Minnesota with the newborn screening blood spot collection and the public response to the Henrietta Lacks’ story, the Havasupai case that I believe we will hear more about today. All of these cases point to the fact that our existing regulations that allow us to use anonymized datasets for research purposes without obtaining further permission or notifying the participants of that use is perhaps not enough to assure people that we are working in their best interest. What can we do above and beyond what is required of us in these regulatory requirements?
I understand the logistical challenges. If you have a nationwide dataset or a statewide dataset, really how do you go about an appropriate notification or letting people know what kind of work you plan on doing with that dataset? I think these are public disclosure challenges that need to be discussed and overcome.
Personally as someone who works in education, I think this is an exciting opportunity for public engagement, public education, and really doing some more outreach in an exciting way. What are our hopes for public health research? What are the desperate needs with our health disparities and our health inequities that could be begun to be addressed if we utilize some of these existing public health data repositories toward health benefit for our communities?
And if we could help people understand what is the kind of community health research we have in mind, and really how that data can be used to benefit their communities, I think we might be closing the gap in understanding of what we are trying to accomplish. What we are trying to accomplish isn’t what people refer to as career building research or curiosity research or research that might harm communities. What really our intentions are trying to work for a community benefit. There are a number of structures that I recommend that can help communities feel more assured that these datasets are being used for community health benefits, things like community advisory boards, community consultation processes, more participatory models of governance that can help. There is not just a single body that is making a decision in more of a paternalistic way saying just trust us. We are using this data in your benefit.
I will just stop there for now. Certainly I appreciate this committee’s emphasis on trust as being the most essential resource that we have. It is a resource that shouldn’t be squandered. What can we do to really demonstrate respect and accountability for the data that we have? Thank you.
DR. MALIN: First, I want to thank you guys for inviting me or at least inviting the eMERGE network to come and speak before you. What I didn’t say earlier is that I am an associate professor of computer science and biomedical informatics at Vanderbilt. My perspective on this matter may be a little bit more skewed towards the technological perspectives.
I also wanted to start by telling what exactly eMERGE is because sometimes there is a little bit of confusion. First and foremost eMERGE is a consortium that is sponsored by the National Human Genome Research Institute at the NIH. It was designed to be a nationwide consortium that would develop, disseminate, and apply approaches to research that combine DNA-based biorepositories with data derived from electronic medical records. It was not designed to conduct prospective studies, but more retrospective studies to see what was capable with data that was already on the shelf.
The network at the present moment consists of seven-member sites and they are distributed across the United States. That includes the Essentia Institute of Rural Health in partnership with the Marshfield Clinic, the Geisinger Health Corporation, Group Health Cooperative of Puget Sound in partnership with the University of Washington, the Mayo Clinic, and Mount Sinai Hospital in partnership with Columbia University, Northwestern University, and Vanderbilt University where I am coming to you from. You can see we are from all over the country.
When we think about data management and governance in particular, I think it helps to look at this from a lifecycle perspective that gets splits into roughly three different sections: initial data collection, the second is data utilization, and the third is data dissemination. I wanted to go through a couple of these and show you how we have addressed these within the network.
When we solicit data from the community, eMERGE does not do this in a completely standardized way. It was recognized from the outset that we have different locales each with disparate policies and procedures, as well as different populations from which we are collecting the data. Each eMERGE site has the opportunity to consult with its constituent populations to establish principles for data collection and research oversight as they see best fits their needs.
I am not going to go through all of the details of how we do that, but in addition to this testimony I believe I sent you a worksheet that documented what each of the current seven sites is doing from many different advantage points from community engagement.
MS. BERNSTEIN: We have a copy of that. And the print on it is quite small.
DR. MALIN: Look at it electronically. It would take me more than 5 minutes to read that. The only thing I wanted to highlight was that there is — we have been doing this for about 7 years now. That really summarizes what has been going on each of the sites. It is difficult to summarize that in 5 minutes.
What I did want to highlight regarding what people have been doing or what the eMERGE sites have been doing, is three different perspectives on how we get the community involved. The first is a community engagement model. All eMERGE sites have utilized community engagement models in one way or another. I will give you one example.
The first example I give is from Mayo Clinic. When they started their biobanking efforts and their reuse of the electronic medical record system, they adopted a deliberative democracy model. They actually engaged community members in open dialogue over four days of activities. The deliberants were provided with background materials, biobanking, biomedical research, as well as the local efforts at Mayo. They were then given an opportunity to interact with domain experts including scientists involved in genetics research as well as privacy advocates.
In the following several days of the democratic process they spent the time debating the issues and formulating specific recommendations for Mayo to move forward with their system. Barbara Koenig have published on that and you can look at that further in their publications to that reference.
The second thing that we do with the community is community advisory boards. These are in place at every site in the eMERGE network and they have various names such as Community Advisory Group at Marshfield, Community Advisory Committee at Northwestern. The reason why this was done is because when we were going back to reuse the entire electronic medical record system as well as the biorepositories at these institutions, it was deemed that this had some very special connotations associated with it.
In particular, I will use Vanderbilt as an example. We built our system in a manner that it was an opt-out, de-identified by a repository. And the Institutional Review Board agreed that in the way that we had set this up it did not meet the criteria of human subjects’ research.
However, given the potential scale and the potential impact on the local community it put additional safeguards in place. These included evaluation by medical ethics committees at the university as well as the establishment of Ethics, Scientific, and Community Advisory Boards, three different boards that meet multiple times per year.
The Community Advisory Board of the CAB — it was basically established to ensure that the community had a voice. They had the opportunity to interact with people at the university. They also had the ability to bring things back to the community as necessary. It initially consisted of 12 people that covered a wide strata of the community including parenting activities, church groups, civic communities, educational activities. It was not expected that people would have educational or genetics background. They are quite vocal. They are not passive in that they just listen to what you tell them you are doing. They want to know what is going on. They want to provide recommendations. They even have tried to open up the system more than we have actually felt comfortable opening it up which is nice to know.
Beyond the boards, I also wanted to highlight that it is critical that you provide information back to the community in general regarding what you are actually studying and what has been found. The Marshfield Clinic in particular has done a really good job of providing newsletter, for instance, back to their community to say this is the type of study that we have been running with data out of these systems and this is how it may impact you, not necessarily you specifically. This is not what we found for you.
I am not going to talk about return of research results today. That is a whole different ball game. And HGRI has established a whole new subcommittee to deal with that. But providing aggregate results back to the community and just keeping them involved has been I would say a lifesaver. That is about community engagement. That is the first step in the life cycle.
At the second step in the life cycle we actually have to say what is actually going on with the data, who gets to use it, and how are they using it. In eMERGE investigators when they access data, they access it in their local system through traditional mechanisms. They go through the Institutional Review Board because everybody has different management files. Vanderbilt has de-identified data. Other places have identified data.
But at the same time there are these additional oversights that I was starting to allude to, for instance, focusing on what happens at Vanderbilt. For every research study that comes in we do use the data use agreement as you can imagine, but we also register each study with its own specific number and it gets audited in its own. For instance, if you have the same cohort being studied for two different pieces of research that is going to require two different registration processes. And hopefully that will ameliorate some of the Havasupai issues.
This does not necessarily prevent you from misusing the information or misusing the repository, but it will at least make investigators cognizant that what they may be doing is against the policies and against the wishes of the community from which the data was collected.
I was going to say that one of the challenges at this point in a consortium is that even though we were charged with constructing new methodologies for analyzing EMR data in a secondary use fashion, this is a nontrivial challenge because we have different types of electronic medical record systems. There are different business practices at each hospital which ends up skewing what types of diagnoses end up being made and the comorbidities end up procuring the data.
This is where we found that — if you really want to do something for the local community, if you want to develop algorithms for analyzing this type of data, you have to do validation in the context of a network because 9 times out of 10 when an association study is run with these types of data, we find that it is debunked in a further study.
The eMERGE consortium has spent years figuring out how to do these analyses in a round. One site proposes a phenotype model. And this is what a person with type 2 diabetes looks like at Vanderbilt and this is what we think they will look like at other institutions. And then we throw that across the river to another institution like Northwestern and they tried to do this type of analysis in their system and they give us the positive predictive rates and all of that and that goodness. But this type of a feedback where we are using now seven different sites for these types of evaluations has been invaluable towards recognizing where the biases live and trying to come up with generalizable algorithms. This is not so much getting the community involved. This is to ensure that when you actually do research with their data if you want to determine if the findings can go back to them and they are useful, you are giving them the right thing at the right time and you are not reporting things that are premature.
I could go on and on. The last thing that I will say is that when we share data beyond the network, we do adhere to I guess you could call it HIPAA style de-identification requirements. And this is a moving picture at this point. Maybe you don’t have a moving picture, but this is a moving process. But we have spent a good deal of time trying to determine what constitutes identifiability with respect to biospecimens and biological data as well as clinical information. And in this context we are giving I wouldn’t say certifications, but we are managing identifiability in the context that it is a recognized set of users of the data. When we give data to the NIH for these purposes, it is not being given as carte blanche reusability. You have to provide some additional documentation of what you will be doing and agreeing to it here to the security and privacy standards that were set forth at any particular site, not just what the NIH says you have to do.
That is it. I will stop there.
DR. SMITH: The Indian Health Service really appreciates the opportunity to provide some comments to this committee and we are very appreciative of the time that has been allotted to us to do so because your business is very complex.
I thought by way of introduction I would start with a little story in terms of maybe give you a perspective of how we reach some of these things. As medical providers, we oftentimes engage and interface with patients in an unknown setting such as the case with me. Oftentimes when I am out on the reservation the patients that are referred to me are those who don’t speak any English. Because I speak Navajo, my first language, they are sent to me and such was the case with a man that was sent to me that have been seen for abdominal pains and of course went through the gamut of all exams that were there. Every excretion was evaluated. Every orifice examined. I was ready to go in to talk with him and was looking at the x-rays. I think if we had a CT scanner, we would have done that as well.
I was looking at that and his wife happened to come by. And being a good physician I said to the wife — I will give him the name John. John is in there. What do you think his problem is? And she looked at me and says ask him what he does on weekends. I thought well I will go in and I am thinking what does that mean. I go in there and of course everything is normal. Then I go aha. Without saying everything to him, I said John, we have done all these tests. I don’t know what the problem for your abdominal pain is, but I think it is due to alcohol. He looks at me — of course this is all in Navajo — he looks at me with a straight face and says, don’t worry Dr. Smith. I will come back when you are sober.
And that is sort of the setting I think that we are in in terms of talking about this because we have some super experts here who wrote doctrinal thesis on this subject. And then we have individual patients who are concerned and wondering if the information they gave is going to be used and affect them in some way.
I think such is the case with the Native Americans. We have had people come in literally called helicopter research, come in from universities, pop in, take the blood, and disappear. Later their grandkids are saying, oh, in the New England Journal of Medicine here is an article that was written about you that says you did this.
And to this day many of us who read Margaret Mead’s thesis on sexual practices among Navajo, we joke and laugh about it and say look at this. It would be like having an individual go down to the mall and examine the practices of a few people at different malls and then describe their sexual dating practices based on a simple observation.
Today, I think the key that I would say that we are concerned about is we have gone beyond the oral data collection. We have gone beyond the written oral collection. We have gone viral. Safeway knows more about you than your mother ever did.
With that background, I just wanted to quickly go through what I call a little slide show. This is the setting that I think you need to be aware of. This is Havasupai. Many of you have been down there. A small tribe that really felt the impact of research at a university level that was done in a very professional scientific way using the values, where IRBs cleared wards, said everything was very ethical, appropriate, et cetera not realizing that the consequences of how those egregious’ were made. Would that impact the whole people? And then you would see a whole tribe once they discovered that going around crying and moaning and having to go to purification settings because of those settings.
Of course most of them in the scientific world are just pooh-poohing the whole thing and say it is not a big deal to be stigmatized in this way. And through the regular processes they did not have any remedies to pursue in terms of the governance of IRBs, privacy issues, or anything, and so they took it through the judicial system and of course the university in its good judgment I would say, settled with the community. We were talking about this just years ago. We are not talking the Nuremberg Trials and the atrocities of the Nazis. But yet there is a similar setting in that setting.
Having said that I thought I would just highlight a couple of things, not going through history, but just wanting to say historically, there are a lot of things that have impacted American and Alaskan Natives that impact us today.
Just a quick reminder, I think people forget that American history is like European history in terms of certain people being oppressed in many ways. But yet in America, there has been a real transition from a campaign to totally eliminate a race, to now really having an arms around type of approach and allowing a little governance to take place. These things are just highlighted mainly because they have impacts. To highlight a couple of things that the experience is really different than other populations and we are living this setting.
To make a point clear is that these settings are very closely linked. I am the last child in my family and my dad was born in 1896. My grandmother was the one who walked back from Fort Sumter back to Fort Wingate. It is not something that is like the Middle Ages, it is something that is still impacting us today. Sometimes when you are dealing with them, you need to be aware of it.
And then secondly is because we are very nice subject. There is a lot of homogeneity in our setting and there is sort of an exotic setting. There is a setting for people to look at things.
A couple of things just to highlight are to note that the tribes are very different, very distinct in many ways. There are just some things that one needs to look at when we talk about those settings including the gene flow and the studies that have just been referred to recently. I borrowed this from the US Census. That is why the question mark up there on the migration activities is still there. I guess I should have removed that.
But in terms of research just wanting to let you know the key thing I think is communication and as I alluded to earlier and the various forms that have changed really make it very important that one not only looks at the culture of communities, but the culture of science. Just watching and interacting a man of letters with an individual who has none you see distinctions in terms of how this communication level occurs whether it is formally, informally, or otherwise. And also the idea that resources is really there. Research is a real great enterprise. The government invests in a total of $40 billion to $50 billion a year on those types of activities and sustains the lives of many universities. I am not even talking about the impact that the private foundations include in that. The setting there is really set up for many settings.
These particular things that I wanted to highlight before I jump into the last part of the IRB process, are just some things to highlight and you have that so I am not going to really go in there other than to say safety, privacy issues, are really at the forefront of the concern of the tribes. And not only the privacy of individuals in which we are focusing on, but the privacy of communities, of families, of nations, in that setting and those settings need to be addressed. Those lie around the respect for culture, and not to ridicule things because you don’t understand them or have never seen them.
It is really interesting because of the laws that have come into play over the years, is that a lot of culture was driven underground and we weren’t allowed to practice those types of healing practices or religious practices or information before, and with the recent laws that have just been passed, maybe a decade ago, those have resurfaced, and so presents an interesting context in which research can be addressed.
I think it would be wise also to highlight with the relationship with the Indian Health Service, as well. We were under the Department of War for a long time understandably so and then went to the Bureau of Indian Affairs and even to this day our budget is there. When talks have been discussed about the preservation of owls and others, then we have a big discussion about the health care of Indian people and the budget deliberations. We don’t mind that because we do get some funding that is tied to the US government. We know we will get money as long as there is the United States.
And the theme that I think I would highlight here as I switch into the last few minutes of the setting, is the IRB process. IRBs in the Indian community were not around. Other people made decisions on who participated. There was no consent given for all the research of the efficacy trials in the ‘70s of vaccines that we now use. We were really lucky I think as a people, that we didn’t have it to keep Tuskegee experience. We could have easily had that because the public health service was involved in the setting.
Luckily we have really altruistic physicians and nurses that really cared about the people who went out into the community and said we are going to save these babies when the infant mortality rate was 150 over 1000 births. Maternal mortality was maybe 100 times what it was for the rest of the world. And they brought it down to a setting where we are now close to the healthy people guidelines, and to do those trials where have 1000 times the rates of hepatitis and through those vaccine trials of hepatitis A, B have now gotten rates down to a setting that is the best in the world because of the utilization program.
I think the other thing to say is that the framework in which we work — we don’t eliminate the science of it. The science needs to be there or it is not any good. And the bioethics aspects of things are very good. We accept this definition because it applies to value systems which may be different, and we also adhere to the fact that the values of science cannot be superseded by the values of humanity. And people may say, what do you mean? I said we can’t be like the Nazis. They did good research, but it was wrong. And then they go okay, and then we move on.
I think the thing I would say as I move along, is these particular settings I think you are quite familiar with in terms of why we have IRBs. The thing that I would say is the differences within the IHS IRB, we have a central IRB of 15 members, of which 12 of them are Native Americans all with degrees. Community members really are important with the IRB that we have. The IHS has provided this information to you. We have 12 administrative areas, and each of those areas have IRBs so they can review materials for their local IRB. And then we have allowed the tribes to assume responsibility if they wish, and we now have about 12 very active IRBs from the tribal community.
When they look at the protocols, they look at all protocols, not just things that meet the criteria of the common rule, but they look at public health assessments and they would big debates about genetic studies. There is moratorium of all research on genetic studies on the Navajo reservation, and that was a tribal decision. And the reason was not because they did not see the benefits of it nor the expertise that people would do nor the utility that it might for the tribe as a whole. The reason they did was they did not understand what it meant. What was genetics? What was it? There are no words in the Navajo language about haploids and splits and the association and gene pools, et cetera. They have no idea what that means and they said we totally understand a little more of that. We want to put a moratorium on it and we have given them that autonomy.
The respect for a person extends to the tribe and to the community and the benefits likewise the same as well. And this particular setting is I might say is that they want this to occur.
Of all the things that I think when I work with the tribes I tell them they have the ultimate authority. Research as a whole is usually elective. You don’t need to do it. It is a good thing to do it. It is the right thing to do, but it is elective. And you could always say no. What would happen to one of us in here if somebody came in here and says I am recruiting for a new vaccine trial for a live AIDS virus? How many of you would be injected with that AIDS virus and say we have some medications that we want to give to see what the risks are? How many of you would participate in phase I trial? I would say most people won’t. You want to do phase IV. In that setting I tell them that. And we give them that right.
And I think the key is in terms of the relationship with the tribe is to have them have a say of somebody who never went to school because there is wisdom there. If a person has survived in this world for 80 years, he has some wisdom. He has some experience and often times have relationships about things which we people who have letters don’t often times understand. I can tell you the divorce rates among internists, surgeons, doctor level says that. The number of kids of my colleagues at NIH whose kids are in special institutions says that there is some problem there. They don’t know. Particularly in the aspects of not the physical, but there is research in the mental, in the social environment, and in the spiritual component. We never touched spirituality. In fact, we are by law aren’t allowed to. We sort of stay away of that though we do some studies along those settings.
What I might do is just say the data that comes is very useful. And I might say that the Indian people have contributed much. The rotavirus, hepatitis vaccines both A and B, the pneumococcal vaccine that everybody in the world enjoys today, the Pedialyte solution. All of those were researched that was done on the reservations in its initial stages. There have been benefits and the innovative system of how services research has been there.
I think what I would say is that the key has been respect of the tribe and willingness to listen and willingness to have shared decisions. I think in closing would say that it was only natural for the Indian Health Service to do that maybe because we have been mandated by law to consult with the tribes. It has only been in the last several decades where we have really taken it to heart and the tribes have been able to soar in addressing their own health care needs and reviewing the research protocol. And they have done a very good job of it in my opinion of reviewing protocol and addressing some issues, not necessarily the science, but assessing the risk and assessing the data to make sure that it was appropriately analyzed. Thank you.
DR. FRANCIS: Thank you. We now have close to an hour for discussion, which I am delighted by. I am going to just invite committee members to jump in with questions and after that staff, and hope for a wonderful conversation.
DR. TANG: I just want to open up by saying this was incredibly inspiring. I thought it was very informative, but it was inspiring because it was so clear on what the right principles are. It is such a hard subject, but it just seems like it is amazing clear after listening to you three almost what the solutions are. You all talked about the community advisory — in other words, it is for the people and by the people. The notion that — like you said, the Navajo language doesn’t even have this concept. How is it possible we could force something on them? It is the same thing. We need to educate the public on the benefits and what is in it for the community and them. And then we very much like to gain their trust. But then the other piece was, and then tell them what happens when you get it. These are all simple concepts in retrospect, but they are beautiful examples of how to deal with the hard issues.
What I also liked was when Bradley talked about the use of this for the integrity of the science. I loved the evidence behind your 9 out of 10 are debunked because that is awful scary, but also it is an important thing to know. Immediately you can do the retrospective analysis in your head and go there are lots of biases in any single system and once you combine it, it just sort of falls apart and you are often going to get stuff that is not going to be true. The scariest thing is to do research and then find that we are imposing bad things on people. That was an important piece.
The DUA saying it is not as if I just hand you a carte blanche thing. It is all part of the trusting that you really so elegantly have worked out, all three of you. And one of the things that I found very interesting about the Indian Health Service is the notion of a tribe — your comment that stuck out is we often think as privacy as an individual thing, but it is really about families and in your case nations and it is really true. But we see some examples of social networking like PatientsLikeMe where you think that is helping individuals, but actually it is disturbing the privacy of an entire segment of our population. That is sort of what we have to deal with.
I am just really struck by how clear you all have made the issues, but also how clear I think you have made a big part —
DR. FRANCIS: I will invite testifiers to comment.
DR. EDWARDS: I am really heartened by your summary, Dr. Tang. I consult with a number of research repositories, and it does seem there is some common sense and core, who is really returning to our roots of our ethical foundations here. Some of these solutions, I think, are straightforward and yet not often done. I think we have had our head down just trying to move forward so we have lost our way a bit.
So things like let’s add a community advisory board. How can we assure that this is for the people by the people? How can we close the loop and let people know what we have done with the data that we asked them for? I think those are very simple things, and yet remarkably not done in a big stretch of our research work.
Frankly, I think it is because we have had an over reliance on IRBs and regulatory bodies. People say I have sign off from the IRB and so my obligations are over. I think really what we are asking for here is how do we go — that they have set the floor for our behavior and really how do we go beyond that. We can actually go further if we invest and if we take these small investments in the integrity of the system. I appreciate your reflection.
MS. KLOSS: Could I ask a follow up question really of each of you? To empower the communities or to reinforce the power and the authority that they have I would think some training or guidance or general principles or something, must help enable them. I think it would be useful to us in framing more generalizable recommendations if you could share how you would open the eyes of a community to the value of a community engagement. And what principles you would encourage them to use in really all three of the applications?
DR. SMITH: I think I can give you maybe our perspective from the Indian Health Service. I think the setting is time meaning the investment of time in the engagement, not having just a snapshot meeting and having decisions made on things that might have lasting impact, but have a sustained relationship in the setting where you are collecting vital health statistics. This is for the long haul over decades looking for trending. The interactions and constant feedback over decades of time develops the trust that really can really enhance the relationship.
And then quickly in terms of assessing with input from the community who speaks for the community and what is the community. I think we have many papers on that and say, who is the community. It is very complex in the United States because we are a melting pot. But I will tell you one thing is that there is a real collection down at the bottom of the mixing bowl where you will find that is a community just as much as the things that are floating at the top using the metaphor of the melting pot with the mixing of all the people. But you will I think is that within a community you will see individuals that come forward and those are the individuals who just don’t show up for one meeting, but are there constantly for a decade in concern about the setting and willing to be the interface with the community.
In our tribal communities we often times have that in the leadership of the council that usually is an older individual who often times may not have, what do you call it, the demeanor that is expected in government bureaucratic setting where efficiency is highly valuable and valued and effectiveness is not meaning that you want to have the meeting from 10 to 10:15 and we want to be done and they want to say if this is important, let’s spend a few days on this. And that gets in the way. To remember that sometimes to be effective you need to be less efficient in terms of the management teachings that Peterson and other gurus from the major institutions have told you on how to manage a meeting and to be very open and just follow your grandmother’s advice and say sometimes you just need to listen and not say anything.
I am being very colloquial and I suspect my colleagues will be more prolific in their vocabulary. But my feeling about it is trust is a very basic human element and we learn that as children with your mom when you are born and those things you cannot forget just because you are now the director of an agency or handling something very complex that only you and a few other people understand.
In the management of privacy I think that is true because we don’t see the full ramification that this committee deals with. We only see the outputs sometimes and now knowing how much effort was put to try and address the protections and occasionally something is missed or to decide what is most valuable in terms of collection.
DR. FRANCIS: Others want to take that up on principles that we might want to be aware of for training and guidance? I was just asking if the other two — maybe I can push you or push a question. Go ahead and respond. I wasn’t sure you wanted to.
DR. EDWARDS: There is so much to say on this topic. Leslie, did you have a focus you wanted us to —
DR. FRANCIS: One focus I was going to ask actually and then you can speak more generally to is that in some ways the Havasupai example is potentially interestingly different from some of the other community examples. For one thing it is a very small tribe. For another thing I think it is fairly geographically condensed and isolated. And of course it has a governance structure, a tribal governance structure, and that is not true of a lot of communities.
My specific guidance question was going to be about how you think through in much more diverse settings. But actually I was going to ask Dr. Smith on the other side in highly concentrated maybe not so diverse settings who should be speaking for the community, how you get on community advisory boards. You actually gave an answer in terms of who has been active. But activists may not be always the people that we want to have speaking for communities. They may just be the squeakiest wheels. I would like very much to hear general guidance, but also what are your specific thoughts about who the community is and who should speak for the community in community engagement through advisory boards or whatever other mechanism it would be.
DR. SMITH: I have some thoughts on it. I know that both Bradley and Kelly have others as well. I think what I would say is that you do have those environments with the urban Indian programs where you have that diversity. Very complex where you might have 80 different tribal members coming from different settings and yet they are a cohesive group mainly because they are Indian and they feel a component of that and they want somebody to speak for them because they are a body that is a band in Los Angeles or whatever or Detroit or Chicago. That is a trend now is where people are congregating there.
I think what I would say is this. The first principle I think would be is in terms of transparency is to be very broad in terms of sharing the information of your intent of what you are trying to do and the data that you are collecting. That casts a wide net over individuals that you might potentially work with.
And then secondly is to do a little bit of the accountability. Letting them know that you as an entity are not totally accountable but is a shared accountability with the community as a whole that when they make decisions they have to be accountable for that as well and sometimes those decisions might be wrong. Sometimes they may be on target. But there is a shared accountability and that setting is to be able to be inclusive I guess is a good way to say it.
And then third was in terms of once those communications have been done is when you do the feedback is to let everyone know who had the input and who had the concern. I think that is when things don’t go well. People are accepting of that because there was a real good effort in terms of addressing that.
And then third was to discuss the future. It is sort of like a hope of what this is going to do and to talk about the benefits of this interaction not necessarily saying we are collecting this information because somebody said there were disparities. And you focus on the negative so much particularly on very sensitive subjects about AIDS, sexually transmitted diseases, et cetera. Nobody wants to come forth and say this community has got 10 times the rates of anyone else. But the idea is having this information to say we can address it for the future. I think when the community is looking at the benefits not for themselves as well but for their children and others. I see people coming out of the wood work wanting to help and be engaged.
DR. MALIN: I am going to lay out a little more of a process model here. Before you can say who do you need to talk to, you first have to know what your internal policies and procedures are. One of the first things that we learned at Vanderbilt was that the community needed to have somebody to talk to. You can’t just say we have this new project or we are going to be doing research with all this data and then everybody goes who am I supposed to talk to about what is going on and how do I get out of this or how do I participate. You need that focal point. A lot of times it is called the ombudsman. You have the person who is representing the organization of the figure head of the organization that will interact with the external world and act on their behalf. If not acting on their behalf, at least talk with them. You have to have these types of policies and procedures set up.
Secondly, when we went out to figure out who were the leaders that we wanted on the Community Advisory Board, this happened in two ways. First, we talked to our constituents. We talked to the patients and we said who do you listen to. What are the organizations that you participate in? Where do you believe we may not have properly represented the community? We had a whole series of town hall like forums where we interacted with the local community that were a member of our — for the most part we already patients. Not everybody was already a patient because we didn’t want it completely biased. There are people who continually get referred into our hospital.
It is kind of like an election level. It is kind of like a system in which you say if you do not participate, if you don’t show up to vote, you don’t have any say in the democratic process. We are going to let you know that this is what is going on. We are going to give you the opportunity to step in. If we haven’t taken into account your views, you have the opportunity to present them. Basically what I am saying is that we ask people what organizations they are members of, but we also say it is an open call to participate in these types of forums. It has to go both ways because you won’t cover everything otherwise.
The other thing that I wanted to say about how do you find these leaders is that you want people who are willing to serve for short terms. They have term limits because you don’t want the advisory boards to be dominated by people that have their own views. You have to be careful about rotating and having these policies set in place as well.
I also wanted to reiterate the accountability aspect, but I wanted to put it in a different light which is liability. One of the things that Mayo found and I hope I don’t speak out of turn here. One of the things that Mayo has found was that when they talked with their patients about who do you trust to do research — they basically asked them questions. Do you trust the federal government? Not really. Do you trust Mayo? They said yes. We trust Mayo. And then they said do you trust the federal government if Mayo gives them your data. And they said yes, I think we trust them. Wait a minute. What just happened? There is this chain. There is this accountability chain.
And basically if something went wrong, they wanted to know to a certain extent that they could hold Mayo accountable for their actions. But they also wanted to know that Mayo was going to act as a credentialing body for them. They said I don’t have the ability to determine what the federal government is doing is right or not. But you have much greater interactions with them. If you know what is going on and I can trust you, I may not need to know everything that is going on, but at least you are giving me some piece of mind. And this is where you have to recognize that the community has limited bandwidth. They deal with their daily lives. They are not constantly thinking about health research. Instead they want to know if there is somebody who is looking out for their best interest. The trust that you build up with them is important because it allows them to offload some of that responsibility and they feel better if you don’t violate that trust. That is why I say it is a process model to a certain extent.
I will say one more thing. It is very easy to allow patients to be negative and to allow people who have negative voices to get up and speak. It is much more difficult to encourage people who feel strongly about the positives. And I don’t know why this is. I do not know the psychology behind it, but we recognize is that we have to be very careful about not letting discussions in our community engagement settings turn towards things that have a negative stint. It is not like that we are trying to obscure what is going on in that side. It is just that we have recognized that there is a little bit more — it is easier to be vocal on that side.
DR. EDWARDS: I agree with everything my colleagues are saying. I think there are a couple of things that I want to underscore are I think people in the United States tend to say we don’t like the idea of someone speaking for us. This idea of how on earth do you get a representative body, a community advisory board that is going to represent the interests of the American people. And yet we do this in our democracy. We elect officials to speak to act in our best interest. I think as Brad is suggesting it needs to be a both end kind of solution. We do need those invested people who can do their best to work on our behalf and work in our best interest. And you do need the outlets for the individuals who are never going to see that kind of control to others. Clear opt-out models, clear 1-800 numbers to get answers to questions, clear websites that describe what is happening with the collection and the resource so people can discover information if they want it.
I had a one-on-one opportunity with Twyla Breeze, who is one of our more vocal privacy advocates. She was even clear on if a newborn screening collection had — if the opt-out mechanism was easy enough, transparent enough, and clear enough, that that would be sufficient for her. The fact that all of this was buried and shrouded and just wasn’t even clear what the uses were was part of the offense from her point of view. If we can just make this as transparent as possible and give people options, we can meet both those individual needs as well as the collected needs.
To press a little bit on the basic education — where do we begin with education to help connect to the kind of research we are doing? I would say the principles there are making it relevant. What is going to make it relevant for people and help connect to something in their life? It might be a health issue that they are facing or their community is facing.
A very personal story about how research has led to advances in a particular issue in their community. A very human story is about the researchers. What is a day in the life of a public health analyst as far as what do you hope to — what do you use the data for and how does your discovery work from including setbacks and including ways forward. We found stories both community stories as well as research stories really help humanize the whole process and give us a connection a way to see that this is relevant in my life and then I can start to care about it. We have to want to find out more information.
And then certainly our K12 science education is a foundation for all of this in terms of what is the role of public health research and translational science. Just how does that work. And how does such a long discovery path of research. How do we help people — most of us get our science information from the news headlines. How do we help convey that 9 out of 10 associations end up being not what we had hoped? How do we help prepare people for that kind of scientific reality?
DR. SMITH: Can I just maybe just give an example of the interactions that we have had that have proven to be successful. This is a collaborative project with the National Institutes of Health. It is a very unique program. Because of the civil rights laws of course, they are really restrictive in terms of research that they may be able to do. You can’t let out a grant specifically to study Chinese or Hispanics. But with the relationship that you have the federal government has with the American Indian and Alaskan Natives we do have some provisions that are very race based. That is why the Indian Health Service exists.
Within that context they let out a grant specifically directed toward American Indian tribes and they called it the Native American Research Centers for Health. And the framework was to allow the tribes to come forth and be the center of the setting and have research institutions that are skilled in the science of research form a partnership and the funding go to the tribe and then the university subcontract. That relationship made it such that they have to talk because the money even the researcher have to come from the tribes and the tribes were engaged in developing these grants that were then sent to NIH and reviewed along with 50,000 applications from Harvard, Stanford, et cetera and funded and evaluated on the same basis.
We have a project with Vanderbilt, for example, with the United Southeastern Tribes and very effective in terms of that relationship that has really been built up. One of the key elements was that most of the research institutions are far away from our communities. The way that they have addressed that is to have a local present. And the elements of contacting individuals about problems with the research — it is very difficult to call an 800 number from the middle of the Navajo reservation and you are talking a graduate student that this is not a consequence of research, et cetera. You can imagine maybe an unexpected event had occurred and you are trying to determine if it was part of the research or not. But it is good to have somebody local that they could just go over.
And the way the system has worked is training the local tribal members in research methods and taking the youth and putting them on tracks to get masters of public health to consider working on their doctorate as they move along. And because of the program through the National Institute of General Medical Science, who has these training programs like MARC, COBRA, and INBRE, et cetera, which many of your institutions have students at using those settings to get their doctorate and have postdoctoral training.
It has allowed some funding to come and these are institutions, as I said, like Vanderbilt, Johns Hopkins, Harvard, Arizona and Colorado. You do have these really skill settings. University of Washington has been very successful in working with the tribes and developing these programs that have a training program. And these individuals go back to the community and tell them about data, talking about epidemiology, how statistics come out, just simple arithmetic, how statistics doesn’t work. You have to use biometric measures. And why you can’t do research on 12 people in a tribe that you need a larger number. And talk about power.
And what has been very surprising to many people is that they have been able to incorporate that knowledge and make informed decisions on things. Like I said for me the IRB true test if it is functioning well is if you deny some research. To me that has been the test because it seems like when you evaluate the IRBs around the country, you will find that their approval rates is near 100 percent. There might be a good design, but there might be an ethical concern. Most university IRBs will not say no because of that. They want to find a protocol. To me I think the training can be done.
DR. FRANCIS: I want to ask other committee members who would like to jump in with questions.
DR. MAYS: I want to thank you all for just incredibly rich and wonderful presentations. I was thrilled to come and thrilled to be here given the nature of what we are discussing.
One of the things I wanted to talk about and I saw it very specifically in almost all your talks, but in your presentation, particularly Kelly, is about the issue of relationships and how important relationships are and I think that is probably all we have been talking about.
One of the issues that is starting to emerge and I think we are struggling with this in terms of community data and some of this comes from — I happen to be editing a special issue on the legacy of Tuskegee. And one of the papers came in talking about the nature of relationships that exist in communities. That whole notion of who represents whom and who speaks for who is really emerging as being very critical. The case really is that. Sometimes you have hierarchical relationships. Sometimes you have gender dynamics.
I am wondering if you could talk a little about that because I think the big case that came out about this was the millennium pharmaceuticals that were working with Harvard in which they went into a small province in China and collected something like 60 million DNA samples. In that collection it was like the rate of agreement to participate was something like 95 percent. And when they looked it was like a postdoc who had grown up in that small village, went on behalf of Harvard. And the saying yes and the participation had a lot to do with the nature of relationships within communities. Now this is really becoming an issue that even IRBs are trying to think about in terms of protection. We think having the community be the forefront is actually the most positive.
And now they are asking that we start think about whether a person from a community going into the community kind of being the advocate if there is enough distance from the nature of relationships that they can actually give consent. Could you talk a little bit about that because you have been doing great with this notion of what to do and this is now emerging as a really big issue?
DR. EDWARDS: Thank you for bringing that case forward because that is the flip side of it, isn’t it? I think the concept I might go to is the one Brad mentioned is the chain of trust. IRBs are very sensitive to this issue. They actually are concerned that community advocates can be — it is on the radar for being coercive. They believe so fiercely in their cause whether it is a disease-specific advocacy group or for a community that they would just go out and really could potentially be coercive to getting people into the study simply because of the commitment of their belief in it and the personal connection to the community. How do we protect against that?
In that chain of trust if the people of Rochester, Minnesota are going to trust Mayo with their data including giving it away — we will trust you to give it away to someone else. That means just like the people of China trusting that postdoc who was from there. We are going to trust you if you think these people are good people and are going to have our best interest in mind then we will give you our data. But I think it is on the integrity of that postdoc to say can I genuinely stand with that being accountable for the good will of all these people that they just entrusted with me. Is that in fact true? Is this research team in fact working on their best interest and how is that true? I don’t know the details of the study to know whether it was or wasn’t true. But perhaps advocating for an extra what are going to give back to this community to make sure that we have closed the loop there.
If Mayo does give the data away to the federal government, are they assuring — they are now the trust brokers. How do they assure that that is a positive thing to do? It is not a direct answer to your question. I just think it is partly — I have had some researchers who want to use community members to say you are from that neighborhood. Why don’t you go ask them? I have had a number of community members say that is my good will. I am not going to spend my good will in the neighborhood that way because I am not sure what we are doing here. I think we need the community. There is a lot of integrity that is needed to say —
One of my Native Hawaiian colleagues calls it the ohana factor. If she makes a misstep with her community, she is not just outcast from the community. Her children are outcast. Her children’s children are outcast. The stakes are high for her if she makes a misstep in her community. I would trust her to go do that recruitment because she is not going to talk to anyone into something that that is not in their best interest.
DR. SMITH: In my other life I worked as a sociologist to a certain extent. One of the reasons why I think Kelly maybe tiptoeing around the answer is because there is no silver bullet on this one.
One thing we know about communities and organizations in general is that there are all different types of structures. There are cellular types of networks. There are small world networks where you have the 6-degree separation type of model. And knowing who the right people are that represent each of these sub-networks or types of networks is not only easy to figure out. Usually you call these the ambassadors of these networks. And sometimes they are not who you expect them to be.
For instance, when we were establishing the community advisory board, we thought that it would be one of the pastors, for instance, of a very large church. It turned out that it wasn’t the pastor who was actually the important person, but it was actually one of the people within the congregation. We didn’t know this until we started unraveling, not so much the chain of command, but the chain of influence.
What I am saying is that determining who the right people are is sometimes a research project unto itself. How do you actually walk into that community and establish trust with them? This is also a really tricky thing because different communities value different things. You have to figure out what exactly or who exactly — no, it is what exactly — the concepts are that they value in terms of saying is it about being a member of the community. Are you coming from within the community? Are you an educated person of letters that acts on behalf of the community? Things like that and different communities have different belief structures in that way.
There is a whole series of research in I would say sociology and organizational behavioral modeling that gets at this because we have seen in terms of — I will allude to it again — in electoral systems. When different campaigns want to figure out how to reach a community to determine how to get people to vote for their candidate or just to get a message out there, they have established — for instance, the DNC and the RNC both have established very nice statistical models to determine who they are supposed to be interacting with to get messages out there and how to get information flow moving efficiently. I would just say take a look at some of these.
DR. EDWARDS: You asked the question in the context of the Tuskegee story and isn’t one of the examples there is nurse Rivers’ role. We have identified exactly who the most influential person is. What would have protected her in that story might have been having more of a backup for her of how to check out whether this an appropriate study to be recruiting people.
DR. SMITH: Just a side comment on that. One of the things that a couple of the researchers that was done and this has been discussed in the primer is that this human-to-human contact in terms of getting consent is the element you are talking about. And to then go to the use of technology in which you are taking the human face-to-face contact on there and the comparison of consent by responding to a computer as opposed to an individual and seeing if there was a difference. And the idea was to look to see if there was understanding of what they were engaged in because most of the time our problem with consent is that when we go back and talk to the volunteers, they have no idea what they are actually really participating into.
I laugh about it because I have engaged that even in a personal manner because a couple of my children have participated in some projects at NIH mainly because they wanted to and I said do you know what they are doing. And they said no. And I said they are studying diabetes in the Indians and I am an Indian. And besides they are giving me $200 and taking my blood, type of thing. And I am thinking (?) -and all these things are coming out.
And then I said you need to read the study because they may be wanting you to do things that you don’t want to. And then they go oh, they told me the risk was minimum, it is like walking across a street in downtown Silver Spring. And I said that is pretty dangerous at that type of person-to-coast contact. And what they then did was put it on the computer system.
Now when I go and talk with people who have done that, they know what it is mainly because they are into the computer system and we have tested out with individuals because it is a touch screen and it is a vision and a voice. The person doesn’t need to read to touch the screen. I think that might be something that might be helpful in addressing concerns if you think that there has been too close of contact.
Our problem is almost all of our research projects involve the local community being involved, but most of them don’t do the consent. We have the consent being done by another person, not the community member that is there. We have somebody who comes in because the feeling is they are consenting to an activity. They are not consenting to a trust which sometimes in nurse Rivers was the case, they really trust them. I think it is an issue like you said has no silver bullet.
MS. MILAM: Your discussions really have been inspiring. I have two questions I was hoping each of you would react to. And the first one that comes to mind is that your approach to community involvement is really exceptional. You can hear how important both the individual and the community are to you, along with the research. I have seen other researchers who don’t always approach it that way. How do you incentivize that approach to merely getting through the process to get to the research? What makes that important to you all that could be shared or included so that it would be important to all?
DR. EDWARDS: I think that is such a great question and I was in fact wanting to turn this back out to you all too, because I know a lot of the community health data that is under your consideration isn’t the kind of data that you would be having consent for at all. This is data that already exists in existing collections. There really isn’t even this face-to-face opportunity for understanding.
What we are talking about is even more challenging of motivating a retrospective sort of how do we now go back and start talking to people about what we would like to do or what we are doing with these existing collections. How do we incentivize that?
I think there is both the carrot and the stick approach. I think the stick has been for better or worse I think the Texas newborn screening story did some of us a favor who care about these things where there is the real cost. There are five million blood spots that literally got destroyed because of the — not because they did anything illegal, but because sufficient people were concerned about what was happening and felt cut out of the process. What is at stake for us here? What is at stake is literally our ability to the collections themselves. We want them to stay. And we want to be able to utilize these collections and we want to be able to utilize them to maximal purposes.
I think starting with we don’t want to lose this and then working towards the carrot of what could actually be gained. I watched this happen just individual by individual in terms of when they go to have a community meeting or go to a consultation or watch an engagement activity. They actually are reconnected with why they went into research in the first place. We spent so much time behind our computer and desk that we know we went into research to help people or to make an impact in public health. But it has been a while since we have had that interpersonal contact. Sometimes just getting back into a community and remembering why this is important and how it can help can literally refocus and re-motivate and help give meaning back into the work. I really have seen that happen. You can’t script it. You can’t plan for it, but it is each person’s own transformation to reconnect with why they are here. But that can happen in those —
DR. SMITH: It is very difficult to, how do you say it, provide incentives when you are dealing in a subject that is very linear in the process. You are doing it in environment that is three dimensional. And what I mean by that is research is very — a scientific model is you have a thesis of a question and you have a really rigid standard that you follow in addressing or making adjustments for biases, et cetera, and you are moving toward a decision based on the analysis of the data that you have and the instruments that you have biometric measures and then you make an interpretation and for the most part success is when you have a peer review publication or a continuation of an RO1 grant. Often times there is a little more of a dragging force than the thousand lives you saved because you often times don’t know about that over a period of time.
And what I mean by three-dimensional setting — that is the life we live in. That is the getting up in the morning, getting the kids to school, and going to work, and looking out from the sky lab at night and see if you can see it. And these are all life settings.
To me I think it starts really early in terms of this inquisitive spirit. It is sort of like rejuvenating the American mantra I guess is a good way to say it, but I see where people engage in these types of studies where they are doing secondary view of data. And when you see about it, it is something that they did really early in their high school setting or in their center that continues.
And then secondly as I think and as a federal employee I probably shouldn’t say this, but I think there needs to be advocates from the community itself to talk about the funding for issues that are not purely by medical and in the new frontier of science, but dealing with how services research, dealing with why this group is able to be very resilient in a setting where these other three groups are not. Yet all of the social dimensions appear to be the same and providing some support for that. And for us in the Indian health arena we are very happy that CDC is around because that is their realm and often times we have big debates of whether something is research or whether it is public health assessment. To the Indian community there is no difference. There is no difference whatsoever.
DR. FRANCIS: I want to ask others to jump in who have questions. Jack, did I see your hand?
PARTICIPANT: We didn’t hear from Brad though, on this question.
DR. MALIN: I have been thinking about how to answer this question. In some respects it is a loaded question. What I will say is that there are some investigators. There are some researchers that absolutely do not want to have any type of a relationship with the individuals from whom data was collected. They want to be hands off. They don’t think that that is their responsibility to act in an ethical manner. They don’t want to think about the ethics, not that they don’t act ethically. They just don’t want to think about it.
In some ways you can’t pull everyone kicking and screaming in front of the patients or in front of the individuals from whom the data was collected and say you must talk with these people. You must act ethically. You must interact with them. And this is where there are filters. We don’t have an unstructured interaction between scientists and communities. There is some type of a buffer. There are people that act on behalf of the community because there are some people in the community that probably shouldn’t be interacting with the scientists. And then there are some scientists that you probably don’t want to be putting in front of the community. Establishing that policy and procedure that is I think where you get the best bang for your buck.
It is kind of like a telling process. You want that board or you want that group to be cognizant of the research that is going on within the institution. They have to be. If you don’t know how to report back to the community with the type of research that is going, you have lost them because then things get published and they go I wasn’t aware of this. And at the same time you want to have a way of filtering the relationship to the community because not everything that is said by each person in the community should be getting back to the investigators. Every institution, I think, sets that policy or in the oversight board up differently.
I think what has worked for us is having people who are — there are several scientists on our board. There are ethicists that are members of the medical ethics community for the institution as well as administrators for the medical center and university in general just so that they are aware of what is going on and recognize how these relationships exist. But it doesn’t become always interaction.
The other thing I will say about this is that I think there is — this is where I could get in trouble. There probably should be more training for scientists on ethics. I will say that even in our own curriculum, even in the informatics curriculum we do not have an ethics course that is required for all of our students let alone their investigators. The investigators have to show that they are familiar with the Belmont Report. They have to get IRB certification of human subjects training. But at the same time it begins in school at that point.
And if you don’t get the appropriate training, you don’t get shown that beneficence and proper interaction and respect with the patient community is the right thing to do. It is not something that is going to happen once you start working as a PI on a particular project. It does come back to taking a step back and possibly changing the workforce a little bit. That is a really challenging thing to do. And I am not saying it has to be done, but it is probably something that if more institutions took that up, it would be better in the long run.
MR. BURKE: Thank you very much for your presentations. Not a question, but an observation. Bradley hit it right on the head. I am concerned that as we develop standards for good stewardship of data to do this work we need to focus on the research participant, but more importantly on the person conducting the research whether it is the principal investigator or someone engaged by that person. The transactional, mechanical acceptance of consent is only limited to that piece of work. And to develop the trust you really have to go beyond that. The ethical dimension of seeking and receiving consent is doing more than you are required to do. It is not the mechanical checking the box, I uttered these remarks, I heard this answer, and here is the signature. It is how do you invest in creating that trust whether it is for research or use of cumulative data in either direction.
DR. MALIN: Let me jump in on this one. This is where I think we have to be really careful as a society about how we open up repositories of data for research purposes. One of the things that I advocate is that there has to be a way of establishing a relationship with the investigator. What do I mean by this? There have been proposals to take datasets and make them full public use all the time even though we recognize there are potential risks to the underlying individuals. And our argument is that those risks are small even though we may not know what all those risks are at the present time.
If you establish that relationship with the investigator such that you know who is doing the research study, you may be able to have some credentialing body that says what goes on in their institution and are they willing to certify that it is all kosher. You are in a better situation than just letting the data fly out the door and creating one gigantic citizen science environment.
I know that what I have just said actually flies in the face of what a lot of people would argue for right now. But at the same time not all people are good. I hate to say this. Some people like to do demonstration projects just to show that there are vulnerabilities in a system just to point out that you can’t protect the system always which is a dangerous thing to do. Having accountability for who is doing what is a nice thing to start with.
The other thing that I will say is you also hit the nail right on the head in that the chain of trust goes from the patient to the institution that has collected the information if there is one to the investigator to everybody that works within the investigator’s team. And at that point that investigator is I guess credentialing all the people that work with him or her. I don’t think that there is a simple way of giving a guarantee that each of these people are going to be completely trustworthy. Rather it is due diligence.
It is a matter of just trying to keep the investigator and their team vigilant with respect to how information is being used, how it is being disseminated, and that the policies and procedures that are set up within an institution are being followed and that requires audits. That is time consuming. You have to be careful about how you do audits. But at the same time if we don’t have given any guarantees you do have to take a step back and do some investigations.
MR. BURKE: The observation that you made is valid. The missteps that we have seen in conducting research have typically been traced back to the research assistant working on behalf of the principal investigator. While the research assistant is responsible for the misstep, it is the principal investigator who is accountable.
DR. FRANCIS: We have about 5 minutes left. I am going to invite Kelly to make the comment that she did. Then I am going to invite anybody to raise very briefly a question that they haven’t had a chance to put out on the table. And then I am going to ask each of our speakers to answer any of those or say anything that they haven’t had a chance to say, but that they wanted to say. Kelly, quick answer.
DR. EDWARDS: Quickly, we are supposed to be beyond data use agreements, but I do think your question gets at the heart of it. I think too many of us click through. We are too used to internet click through agreements that we don’t really read what they say when you update your iTunes. People have talked about using video and more interactive approaches to help research participants understand what they are getting into. I have been advocating for more video slowing down the data use agreement on the data access side for researchers so they understand what they are getting into and agreeing to. How do we make that not just a click through process, but really a meaningful — I understand what I am taking in responsibility for by taking this data. It is one place.
I think anybody that owns data or is stewarding data and giving other people access to it you have a lot of opportunity to set requirements for your users including dissemination plans. You could have free and open access data repository, but require application for that use so that you know who is using it and require that everyone using it has a dissemination strategy in mind for what they find which could include giving findings back to you or to their local community or whatever their intention was.
This is another response to you, Sally. Researchers will do whatever a funder or a data accessing group tells them to do. And if we have to come up with a dissemination plan, we will start working with a communications group in our institution to try to figure out how to meet that requirement. I think there are different things we could build into the structures without making them onerous. I think we have to help people meet those obligations, but to help really frame the kind of research environment we want to create.
DR. FRANCIS: Other burning questions?
MS. MILAM: I think we have heard a lot about the important role of community participation and research. When I think about data, I think of a spectrum. On one end we have data collection from the patient, specimens from the patient, clearly research, clearly IRB approval. I would like you to take me to the other end of these large anonymized legally de-identified datasets, perhaps not an actuality. If Brad had his hands on it, he probably could do a lot of things with it. But if you could take me there, would your thoughts about community participation apply in that situation as well both for people in government and those using government data?
DR. FRANCIS: Anyone else? Vicky, quickly and then I will turn it back to the panelists.
DR. MAYS: I want to hear your thoughts about whether or not you think we should change the CD training in order to include some of the community data issues.
DR. FRANCIS: Why don’t we just go down the line?
DR. EDWARDS: I think the city training should be updated to include more of these community health database issues. I think this is one of those challenges where that training has been narrowly focused around the regulator issues which are individual protections and identifiable patient information. But I think this does connect to your point, Sally, which is — and my interest in our broader public education. I think we have so much opportunity to help people understand health and science if they existed of these government collections. Just the fact that they exist, what they exist for, what their opportunities for use and contributing to public health insights are there for. I think this is just a real public education opportunity. Perhaps look at it as an opportunity and a responsibility more than an obligation. It is not that kind of regulatory obligation, but it really is an opportunity.
DR. MALIN: Short answer to Sally’s question is yes. The moderately long answer is that we do not have sufficient transparency in the use of de-identified data in this country. And that is problematic because people have absolutely no idea how the information gets shared or gets used. And I am actually kind of happy that Sue is not here to hear me say it, but I am sure she will hear, is that this is something that I believe OCR should be looking into. I don’t necessarily believe that all patients should be given the opportunity to opt out of all possible studies because it runs into bias issues. But you should definitely be transparent about the research that is being done with this type of information. That is clear. And right now there is no federal requirement for doing so. Once the data is de-identified according to HIPAA, it is gone. It just falls off a cliff.
The second thing I wanted to say really quickly and I will paraphrase Bill Stead for this one, which is you start small, you fail small, and then you succeed large which is when you do secondary analyses or you are going to do broad data collection for secondary analysis at the end of the day, it is a good thing to do a test run. It is a good thing to do a test run to get the kinks out. When we started working with the community in terms of telling them we were going to be collecting specimens, collecting data, and then doing research, we did this for 6 months even though we actually didn’t collect anything. And we just wanted to see how people responded and we gave them the opportunity to opt out at that time and it worked out a lot of the kinks when we actually went live.
DR. SMITH: I think what I would like to do is answer the two questions. The answer is yes. And really support that with a caveat that to be broad in terms of selection of teachers or the contributors and not to be so focused on people of letters, but to be broad in terms of having inclusions of the community members who may have experience in dealing with research because they have been trained and that had gone through the city course. An individual from the community we would require them to do that. And they become very good bioethicists even though they are not on faculty at some school. They are really true blue bioethicists and to provide that.
And then secondly is on the issue that you brought up. For us it is a very sensitive subject mainly because the tribes are very concerned about the data which we collect often times without them knowing it. We have tremendous profiles in the millions of biomedical samples that have been collected over the past several centuries that are housed in government settings so you can do genetic studies if you want to of any communities that existed centuries ago. They have a real concern about that.
And then maybe last in closing is that it is very easy to develop policies and procedures today because what I have seen people do is they just look around to see what everybody else is doing, grab it, put it on the system, and replace the institution with their own institution and then the problem arises when the application of those come about. And my feeling is these need to be developed the same like strategic plans, over time. You have to have that discussion and see if there is buy in of not only the faculty, but the postdocs and staff and whatnot.
To me I would say about that is we still need approval for those de-identified. It involves the tribe. We ask for their permission to use it. I have to say many of them have said no. And then what we tell people is here is another tribe that might work with you. They have taken their project elsewhere.
DR. FRANCIS: I hate to cut this off. This has been wonderfully rich. I want to thank the three of you for an enormously helpful discussion. Thanks so much for being here. I apologize that we ran over a few minutes. We are now going to take a 10-minute break.
MS. KLOSS: It is the sign of a good hearing that we have this much dialogue spilling right over to break and just continuing into our panel number two which we have focused on considerations of protecting of small groups in community use of data. We certainly understand that some of the concepts we have begun to explore will be continued and built on and we certainly look forward to that.
We have two panelists in panel two: Dr. Paul Spicer from the University of Oklahoma and Dr. Villegas from the National Congress of American Indians. While they are in reverse alphabetic order on the agenda, we are going to flip because they have agreed that Dr. Spicer’s comments will be a very good set up for Dr. Villegas.
DR. SPICER: Well thank you very much for the opportunity to come and testify. I don’t think I have ever testified.
MS. KLOSS: Now you know it is a conversation.
DR. SPICER: I can do that I think. I was struck today on this morning’s session by the emphasis on relationships and I think that is what I came here to talk about as well. I have participated in multiple rounds of meta-reflection on community engagement and research in the context of genomic science. I have had three LC grants. The most current one is from the human microbiome project where we are working both in Peru and in Western Oklahoma to understand Native communities, hopes and fears about genomic knowledge and the opportunities for finding a way forward.
We had a gathering of all the LC investigators back in March at Baylor College of Medicine to talk about issues that have emerged in the context of the human microbiome project. I was on a panel about informed consent and we didn’t talk about the document at all. We talked about the importance of the relationship and this was true for me certainly working in Native communities, but this was true for any of the other investigators there as well. We were all drawn to I think what is a lack of knowledge, a lack of scholarship, a lack of articulation about how we should attend to those dynamics. I think the bottom line as emerged this morning I think to my mind is that relationships are key.
We have theories of individual relationships. I have worked in developmental science. I am very interested in the relationship between parents and children or grandchildren and their grandparents. But I think what we lack is an ability to scale that thinking to a larger relationship, to relationships with public, to relationships between institutions, to relationships within organizations. I think there is a clear need for a lot of thinking there.
I especially liked the comments this morning about the need to think about institutional arrangements because I think that those are key and I know that Malia has a lot to say about what is emerging in Indian country in that regard.
I think the big danger that I have been made aware of in community engagement is the danger of overselling the knowledge. We want communities to participate in our studies. We are excited often times about the prospects of the value of our data, the value of our knowledge. I don’t think that anybody would disagree with me that in genomics we have had very high levels of disappointment. Most of what we have hoped for in genomic knowledge has been — we have failed to realize.
I think that is good from the perspective of Native communities that have been fearful of what genetic knowledge might mean. I think that is very disappointing for a health care system from a public health perspective in terms of what we have been able to act upon with that knowledge.
When we first started our community engagement around genomic knowledge in 2000 in the wake of the demise of the Human Genome Diversity Project we tried to focus on medical genetic research and struggled mightily. It involved everybody we knew who was working on these issues in alcoholism and diabetes, in heart disease to come and talk to us about the potential values of that knowledge and the persistent response from the Native communities that we brought to the table as well was that this has been meaningless. This is a waste of our precious resources and we have much more direct things that we can act upon and should act upon.
I think it is pretty clear that when we are talking about Native communities, we are talking about communities where you are not maintaining trust, where you have to create that trust, where you have to regain that trust because there is all the reason in the world to be suspicious of anybody who is coming in with the next big data projects that is going to give us the answer to this, that, or the other problem.
I think the joint management of data is key here. We need repositories that are jointly managed with tribal governments, with tribal authorities, and with researchers who have the capacity to use that data and manage that data.
But I think the biggest problem, and this emerged again this morning, is with dissemination. It is what we do with the data afterwards. And in our discussions with communities we have been pretty good. Those of us who work with tribes have to work with tribal governments, need authorization to do our research at least those of us who do that legally. We have been pretty attentive to getting approval upfront. We have been much less attentive to giving those results back. I think that is the biggest problem.
What I think is something that we should all be thinking about, are models of community coauthorships. We have review and approval in most of the studies I do with tribes. My publications have to clear a tribal review before they get submitted to the journal. And increasingly now, as a way of dealing with that process we have involved communities in writing in articulating that knowledge and list them literally as authors. That runs against some of our grievance around community confidentiality which is something I talked about in the statement I prepared for you.
I think that kind of community participation in the dissemination is really one major antidote to thinking about some of these problems. And I think they are relatively easy to imagine because I think communities have perspectives on the data, ways of interpreting the data that are unique and valuable in articulating scientific knowledge.
I am with Kelly though this morning in thinking that scaling from my individual relationship with a tribal government in the context of a particular study to what you do in the state dataset or in a national dataset is tricky to imagine. It requires some creativity. And Kelly really emphasized this morning the importance of public outreach and engagement and education. And again the danger there is overselling the knowledge. I don’t think we want to do that. But we can walk a fine line. We can talk about why science is interesting. We can talk about why this data might matter. We wouldn’t be acting with ethics as scientists if we talked about how this data will matter because otherwise we wouldn’t need to the study if we already knew it mattered. We are testing hypotheses and we need to keep that spirit in terms of our outreach and engagement with the public.
But I think the challenge there is in thinking about what those relationships are. What are relationships between institutions? What is the affect in those kinds of connections? Because certainly the relationships I have with individual tribes, with individual tribal members are individual relationships between me. They trust me to do what I say I will do. They don’t necessarily trust the University of Oklahoma or my former employee the University of Colorado and in our survey work as well they certainly don’t trust the federal government to honor my agreements. It works for me as an individual, but how will that work for the University of Oklahoma? How will that work for the State of Oklahoma? How will that work for the CDC?
I think we need to think carefully about if we want to build up from trust, which is an interpersonal dynamic I think fundamentally, to institutions and organizational issues. We need to think carefully about that scientifically in terms of the social and behavioral sciences, the human sciences. But I think we also need to think carefully about that ethically.
One thing I didn’t hear this morning which I think is important for us to entertain in this context is that we need an ethics of relationships. We have an ethics of duty. We have an ethics of rights. But we don’t really except with one exception I am about to tell you about that you probably are somewhat familiar with. We don’t really have an ethics about what happens between those of us who have duties and the folks in our studies who have rights. What is the nature of that relationship? There is the ethics of care, which is articulated largely within the context of feminist scholarship and bioethics, but that really attend fundamentally, to those relational dynamics and recognize our relational nature.
And I think that there is a lot there that we can build on as we think about getting beyond my duties as a scientist or your rights as a participant in my study, and thinking about what we owe each other and what those mutual obligations might be. I will just leave my comments there.
DR. VILLEGAS: Good morning everyone. I want to acknowledge that we are meeting today on the traditional lands of the Piscataway people of this region. My name is Malia Villegas and I am Alutiiq/Sugpiaq with family from the Islands of Kodiak and Afognak in Alaska and O’ahu and Lana’i in Hawaii. It is a real honor to be invited and to be here today.
I currently serve as the director of the Policy Research Center at the National Congress of American Indians. NCAI is the oldest, largest, most representative organizations serving in the broad interests of 566 federally-recognized tribes, state-recognized tribes, and American Indian and Alaska Native citizens. The Policy Research Center was founded in 2003 and our mission is to provide tribal leaders with the best knowledge and information to guide their decision making within a framework of Native knowledge to inform the best futures for our peoples.
Toward that end research regulation is a huge aspect of our work, a huge part of what we try to support tribal leaders and communities in developing processes around. In particular, we have developed a research curriculum that we are in process of preparing to travel out to three tribes in Montana in the month of May to talk about how you set up research regulation, oversight processes which could include tribal IRBs, but are not required on that level.
Today what I want to share with you — I have prepared some comments and they are much longer than what I will talk to today but I will just give you some of the broad strokes. The comments were really drawn from other comments that we submitted to OHRP on relieving researcher burden and really hit on the complexities of doing this work in tribal communities. And for us in this work there are at least two anchors. I think as Dr. Spicer and I am sure Dr. Smith alluded to there is a long and troubled history of research in Native communities. There is a real need to focus on the ethic and I really appreciate this ethic responsibility and relationship that you talked about, Dr. Spicer, there.
I think the counterpoint to that also is that we recognize that we don’t want to be dis-included from research and there is a real diversity of engagement and experience in communities as you can imagine at least 566 with the federally-recognized tribes at that level. In addition to an ethical approach we really want to make sure that we are supporting meaningful research and that includes both benefit, and a recognition that there is deep innovation to be had in work that is being done by tribes and with tribes that can inform the work happening at a federal and regional level. Similarly a lot of our work hinges around both enforcement and accountability efforts with federal and university partners, but also finding ways to not just encourage and support Native people at participating, but at engaging and driving research themselves.
In particular I think it is always important for that history to come into the room when we talk about these. I am just going to read for a moment. I think the most recent example that we need to be assured to have these kinds of conversations is what happened with the Havasupai Tribe in the State of Arizona.
In February 2004 the tribe filed a lawsuit charging that researchers from Arizona State University misused blood samples taken from tribal members. The tribe claimed that tribal members were told their blood samples would be used for a study on the genetics of diabetes. However, the samples were also used for studies on schizophrenia, inbreeding, and possible migration patterns of the tribe’s ancestors from Asia to America. The case was recently settled out of court, but I think it is an important reminder that these are real present day issues. This is not just a part of our history. This is a part of what happens every day in communities and we want to acknowledge that space at the same time that we see that tribes are actively continuing to engage in research and how do we support that ethic and that meaningfulness of the work that they
The comments that we have put together today, on the second page of the packet that we prepared, I am just going to give you the five points of recommendations about what we suggest when we are looking at protection of small populations, but also as I said engagement in a meaningful and ethical way. The first point that we emphasize, point A, is continuing and close review of research with American Indian and Alaska Native tribes and peoples by IRBs. We understand there is an expedited process that is often used by research regulation boards, but particularly in the case of Havasupai and some others. If there was a situation where the review was continued and ongoing at least at a 2-year mark is what we recommend, but those where there is a more than minimal we recommend at least annual review and ongoing review of the research proposals that way.
More specifically the second point there is we think that it is very important to continue to provide oversight of and protection over informed consent processes. And I know you have heard quite a bit from Dr. Smith with his extensive work on this topic. Specifically here we have three points. One, where instances of informed consent must consider both the individual and tribal consent. A lot of the Belmont principles are critical and essential to the work that we do as researchers. However, they often do not account for community and tribal consent. How this plays out in some of our recommendations is that we advocate to not support standardized or blanket consent forms. We think that they need to be specific to the research project and to the community in that sense.
We also recommend that options be given to research participants and to tribes about how that data will be used. This is particular to the secondary use of data issue. That second bullet point there. But ultimately ensuring that waiving consent is not an option particularly when you are looking at biospecimen work given the concerns over how that has been used in the past and the meaning that human body parts and tissues have to many tribal communities. Ultimately though it is about advocating for a tribal consultation process and research, a tribal role not just to say will you participate or will you not. It is how can the tribal perspective be engaged in the design development, the why question that Dr. Smith put on the table earlier here.
This is an issue I think not just for those who are new to the context, but those who have been working in indigenous communities for a while. We have recently been made aware of several situations where well-meaning researchers coming from a particular ethic have developed agreements to ensure confidentiality of individual participants as well as to provide the data back to the tribe. But given the small size, given the context it is very hard to do both. To ensure confidentiality and to provide a dataset back to a tribe that you are protecting both the individual story and data points, but giving the tribe a set of data that they can actually use. This is a point where we had the set of values around tribal consultation and tribal role, but we recognize that there are real complexities when it comes to the impact and to actually doing this work.
I mentioned the secondary uses of data collected. We have some recommendations in there really around consent, and then of the de-identification of data, some of the complexities around that. I think we are well aware of issues around group stigma and stigmatization, but I want to emphasize that it is not just about what people outside of a community come to know and understand in terms of what is reported, but it is about the harm that can be waged internal within a community particularly around some of the genetics research, for example, looking at research that might tell a different story about where a group of people come from or the lineage relationships between them, could have some really detrimental impacts in terms of how people come to know who they are as people of a community on that front. I think it is important to look at both group stigmas outside and group harm within.
The third point there is really about ensuring that we include tribal IRBs and other research review processes in health research review regulations recognizing that tribal IRBs will have a much more local and particular sense. This includes support for establishing infrastructure around tribal IRBs, but also supporting tribal members to serve on IRBs that might be set up at a regional or federal level as well as not requiring or supporting the idea that multi-site IRBs are needed in comparative work in indigenous communities recognizing that might place additional burden on a researcher to manage multiple IRBs, but pushing for, advocating for one model, one body making decisions for multiple can be problematic when people from those communities don’t have an opportunity to engage the decision making rather than just the participation. I am hoping you are hearing that in what I am offering.
The fourth point there is this piece that at the end of the day it is really about getting us to coordinate in better ways around research regulation. We recognize there are very complex issues. While you can again take in a set of values when you come to doing the work, there is a way perhaps not to do work. But in terms of how to find a pathway through it is often you need the partnerships and you need perspectives from various levels especially including from the people who are participating in that.
What we would suggest here is a move towards a single guidance document. There are some great models in Canada around work with Aboriginal communities particularly in the public health domain that have been useful for us in the work that we do. I would be happy to provide some links towards that front, but that uphold a set of values and ways of doing this work, but also provides some guidance to Native and non-Native researchers trying to make sure that the meaningfulness of the work comes through and that tribal sovereignty at the end of the day is affirmed.
Finally, this piece here around — one of the key aspects of protecting small populations we believe is making efforts to include American Indian and Alaska Native peoples in public health research. And that includes data collection efforts. We are very concerned that we with the National Childhood Study, American Indian and Alaska Native subsample, was ultimately taken off the table because of the difficulties of including a sample due to our remoteness and our size. We realized that there are challenges towards that research design methodology. We are seeking alternative research methodologies that recognize the issues facing national bodies including American Indians and Alaska Natives and other small populations, but that also allow us to develop research that is meaningful.
We refer to ourselves often as the asterisk nation in that you go into federal report and you go down the line of racial and ethnic groups you get to our line and there is an asterisk because we are too small and that leads to our invisibility and our inability to participate in large-scale conversations. We want to be included in national large-scale datasets as well as to support rich, deep, contextual qualitative work in communities. It is both and kind of space here. We have some ideas.
Dr. Rubidoux of the Indian Health Service has recently released — not released the report — we are still trying to get our hands on a report that embraced translation science methodologies and 66 demonstration projects. I put a little bit of information that I have been able to glean that was able to demonstrate similar validity and strength of findings to an RCT model. We are looking at translation much like you talked about dissemination as a way to look at alternative designs and a longitudinal piece.
The final bit here is about data reporting. I want to draw your attention if you are not already aware to the OMB 1997 regulations around race and ethnicity. I haven’t been able to find what the HHS guidance is on how that guidance is going to be implemented, but I know for a fact that the Department of Ed, for example — this is about Hispanic or non-Hispanic as an ethnic category and then a minimum of five categories of race which includes American Indian and Alaska Native, Asian, African American, and so on.
Our concern is with the Department of Ed, for example. They are collecting data on all those fronts. But what they are reporting is if one of our students or family members selected Hispanic and American Indian and Alaska Native, they would only report that as Hispanic. If they chose non-Hispanic, but selected American Indian and another racial category, they would only be reported as mixed race. It is not a matter of not having the data. They are continuing to collect the data, but it is the way that they are implementing the reporting that is challenging our data quality and access issues. I just want to draw your attention and pose a question really to the committee to look at what is HHS’ guidance on how that data is going to be collected and reported and to request that even when it is not significant, even when it is too small of a population that that data be released. We need data in Indian country and we are committed to data quality on that front.
And the last point really is in order to do that we are committed to encouraging our people to go on science careers. And we think that a big part of protecting small populations is in supporting more Native people into the sciences. NIH recently released a report on the race and ethnicity of investigators and their success at achieving NIH grants. And we are looking at those instances and trying to find ways to work with bodies such as you to clear the pathway and to make sure that we are at the table in these decision making —
I will leave it that to say that we have a few other examples in here from data enclaves to the Alaska Native specimen bank that shows examples of how communities are trying to do the both and work here of protection honoring that ethic and supporting meaningful engagement with research. Thank you for your time.
MS. KLOSS: Thank you so much. We have a good amount of time again for discussion. I encourage questions. Is there anyone who wants to start out?
DR. TANG: Thank you again for the testimony. I think it continues the illumination of some of these issues that we started with in the first panel. What I heard was — and it was mentioned — expanded on the notion of there is individual consent, but there is tribal consent. Your examples just bring it so much to light, but we can extrapolate to say really each person actually doesn’t even own their own data. It was part of the family thing, but also we are all parts of small communities especially in America that we need to both recognize the benefits of doing research that includes small communities as well as what you emphasized of the protection.
I think the main theme that running this whole morning is if only we involved people and that is that is both the governance, the participation, and in your last point which is the training and education so that you are part of both sides. I think that is the answer. It is not easy to accomplish, but thank you so much for helping us think that through.
DR. MAYS: Again, thank you very much for very insightful presentations. I have a question and then a comment. I want to make sure I understand something. In terms of multi-site studies in American Indian populations part of why they went to having one place be the keeper of the consent and the forms, et cetera was in order to move this along. I want to make sure I understand. Is what you are saying is that it really should still have some consent activity that is done in all the different sites or it should — I am not sure I understand that well enough. If you could help me with that?
And then the other some of the research particularly in terms of American Indians is done in areas where there are large urban populations. For example, I live in LA, New York, places like that. I would like to hear what you think should be responsibilities when we are doing community data. They are being included, but yet there is not a tribal entity that always is available without going back to maybe 15 different tribes so that I can make sure we are doing the right things.
DR. VILLEGAS: Thank you for the question. To clarify in the multi-site and I have some more specific information on page 6 and 7 of the comments. What we are recommending is that first of all that there should not be a requirement for only one IRB of record for multi-site studies imposed. If the tribe who will be participating in that data will be reported at that tribal level chooses to acknowledge another body to oversee the decision making on that then we think that that is absolutely their prerogative. That is part of our role in supporting tribal decision making and sovereignty on the front. We don’t encourage or discourage either way, but we definitely want to identify instances and models of that for researchers and communities looking at different approaches there.
At the same time I will say that in the first month of my time in this role as we were preparing these comments for OHRP I got a call from a dear friend and colleague who is an indigenous researcher here in the states saying that she was working on a multi-site tribal project. We talked a bit about these comments. And I have known this woman for 15 years. I know her family. I know where she is coming from in terms of ethic. I know what she is trying to do. We both share a belief in the need for comparative work tribe-to-tribe, nation-to-nation, and native-to-native kinds of strengths-based approach. She just continued to explain her desire to do that, but the struggle with managing 15 tribal IRBs on this front. That is to say that we realize that there is not one way to set recommendations on this front.
We are actively engaging in the complexity and are excited about the potential thinking with you folks about how we do this in a way that gets that comparative work out there that gets the rich data done in the right way at the same time that we acknowledge because we know it takes time. We know it takes money. We know it takes that particular piece.
We have put these recommendations in place to really first and foremost honor that tribal consultation and opportunity for tribes to determine. And I would say that that carries forward in terms of the second response to the question about urban as well is it depends on what is going to be reported. If it is going to reported in American Indian, Alaska Native, and urban community that is that piece. If as a researcher or a partner to a project want to report it out as Navajo people in an urban area, you really do need to be ensuring that consultation with tribes, but also really what that process looks like and that identity and how you can make the claims that you are making about whether or not that is a product of being a member of a tribal nation versus a different population that way.
I think the urban question is a huge one for us. It is very complex. We are working with the National Council on Urban Indian Health, NCUIH, to look at different research designs and how to guide researchers working in this domain. I don’t know that I have any more specifics to offer, but I think the reporting question is one that we are talking at this point.
DR. FRANCIS: I actually want to pick up on reporting out and co-authorship because — one of the things that I heard in the last panel and here is that there should be — that it is really important to get important information back to people closing the loop that way. But there may be very different views about what information should be conveyed, how it should be conveyed, whether it should be public or in some limited way shared with the group. There may even be disagreements about whether to publish or to report out. You suggested a co-authorship bottle as one possible solution to that.
I have heard a lot of discussion about getting the community engaged in defining the research questions or setting the public health questions, but I haven’t heard a lot of discussion about how to negotiate the conflicts that might emerge. I don’t even have any good examples. The Havasupai wasn’t a conflict about whether to publish it. It was a conflict about the use of the data. I would like to hear you all address that question if you have any thoughts about when their intentions about — one of the principles we might adopt or recommend or something like that is there should be reporting out, but how to think about negotiating conflicts about reporting out or publishing. Anything you have to say about that I would love to hear.
DR. SPICER: Since I started it I can start the answer. I proposed co-authorship as a more collegial, more respectful collaborative arrangement with communities in terms of articulating knowledge about them. We already agree in multiple tribal settings that tribes have the ability to approve what we publish before we publish. There is already a distinction between the quality and depth of the information that we give back to the community for their own planning and programming and what we would ever publish anyway. There are already things that we are not publishing based on work we have done in communities because there isn’t really a good way to do that. Certainly in the case of me as an ethnographer I know the people I am collecting the data from. There is a bunch of things I know about that I would never say in print or in a presentation about them. There is already that kind of protection in terms of disclosure or potential protection. There is already that kind of option available to researchers in terms of deciding what to report out.
You are right, I think, that colleagues don’t always agree. Co-authorship isn’t like magic dust that you wave over a collaboration, and suddenly we agree on how to present things. Fortunately in the cases that I have been engaged in we have agreed. I think the real challenge is if you have an investigator who says I need to publish this. This is important for the field. If you can’t agree then I am going to publish it without you. And that certainly is an option that we as colleagues have more generally. I would suggest that is not an option you have working in a tribal setting. And if you can’t agree on how to say something then you need to move on.
There are multiple findings in the course of any study that never see the light of day. Some of us publish with abandon, but most of us are pretty slow. There are lots of findings that are in file drawers and this would be an example of one of those. That would be my default assumption is that if we can’t agree then we don’t go forward with that particular publication.
The one instance that has come up that is a difficult to discuss in detail had to do with the sensitive issue of native spirituality, ceremony, religion, but it was a broader issue. It wasn’t about providing details about specific ceremonies. We didn’t even ask those questions. That was not a part of what we were approved to do. But people often feel that spirituality is important in healing from the kind of psychiatric issues that I generally work on. And the question was whether to call it religion or spirituality. And in sociology it is of course religion. We have a defined body of knowledge and the sociology and the author herself Native wanted to call it religion. And the community was insisting on calling it spirituality because they had other associations with religion that they didn’t want brought into that context. And in the end we were able to agree and we called it both. We had a footnote indicating that sociologists call it religion and we are calling it spirituality. We were able to move forward.
I think in general the emphasis I would place is on the importance of dialogue and on attempting to work it out. And the respect of our tribal collaborators if we can’t work it out that we are just going to move on to a different publication. I have ten that I meaning to write so I can move on to number eight or seven or six, but that is from an individual investigator’s perspective. But I think already there certainly is a fundamental distinction between what we give back to communities and what we would say in public.
DR. VILLEGAS: I would add a lot of the conversations that we are having with tribal leaders and communities particularly around this research regulation curriculum raises questions about enforcement and we are trying to reframe that, but just to give you a sense. Talking about research design, talking about reporting, talking about the whole process of research with tribal communities. A lot of questions come up to the point where what if you disagree. What happens in that instance? What authority do we as tribes have to prevent that from getting out there, to shape it? What does that look like?
And what we have been able to say at this point is if you haven’t been involved in the early stages, you haven’t set out those partnership kinds of parameters and even legal parameters about what that looks like. You really can only appeal to funding agencies who might be funding that work or to the professional ethic bodies and communities that way. We try to advocate for tribes to get involved very early on in terms of the partnership but also to begin to develop and design their own research and to select partners with them in that authority position to look at partnerships from a different angle. Again, tribes not just as participants in the research but in the lead on that. I think there is a lot around what is the authority when we look at decision making.
The other piece is what do we mean by publish, and what forums are we talking about. I think you raised the term about dissemination. We are working on a project with NIDDK in Saint Louis University around diabetes-related translation research and looking at health literacy, looking at health communication. This idea that from discovery to delivery 17 years is not good enough in many of our communities in any of our communities. How do we work on prevention, on behavior, on getting the information out to have that more immediate impact in ways? I think it is both around — I think co-authorship is really important when we are talking peer review, when we are talking about those kinds of forms.
But when we are looking at other types of publications and ways of getting information out, webinars, a lot of our folks are — many don’t realize. We are online. We are on the internet. There are ways to get information out in different aspects. I think, again, re-centering where that impact conversation happens the publications are critical for academic fields and others. But there are other journals and other mechanisms to get that out there. That is something we are looking at. How do we develop different approaches to dissemination and development that are meaningful?
DR. FRANCIS: Just a follow up to both of you. What if there is disagreement within a group where some — I am asking this partly because I really appreciate the relationality points. It may not be possible to respect some people’s desires not to have the information be known and other people’s desires to have the information be known because of the relational implications. That is sometimes true within families, for example, of genetic studies, but outside the genetic context. If I know that there are toxics in my community, I know something about the toxic exposures of my neighbors. Do you have thoughts about how to handle intra-group disagreement given the relational context?
DR. VILLEGAS: I think it is a really important question especially the work around tribal communities. There is often a sense that there is one kind of perspective as with any human community rich diversity. What we tend to again prioritize at NCAI is formal tribal processes of decision making. I think this is an important point because I situate myself as a researcher, but I know over time there have been outside researchers, researchers who do not from the communities who form relationships with one set of families, for example, and may develop a set of knowledge and insights from that piece from those more informal relationships that way and then begin to make claims about a whole community without recognizing that there is a wealth of diversity and perspective on that front. That is where we would always defer to that tribal council level, to that formal decision making body to be involved in the early days on the design and decision making and dissemination on that end.
I think the other example I would say — I know you have referenced genetics. We are preparing to release a genetics resource guide at NCAI based on the interest of tribal leaders at how to navigate, researchers coming in and wanting to do this work with them. In the Southwest we are seeing huge increases in childhood in leukemia and the opportunities for pharmacogenetics to address some of the issues on that front. There is a real range of perspectives on whether genetics is something tribal communities want to participate in or want to know more about. Yet we see our position as again deferring always to the tribal leaders and the tribal councils to say, we are not advocating or saying don’t participate. Here is a set of information about how you begin a community process. That goes back to that kind of formal decision making authority to try to navigate some of that.
There are going to be some instances where that process will not result in one definitive sense of direction. And I think that that is the piece when you cannot disagree. I think I would echo that you need to defer to that tribal council’s authority on that front to say we are not quite ready to have that information out there on that front.
DR. SPICER: I think in that sense working in tribes is relatively easy because you do have the formal authority that ultimately I think you defer to in that context and they have to answer to those different constituencies in their community. We need to be aware that that is part of what is being mediating for us in those discussions with the formal authorities.
I think it is trickier in communities that lack that kind of formal organization be that an urban American Indian community or be that any of the other communities that we are involved in. But I think there are strategies in terms of how you move forward, how you present, what that represents. In the case of a tribe if you are talking about this tribe’s perspective on an issue, you are talking about the tribe as a unity. In the case of almost any other community and in the case of many tribes you are not talking about that kind of unity. You would be talking about a particular segment of that community, the segment that elected to participate in your study. There are ways to make it clear that you are not speaking for the Native people of Los Angeles, for example. You are speaking for this set of families or this neighborhood. Those kinds of approaches I think make it clear the limits of your ability to claim to speak for a community.
MS. MILAM: Taking a step back just to the small group issue, what advice can you give us very particular to the risks involved with a small group in terms of the principles of notice and consent? What have you learned along the way that you think would be useful to others working with other types of small groups?
DR. SPICER: I will just jump in because I think you pose it in terms of risk. I think the other comment that Malia made earlier so I am not going to answer your question directly at least, but the other comment that Malia made earlier is in terms of the possibilities, the potential of this knowledge for small groups and the fact that in multiple national studies in American Indian and Alaska Native communities are systematically set aside so that we don’t generate knowledge that is useful for them. I think on the one hand we should be certainly sensitive to the risks of that information. On the other hand, we should be sensitive to the fact that we don’t engage in developing any of that information to even know what the risks might be and that is I think the persistent frustration in tribal communities.
In the work I do in Head Start, for example, the National Indian Head Start Director’s Association advocated for years for research that would inform American Indian and Alaska Native Head Start. They finally got it in the reauthorization of the Head Start Act and our center was charged with developing some of that knowledge.
But I think that we are very wary of engaging certainly tribes in these kinds of national studies and as a consequence we don’t know what is going on in tribes. I can assure you that what is going on in small tribal communities is unique and offers us potential to think about issues in some new ways that we are simply not getting to.
Some of the challenges in working in small communities that Malia commented on earlier in her remarks, is the fact that often times tribes now want that data provided back to them. They assert their ownership of the data. And so researchers do provide that data back and there is a danger then that people can find people that they have something against in that data, and can say something about that. Certainly that is one of the risks there.
One of the standards that has been operative in American Indian and Alaska Native health research for years is community confidentiality. So that we don’t identify the specific community in which we have done this work which protects that specific community, but as I said in my remarks may just create a more general stigma.
One of the designations we have used in Colorado for a very long time is Northern Plains. There are a lot of tribes in the Northern Plains. That includes the Aberdeen area. That includes the Billings area. At a minimum it probably includes parts of the Bemidji area as well depending on how you define it. These are service areas of the Indian Health Service that Dr. Smith talked about this morning.
On the one hand that particular tribal community isn’t going to be stigmatized and injured necessarily. But then we perpetuate a broader stereotype. It is one of the reasons that Eric Youngst(?) threw his hands up on the notion of community consent several years ago and had been advocated in the context of ELSI, was that where do you stop consulting a community. Maybe you shouldn’t stop. If I follow through on some of Leslie’s comments about the idea that you don’t know if you have exhausted community opinion on an issue, you shouldn’t say you have consulted a community until you have talked to everybody in the community. There are certain practical issues that I think put a limit to that.
But I think that certainly in American Indian and Alaska Native research we are very sensitive to the risks of particular communities and try and protect them. But that does raise challenges then in terms of how useful that information is to them or to anybody else, and it certainly raises issues about the diversity between tribes at a minimum that I think we want to talk about.
But I think the more general comment is that I think we have a lot of thoughts about what the risks might be. But because we are so wary of engaging and actually doing the work, as a society that we don’t really know what the risks are and we certainly don’t know what the possibilities are as they unfold in practice.
DR. VILLEGAS: No, I definitely agree. I think there is a lot that we are learning in tribal communities that can be made beneficial. It is really the comments that I put together in pages 3 and 4, about how do we address some of these risks. I think one of the recommendations that we have is around not — that standardized general consent should not be used, blanket consent processes. It needs to be specific to that research project and options need to be provided for research participants on informed consent forms.
My mother was recently asked to participate in a menopausal study. There was just a note that this data would be — she was having to sign off on this data being available and used for other studies down the road. We had a long conversation about secondary use of data and what are her options. She said there is no box here for me to check that I don’t want that to happen. It was a very standardized form. I was able to look over it with her, and to push a back a bit with the researchers on that front.
So this idea that you just have that standardized form I think is real problematic and something that we need to look at and set higher expectations for our researchers around that.
I think the other pieces are when we are talking about biospecimen and tissue-related work that there needs to be definitely no waiving of consent. Waiving consent should not be permissible under any circumstances for research involving collection and study of existing data and biospecimens. That is our recommendation on that front — Havasupai piece there. But that secondary use of data should actually require additional consent.
And then we have some recommendations here also about how that data that has been collected is shared within the team and with other teams. We talk about leftover tissue and blood, things that aren’t really seen as essential to the study and just throw aways, but have that as still a part of the human body and it has meaning to many communities outside. We need to be specific and transparent about recommendations and requirements on that and that limited datasets should not be shared outside the original research team without asking for additional permission.
NIH has a really interesting model around the data enclave about how to actually set parameters for folks to get access to data to get more of these findings out there, but in a responsible way not just saying that is not important. We have already got the sign off, but going through that extra work to make sure that that consent has been put through.
And then we provide a little bit of an example on the secondary use of data on page 5 around what I mentioned, the Alaska Area Specimen Bank, and the process that they have used which I would say is kind of a more complex data enclave of sorts where they have engaged Alaska Native tribal and community leaders and organizations to make decisions about how that data is accessed. But it is a very rich data source. Again, we want good research to be supported in that way. But I think there are some real examples that we are trying to get at the complexities that these tribes are trying to address that could be useful — populations on that front.
DR. SPICER: The major risk to small communities in research again relates to how it gets reported. As I said, there are a variety of ways in which you can protect communities that have some cost to themselves in terms of the utility of that information, but that maybe the kind of protection you want to put in place.
I think the major risk to individuals within small communities is that people will not like the fact that you have participated in the study that says something about the community and that you may suffer in terms of your relationships and your standing in the community. And those are risks that I think we tend to underestimate probably in any community, but certainly in these communities were uniquely sensitized to the consequences to an individual participant for having done something that the rest of the community didn’t approve of.
Leslie talked about this idea of genetic exceptionalism. I do think that much of the issues that we grapple with in genetics are not unique to genetics. But one thing that genetics has really done for us as foreground is interconnectedness. None of what we do is just about us anymore and us to attend to that in any population.
DR. TANG: Just a further question on the secondary use. The theory behind HIPAA is if you quote de-identified in the HIPAA way that everyone loses control over the re-use of the de-identified dataset. I think what you are saying is if it is used in ways that the contributor does not expect that is at least to me an ethical problem. It may not be a legal problem at this point. And my test for myself is if the individual would be surprised to find out that their data was part of such and such. That is a problem.
I think that is what you are saying and I wonder what arguments you use about the fact that is “de-identified” because I haven’t been able to convince people that that is an issue. It is the re-use of “de-identified” data. Is that an issue? If so, how do you convince others that that is an issue?
DR. VILLEGAS: I think in terms of HIPAA, I think there are definitely — that we affirm the use of those standards first and foremost on that front. I think the issue of de-identification I agree. I think it is an ethical challenge. I see it from two sides. One is that consent — what was consented initially for what was collected and where the decision making lies ultimately and how that then ultimately gets used.
DR. TANG: I will try to paint the scenario and then you can comment. You gave me an example. Tissue is one thing because it is ultimately self-describing — de-identify it. But once you turn something you gave for a purpose into in theory de-identified then in theory it no longer represents just you. Do our arguments about meeting the individual’s choice persist once — that is the question.
DR. SPICER: I think Kelly said something this morning that was really informative in this regard is that we have made it safe to use the data, but now we have lost the connection of the data to the individuals and that is one of our foremost bioethical principles is that we should be attending to the needs of the participants in our studies. On the one hand, it has made it clean and easy for us to run with the data.
On the other hand then we have severed its connection to the living people that provided it and that raises ethical issues, not legal issues probably I think. I am not attorney. I know we have tons around the table. I will defer to them on that. But it certainly raises ethical issues about our ability to do anything with that information to give it back to the people who provided it.
DR. VILLEGAS: I think that was where I was trying to get to is the re-identification of the data. I am not sure it actually addresses your concern about the de-identification. We believe that there should be a prohibition — there should not be an absolute prohibition from re-identifying datasets that were previously de-identified and this is about being able to return that data to the communities to the people and in terms of thinking about impact. I think for me at least it is a little more complex than just the de-identification and thinking about whom that belongs to and where that is going forward. There is actually more concern for us on the re-identification and the limits on doing that because I think there are some instances where it makes sense to — we need to know where that data was collected, how it was collected, and in order to inform the impact that it is going to make.
I am not sure that that actually answers your question about de-identification and we were sign over for. I agree that is not the individual level there. I am not sure I have a good answer to that.
MS. BERNSTEIN: I just want to poke around there a little bit on re-identification. When I think about it and some of the comments in my policy capacity I made when NRPM came around, one of my concerns has been that under the HIPAA rule that, I am not speaking for the department, that once you de-identify data under HIPAA, immediately at that point someone can begin to re-identify it with impunity for any purpose.
The purpose that you are talking about is a purpose that some people would agree with that there is this duty to the community that we don’t want to sever that connection. We might want to go back and do more research that extends what we have done before or report back to the community and so forth.
But there are other purposes. You could re-identify data that are more nefarious. And there is no distinction at the moment in the rule. Once something is de-identified under the rule, it is just not covered by the rule anymore. And even though the purpose of the de-identification was not to be able — to have not unidentifiable data, you could immediately begin to use whatever resources are at your disposal to re-identify it and there is basically nothing the department can do about it unless there was something in the data use agreement or something. I think it is interesting to think about the purpose behind the re-identification which is really what you are getting to.
MS. KLOSS: I think Leslie has a question, Marjorie and Vickie.
MS. GREENBERG: I am just wondering from the discussion that we have been having and about given the importance of the tribal identity as well as the individual identity and the sovereign nature of the tribes, et cetera whether this whole concept of de-identification even makes sense in that context. It is because if you are studying people in a tribe and that is known then maybe the individuals are not identifiable, but that identity — I don’t know the extent because this is a big issue obviously with HIPAA and with the whole issue of data disclosure about identifiability. What is your position on that or what is your thinking on that? It just seems to me it doesn’t have the same meaning.
DR. VILLEGAS: I think you reference the division, the separation of a person or a community from the data itself. I go back to the question that Dr. Smith left us with which is we often talk about the what and the how in research, but it comes back to the why and that is what I was trying to get at in terms of the impact of this work, why we collect it initially. For me it is about that deep, local contextual work is needed. In that sense you would want to know not so much the individual person perhaps, but how the claims are going to be made about that context and that community.
There is another piece though around the participation in large-scale national datasets. I think you were asking about non-tissue, non-blood related aspects on that front. Where we want to make sure we are not the asterisk that we are getting data reported out even when it is of a small enough level. We also want to know aspects about our community that may be not be related to our race and ethnicity or our political status in that sense. I am still struggling with this question. But I agree that you are putting on the table that maybe it is not a question about de-identification, but the purpose, the role, the ultimate impact.
MS. KLOSS: Vickie, is your question pertaining to this topic?
DR. SPICER: Let me just jump in real quick on that because I think the standard protection that we would add that normally goes into a de-identified dataset is you wouldn’t identify the tribe and that is usually the protection that we give them. You just have some vague, geographic designations or in national datasets it is just American Indian and Alaska Native. But there are considerable costs in terms of data quality and your ability to draw inferences that are relevant to tribes when you do that. And I think that is the point about de-identified data more broadly. It certainly becomes useless to the participants.
MS. KLOSS: We will go to Paul and then Maya.
DR. TANG: I think this has just been such an extraordinary conversation because I think it has brought us ahead of our times of where we are now. In a sense we are all part of small communities. It is just a matter of time. This happens to be prototypic way ahead of its time. Instead of letting us trying to once again try to catch up when we figure out what small communities we are part of, maybe we really need to deal with this front and center. But I even see you struggling with the well you want to do good plus you want to protect. I don’t know where we are — we certainly haven’t found the solution, but I am sure where we are headed outside of education involvement and governance. And what is interesting about this example is there is a formal federal law prescribing governance. We literally have to deal with the governance of the tribes. Maybe we should be dealing with some way of governance of all of us small communities which we just don’t know where we are yet.
MS. BERNSTEIN: As I was mentioned before, the privacy act advocate of the department, I think about individual level privacy for the most part. This is stretching my thinking in a good way. I am thinking about now the conflict between the confidentiality of the individuals and the problem of wanting to identify small groups either — there were two sides of this. You want to be able to report data on small groups that may either identify the group itself or end up identifying the individuals because they are associated with that group, but also we want to have — there is a risk there, but we also want to have the information about these groups.
I am trying to figure out how do we think about that kind of conflict. Are there examples that you came across where there was some way that you dealt with that? I was at OMB when we — my colleagues down the hall were putting together the race and ethnicity stuff. As Paul was saying, I am thinking about what small groups am I a member of. We want to identify those people that are half-Ashkenazian, half-Italian, Jewish, Sephardic, whatever. Eventually we will get to those and we will figure out something about me or we have the BRCA1 gene or whatever that is prevalent in Ashkenazi Jews or something like that. I don’t usually think about the positive side of wanting to report out the small communities.
I sit on the data council at the department. This is a problem that they talk about regularly. How are we going to — essentially one solution to the problem is if we had a lot of resources that we could over sample Native populations of various sorts and we could then have enough people to report on in which case the cell size wouldn’t be so small that you would want to protect it. But it is a huge resource problem to get each one of those respondents, cost a lot of money. It is not that we don’t think it is worthwhile doing it, but resources are scarce in an era of declining budgets. They are cutting — Marjorie is going you are making my arguments for me for my agency NCHS. But we are hearing this regularly every year. How are we going to make priorities for maybe we should do our data every other year? Maybe we should make the sample size smaller. We don’t want to do that because we are coming up with this issue of wanting to be able to report out.
I am wondering if there are particular examples where you have successfully dealt with this conflict, there were agreements made that yes, we are going to — some people may get identified, but we have a larger issue that the tribe wants the data or decided not to do that because we needed to protect the individuals.
I had a conversation in the preparation for this and I am going on too long. I am sorry. I had a conversation in the preparation of this hearing where some of you I may have spoken to, some people who were working for the tribal epidemiology centers which I admit I didn’t even know existed before I was planning this. And they talked about how the tribes own their data and we often send the tribes back their data, but don’t identify them publicly which is what you were talking about. And yet I am thinking if you do that then you have the problem of maybe identifying the individual if the tribe or the band is small enough which I hadn’t thought about. I am wondering if there are recent examples where you had to deal with this or how you might imagine dealing with this.
DR. VILLEGAS: We are actually in the midst of dealing with this. We have a federally funded project and we are the lead organization on this and we are working with two university partners on some case site work with a tribe, again, trying to set this out. I am about 6 months into this position. I inherited this a bit, not to say that I wouldn’t have run into this problem myself, but where we set out a data use agreement, set out an MOU, our university partner and a tribe about returning that data back. The data that is being collected is individual interviews of tribal members about a particular community health intervention. But there was no — now having transcripts and getting ready to present that back, give that back to the tribe. So much of it has to be blacked out because it is identifiable. It is specific to that person.
There also was an issue with the consent form in terms of what was told to the individual participant about what would be done with that data would be reported and returned back to the tribe. We are going to have to go through reissue and consent of those individuals as part of that member checking, returning that data so they can be reminded of what they said in that instance.
PARTICIPANT: So many researchers’ nightmares, I have a feeling. The time and resources to do that.
DR. VILLEGAS: We are in the midst of it. I don’t know that we have a better solution on that front, but just to engage and to try to put as much of this information on the table on the early end.
But I also want to say as you have put the question on the table about is de-identification — is there a different way of coming at that? I want to put the question on the table about confidentiality anonymity. I think that that is in a lot of ways one of the pillars of our field in what we do. But I know that I have been a part of studies where in negotiating with tribal communities with American Indian and Alaska Native peoples, elders in particular, who have said I want you to put my name on this. I want you to name me in this because the information that I am going to give you, my community, my children, my grandchildren, I want them to know this. Rather than going in and assuming that people want to give away their right to be named, I think we need to think in different ways about that impact and how being named. We definitely don’t want to do no harm. That is one of our principles. But how do we get to that do good kind of space — is there a way.
I think part of it is about our elements around reporting. Reporting to whom? To what end here? I have a dear friend, a Navajo researcher, who has been working with the Cherokee nation for a while at their request. They have specifically asked her to come in and do this work. She has 5 years’ worth of data around early childhood that she has been working with. I just saw her up at Vancouver. I said, how is that going. How is the work coming? She said it is phenomenal. It is done. I said are you publishing. What is the story? She said no. That is their data. I have returned it all back to them. They are going to let me know when they want to do something with that towards that end. That is their audience, their community. This idea again about publishing which audience and I think that raises the questions about reporting and confidentiality and some different ways as well.
DR. SPICER: I think the other question you raised is about resources. I think Dr. Smith pointed the way to one model which we certainly used in Colorado as well, which is creating local research capacity. It is incredibly expensive to fly folks to some of these communities to collect the data. Folks who already live there with a modicum of training and certainly a wealth of ongoing support and supervision which is enabled by all sorts of IT that we have available now that we didn’t used to have, are able to do that. That addresses in part I think, some of the resource constraints that had us thinking that we couldn’t include these communities.
I think the investment that the Natchez have made in creating that kind of tribal research capacity, the capacity that exists in the tribal epicenters that are in all of the areas, I think those are resources that might allow us to go forward even in time of tight financial constraints.
MS. KLOSS: We have reached that 5-minute point. We have questions from Vickie, from Leslie, and then I would like to ask — anybody else? We will ask both of you to convey any closing comments you would like to make to the committee.
DR. MAYS: Maya began to address the question that I wanted to ask and that is if you could give some guidance about this balance. You talked about, for example, oversampling or even longitudinal methods as an approach to increase the data. One of the criticisms is that some tribes want to know about their tribal group. And in oversampling what happens is we often get more of the same, but not enough to do clusters. The real push as I understand it is to do cluster sampling or to do clusters so that we can give more information about a particular group that may have a particular health problem. Could you kind of weigh that out? Is it that the guidance is that it is more data about American Indians in general? Remember we are trying often to deal with health problems. This is very difficult for the federal government to do is to think about not only in a national sample to do the national representativeness but then also go in and do cluster stuff. Can you give me some guidance about the thinking of that?
DR. VILLEGAS: It is a great question. We are actually in the process of developing a policy research brief to work with tribes about what is sampling. How does this work? What are the benefits to be had? What are the limitations on that data particularly around a national housing study that is being planned as well as some others at NIH that we haven’t been given a lot of details around, but they have said we don’t want include again the National Childhood Study piece of that.
I think our work is really to talk with tribes about methodologies, about sampling, about how this works and what we can glean from this piece. Obviously if there is money, we would love to have it both and to have that piece, but we are trying to explain that there is a real benefit. I don’t think it is a lot to explain so much, but just to talk about the benefit that can be gained claims about American Indian and Alaska Native populations back to particular tribes. While we might not have that data on a tribal level, it is still really important to participate. It is part of that both and.
We need that deep, local, contextual. We want to participate, but trying to find that blended space and getting that participation out there. I don’t know that there is any particular guidance on that, but to say that we are trying to do work with tribal leaders as well as urban communities about how to participate in sampling and what the benefits and risks are to that.
DR. SPICER: — to say real quick that there are certainly ways of conceptualizing Indian country that would allow you to make principal choices about which communities you might invite to participate that goes to the conflict of culture areas that anthropologists played with forever, but the availability of new tools in geoinformatics gives you all sorts of opportunities for doing that. It has been a while since I attended to the National Children’s Study, but I certainly understood that that strategy was to sample communities first and then children within those communities. It would be entirely consistent with that kind of approach.
DR. VILLEGAS: Real quickly though on that point too I think the regional approaches to data collection are really important. And a lot of the work that I have done in the past has also been to speak against that base approaches so just Native compared to non-Native to really like I said promote Native-to-Native indicators, measures, and research studies that do that work. We have done a lot of this in Alaska Native to look at the regions and to see the benefit. I really appreciate you pointing out the different geographies and ways of cutting at that.
DR. FRANCIS: This may be a question that is too big for the end of the panel, but it is motivated by the federal government’s open data initiatives and also the fact that we are an advisory committee to HHS. Here is the question. We heard in the last panel that people don’t trust the federal government with their data, but they are more likely to trust it if the federal government got it from the Mayo Clinic. Maybe because they trust the Mayo Clinic to ride herd on the federal government or whatever.
My question is given what maybe levels of mistrust of data in federal hands or possibly in state hands — there is a lot of mistrust about data in insurance hands and so on. The question I want to just maybe if you have any quick reaction to this, whether the data responsibilities differ for the federal government should the federal government be more open, less open, more protective, or are the principles that apply no matter who has the data whether it is an insurance company or your health care provider or the feds or whatever.
DR. VILLEGAS: Quick reactions are that I think we want to have greater access and to know what is being collected. My board and executive director have charged me with producing an annual Indian country data book and say you can imagine a massive undertaking but the approaches that we are thinking about that on one hand is baseline level. What is being collected on American Indian and Alaska Native people and tribes of significance that way?
A different approach is what are the indicators that are available and how
can we develop more what we call equity indices looking at the State of Black
America report, for example, and how they develop their equity indices also for
Hispanic Latinos and now they are talking with us about using those similar
measures to talk about status and progress towards a particular set of goals.
Equity isn’t the only goal I think for ours and for many communities. It is
wellness. It is well being. It is on those senses.
The third approach is really culturally-based measures of success looking at more local defined measures. For example, in the education arena the four are typical K12 standardized test scores, high school graduation, drop out, and attendance. Those tend to be the four when we take research back to communities and tribal arenas. Folks say that is all well and good. We have these other measures around participation and community events, leadership, these kinds of things that are very important to how we think about being educated in that sense. We are coming at it from those ways to say that I think that there are places where the federal government, state governments, and tribal governments need to be held accountable for what they are collecting and how they are reporting and how they are providing access.
At the same time we acknowledge our responsibility to community researchers, community leaders in a sense to engage with the data conversation to participate to engage to drive towards that end. We are trying to do a lot. I know, for example, NCES educational statistics has institutes that they host week long where they provide access to federal data that has been collected. How can we as NCI, provide greater access with our federal partners to datasets for Native and non-Native researchers to get some of those stories out that way.
DR. SPICER: Two points by way of closing. First, I think the comments Malia just made suggest that there are important lessons to be learned by engaging tribal communities about what kind of data might matter, not just for us, but for many other people as well. Certainly a hallmark, bedrock principle in cultural psychology — there is much to be learned by looking about the human condition by looking outside of our own cultural horizons. This is certainly an example of that. We might start measuring very different things, once we learn from tribes what seems to make a difference there. I think that is the benefit for the federal data enterprise and engaging these communities.
The second thing and the example of the Mayo Clinic and the reference at the end of Malia’s comments to NCAI, is a question of who is the broker for access to data. I don’t think anybody just trusts some generic federal government. Many people in tribal communities don’t trust a generic tribal government either, depending on who is in power. The issue is where can that data be protected and access to that is regulated in a way so that we have some assurances that data will be used properly.
And this idea of enclave I think is really an interesting one. Kelly didn’t talk about, but we have a colleague at the University of Washington Ron Whitener who is a Native attorney who has been funded by the LC program who is creating regional repositories for tribal data in the Northwest where the tribes then sign off on appropriate or inappropriate uses of the data. I think that there are some models that involve brokers and I think the Mayo Clinic is an example, and NCAI is an example, and I think Ron’s example in the Pacific Northwest is an example too.
MS. KLOSS: That is a really important concept to bring out to our group as we close this morning. I think that issue of who provides the overarching stewardship and protection has certainly been an interest the national committee for some time. That work is not done. Thank you so much. We have enjoyed meeting you and learning from you. We will now have an hour break for lunch and reconvene at 1:30.
A F T E R N O O N S E S S I O N
DR. FRANCIS: I would like to welcome everybody back this afternoon to this third session now of the NCVHS subcommittee on privacy, confidentiality, and security meeting on next steps for community data use. I would like to ask people who haven’t introduced themselves before, to introduce themselves because we have been joined by a couple of people.
DR. PETERMAN: I am Jeri Peterman and I am with the University of Massachusetts Amherst and I am a dietician. My background is as a community dietician, but now I am also an academic researcher.
DR. SILKA: Nice to meet all of you. I am Linda Silka. I direct the Margaret Chase Smith Policy Center at the University of Maine and I am in the School of Economics and a social and community psychologist.
DR. CHAPPER: Amy Chapper, Centers for Medicare and Medicaid Services, staff to the subcommittee.
DR. FRANCIS: Are there others in the back, who haven’t introduced themselves?
DR. KAUFMAN: I am Dave Kaufman from the Genetics and Public Policy Center of Johns Hopkins and I am on the next panel. I will tell you more then.
DR. SPELLECY: Ryan Spellecy from the Medical College of Wisconsin.
DR. FRANCIS: We are expecting to be joined by a third panelist, but we thought it made a great deal of sense to get going. That is what we are going to do. I would like to ask — do we have an order? Why don’t you tell me the order that works for you?
DR. SILKA: You have my written remarks and I am going to talk just a little bit about, so we have the data and we want it to get to the community. We want the community to use it. What are the kinds of challenges and opportunities that are there? And my remarks are based on — for three decades I have been working as a faculty researcher with community partners on a lot of different kinds of topics on how to do research together. There are just all kinds of interesting challenges. A lot of the work that I did for three decades in Massachusetts was with refugee and immigrant communities. Now in Maine it is with rural communities. And although those are very different groups, there are lots of similarities in terms of the distrust and the worries that people have in terms of working with researchers. Feeling like a lot of research gets done, but how does it help the community? Or having the research come back to the community in a form that works for us as researchers, but they really struggle with in terms of how to understand it and how to use it.
There are three overall areas that I hope that your committee will have an opportunity to give some consideration to. The first is the importance and challenges of taking diversity in communities into account. But sometimes we are fortunate in that we have a community that is all the people that are in the community are very similar to each other. In many places in the US that is no longer true. They are very diverse. What works for one part of the community in terms of data dissemination and data collection isn’t going to work for a different part of the community. How do we begin to think about that and address those issues?
The second point that I want to make is that universities are changing. There is a lot that is interesting that is going on in universities that might make them better partners in terms of data collection and data dissemination than they have been in the past. Those changes are going on not just in the parts of universities that have to do with health, but have to do with a lot of other areas where universities are really reconsidering their whole worry about not only being ivory towers, but being siloed and how they can do a much better job of being engaged in interdisciplinary work and that creates some opportunities for resource that are inexpensive for communities in order to do things.
And then the third point that I want to just touch on is that what is going on in terms of health data and thinking about how to use health data and issues of privacy and confidentiality and use is happening in parallel in other very different areas. People working in biological areas, people working in sociology, people who are working in physical sciences are going through the similar kinds of struggles. They are writing a lot about it and what is ending up in journals that are being read by physicists or by chemists aren’t the same journals that are being read by people in health so we are not learning from each other and there are real opportunities to look at how different areas are struggling with some of these issues. I want to just touch on each of those three and I would love to be able to talk with you at more length about them.
The first is in terms of diversity in communities. Jeri and I both, have had the good fortune of being able to collaborate with refugee and immigrant communities. And one of the things that has been very important and a real learning experience in doing that is that I felt maybe as an Iowan growing up in Iowa that everybody is kind of alike. Well, yes and no. I didn’t begin to understand just how great the challenges are when you are trying to work across cultural differences and understand each other. And in my notes I just mentioned a few examples of all the ways in terms of health area this can impact us.
People who are coming to the US who are refugees and immigrants may have had different health experiences, problems, and practices, different use of language to describe health and illness, different levels of comfort in answering interview and survey questions particularly when asked by strangers feeling like I may have signed something, but if somebody who is in power is asking me to reveal something, I must do this.
Differences in customs from their home countries, in how illnesses are expected to be treated. Differences in practices at the individual and family level related to health decision making. We did a house-to-house survey where we started out at the very beginning saying we would like to interview the head of the household. And people stepped back and said you don’t want my family to be a part of this. Thinking about those kinds of differences.
And differences in practices for developing health programs and implementing initiatives. How do you use data? What does data mean, and differences in previous interactions with health providers.
We have used those wonderful opportunities to learn to think that we have to change almost everything we do when we are thinking about data from beginning to end in terms of the cycle of research. Thinking about data collection and how to involve the community in the data collection.
One of the interesting kinds of research that one of my graduate students is doing is she is actually asking communities whether they want to be involved in every step of the research project. Formulating the questions, collecting the data, analyzing the data, writing it up. Groups really differ. Everybody seemed to be saying I want to be involved in helping to think through what it is going to look like, but other things there were people saying no. You do that. A kind of again challenges of — one size fits all doesn’t help in terms of developing the research or the research dissemination.
The problems of confidentiality like the last panel were talking about. I think Dr. Smith mentioned that or one of our colleagues mentioned that people wanted to have their names connected to the research. We found in some of our research when we said to people this will all be confidential. And it was more than even people feeling bad about not being connected. They said so you don’t think we are important enough that our names can be associated with this. You don’t want people to know that you got this from us?
Struggling with having the community says as we went out to do community surveys. We had graduate students in nursing and community residents who were going to be working on the data collection. And the community leader said nurses, you need to wear a nursing hat and white clothes. And the nurses were insulted. We don’t do that. And our community colleague said but that is how we will recognize you as health care people. We can have rules, but then we need to think about what do we do with this. How do we work together?
Finding ways to convey results. Frequently in American communities if we do our academic practice of saying we will give people written information about the results and we will feel good about it because we will do it in multiple languages. Many people if they are coming from refugee experiences, they have interrupted education. They don’t read in their own language and not in English. Figuring out what are strategies when we are talking about not just data collection, but data dissemination. What is that going to look like? We engaged our colleagues who did a lot of GIS, Geographic Information Systems, to create customized maps where people could come into events. We asked them if they wanted to have a map of a neighborhood. We put at the top their name and it was only their map, but it would have the data on it and it wasn’t in English. No one had to worry. All they needed to know is where the Buddhist temple was or where an important place was that they engaged.
And for us the lesson from doing that was realizing we have to step back from every way we do this research. It still has to be very high quality and rigorous. But let’s not assume that written material is what is going to make a difference.
And then we thought about how can we when we are gathering data, when we are thinking about the results, how do we think of this as being deeply embedded in the community and starting to come up with working together as a community event that would be cultural events, but would have the same data and data presentations built in. Now there is a Southeast Asian Water Festival, for example, in Lowell that thousands of people come to. And it started originally as part of an NIH grant as a way to focus on environmental issues, but actually give back the information so that it was alive to the community that it was a part of what people were doing. Thinking about every step of what we do and moving away from not the rigor of what we do, but the kind of academic style that we do it in.
Let me just say a couple of things about higher ed. Kellogg Commission on
Land Grant Universities, a number of different groups, are now kind of putting especially public higher ed putting our feet to the fire and saying you are expensive places. What are giving back to your communities? And more and more tenured guidelines and a number of universities are starting to say that outreach piece, that service piece, we are actually going to hold you accountable and we want to see you do a better job in that area. And that provides an opportunity for all of us to think about where is that under used capacity on some campuses that could be partners that would be cost effective. Are there classes where students are learning how to do data analysis and that MOUs could be developed? Are there thesis or dissertations that provide opportunities for in effect relatively free labor so that the students get better educated and there are things coming back to the community?
There are faculty looking at community repositories of knowledge. How can we invent new ways to get information out there that actually pay attention to how communities keep track of information? We use academic journals because we know to look there. That is where the information is. Where do communities look for information and how can we think about that as a way for us to do our work?
And then developing partnership approaches. More and more universities are really working to develop partnerships that will long standing universities or anchor institutions. They are going to be there for a while. They are real opportunities there.
I would also encourage all of you to think about looking at some of the work that is going on in fields outside of health. Great work in Citizen Science, that is going on now. We are talking about the loading dock problem, the problem where we collect data and it is like we are working in a factory. We are sure somebody is going to want our widget so we are going to take it out and we will put it on the loading dock and somebody is going to pick that up. Same way we do with data. They are really trying to think about how do we co-produce knowledge in a way that it gets used. They are not talking about the same kind of knowledge we are talking about here. But they are talking about the same kinds of problems.
Sustainability science. NSF is really pushing to look at how do we make sure that we are doing our research. They are trying to say how do we do a research so that it is actually going to end up influencing policies and practices. At U Maine we have a project going on now that involves 100 faculty, all the higher ed campuses in the state, a lot of graduate students, and those graduate students are learning interdisciplinary approaches to co-production of knowledge. Is it easy? No. Do people get crabby with each other? Yes. But it is transforming the way we think about how we are going to do our work so that the research on title power, on marine practices, on land use planning, on urbanization actually get used to make a difference in Maine and then it is helpful in other places as well.
My three messages really are we can look to other areas because they are struggling with the same things. And if we join forces, we can do some interesting work. Universities are changing. If you worked with a university 5 years ago, it may be different now because they are feeling more under pressure to do some things differently. And finally we are increasingly a diverse nation and that is an asset, but we can make it an asset if we use it as an opportunity to think about data collection and dissemination in some new ways. Thank you.
DR. PETERMAN: I also want to say thank you for this opportunity to come and discussing improving community health through results of research. One of the things that I have noticed — I came right at the beginning of the second session throughout the comments from Linda’s comments and from earlier comments is that using results really starts a lot before we are actually looking at results. My comments today are really going to focus on strengthening community capacity and community academic partnerships as a means of improving community health.
I approach this from my background. I have been working for over a decade as a registered dietician in a community health setting mostly with vulnerable immigrant populations. As a dietician, my focus has been very specifically on helping people to make dietary changes with the goal of improving community health. And more recently I continue my work as an academic researcher.
My research really has been focusing on how can I continue seeing these fundamental changes given now that I am in a research capacity. It is something that I have spent a lot of time thinking about over the past few years. Today I wanted to bring together challenges and opportunities from these experiences as both a community health worker and an academic-based researcher.
One of the first things is that effective partnerships and research in community health really depend on reliable analysis and well-conducted researchers to document community issues and document effective ways to make changes in community health. And this research and analysis requires an understanding of what is going on in a community, norms, needs, and best approaches including strengths of the community. But it also requires really rigorous methodology and good research practices. It is really a happy marriage of communities and academic institutions to come together because community agencies can bring understanding of issues and trust with the community and networks to both conduct the research and implement the results which is so important and academic partners can bring enhanced funding opportunities to get money to make it happen for research methods and vehicles for dissemination of the results.
But community agencies and research institutions have different pressures and different biases in their research. The community research biases really tend to be toward urgent community issues, the need for funding continued programs, and whatever community goals have been set by the agency and all of these can include improving community health whereas academic research priorities really are biased toward conducting the research, publish and presenting results, and securing funding for future work.
The good news is that this can also focus on improving community health. Despite the fact that there is this united priority of improving community health, these biases in community academic partnerships can compromise results because inevitably there is a tension when academic partners have more power than community partners and therefore their biases get larger weight.
An example of this is that the National Institutes of Health almost exclusively does not allow community partners as co-primary investigators in research when it is explicitly community based. Inevitably this leads to an uneven balance of power and that can mean that implementation of results after the research is conducted can be compromised.
Some community agencies have taken a novel approach of dedicating funds to support embedded researchers. This is what I want to talk about for the rest of the time is that this provides a research vision for the agencies and a means to identify and secure sources of funding to benefit agency goals. It really elevates the status of community-based agencies by building research analysis capacity of the agencies, but it doesn’t mean that academics are no longer necessary because partners still bring opportunity for trans-disciplinary work. They have access to much larger funding sources and they are able they are able to help expand funding and research dissemination opportunities.
Embedded researchers take a significant financial commitment on the part of community agencies and that may be difficult to sustain. I would like to offer two examples from my own experience as a community health provider and as a researcher to detail the differences between temporary and sustained embedded researchers and then offer a suggestion.
The first example is as temporary embedded researcher in Lowell, Massachusetts. A Cambodian-focused agency was subcontract to agency of one of the large racial and ethnic approaches to community health Reach 2010 grants. And as the sub-agency, it hired a registered dietician to address food issues. I was the registered dietician around nutrition and health around diabetes and heart disease.
As part of this work as a dietician, I looked for information about Cambodian practices. At that time I had no experience with Cambodians and I couldn’t find published literature on it. To my dismay and this really goes back to earlier discussion today a lot of people had done work with Cambodian community around issues of food and nutrition, but hadn’t published it. As a new person in the field trying to understand what was going on, I didn’t have access to the research that had been conducted and I consider that as a practitioner problematic.
Because I was a doctoral student at that time, I was able to work with my agency, the lead agency, and my university to design a research project to really get at some of the core issues around Cambodian food practices that could then be implemented as part of this Reach 2010 project in the Cambodian community.
This research provided a wealth of information that was practically useful. For example, we found that the experience of prolonged food deprivation in the refugee experience appeared to be related to poor dietary practices and weight outcomes in the United States in the food environment. And we were able to include specific advice from community members on how to address these experiences when promoting dietary change in a clinical setting.
And for the duration of the grant and a short time following the grant we were able to educate health professions and community members on very specific dietary practices and this was a really great thing. However, this was temporary because the grant was not renewed and in addition to many significant effects on how the lead agency and how the Cambodian agency were able to address health in this Cambodian community my position as a nutritionist was terminated.
Currently the information from this work is being used only sporadically to programmatically address community health. I am able to use it as a researcher to inform my research and work with community partners, but that is not the same as actually implementing it in a community setting.
The second example is sustained embedded researchers in Springfield, Massachusetts. A community health center in Springfield, Massachusetts is a federally qualified health center and the only refugee health assessment site in Western Mass. At the core of this agency’s mission is meeting the needs of culturally and linguistically diverse community and they have decided in order to meet this goal that research is a priority for this community health center. The health center has dedicated funds to a director of community research. And they actively seek out research partnerships with academic institutions based on trust and respect. They work to ensure that studies build upon previously studies and specifically address community identified needs and used culturally appropriate research methods.
They have also devoted funds to this director that she can use in other center needs particularly when she is not actively supported by significant grant funds on the specific grant and that makes the position sustainable and good for research and good for the community.
Some benefits of this position have been that the director of research has been able to partner with an institution with a recently completed RO1 NIH grant. That has resulted in presentations and publications, community agency partners or authors on these and sometimes the first author on them.
Currently the director is preparing a joint research proposal and I am one of her academic partners on this to build on the previous NIH grant and also bring in my research background and the research background and experiences of another researcher to serve specific needs that have been identified in this population.
Regardless of the success of the current application we have committed to continuing this research partnership long term. This is an example of when funds are committed what can happen. The results of this research will be implemented directly into standards of care for practice that serves a large portion of Western Massachusetts.
These examples highlight the impact that sustained embedded researchers can have on addressing community health. Many study mechanisms are time limited. And they do have value in achieving specific targeted goals for both communities and for academic researchers. But for sustained impact on community health they are just not effective. Funding that provides support for embedded researchers or incentives for agencies that commit some funds to it really have a potential to make long-lasting effect especially considering that community issues and solutions can take a significant amount of time to develop and then to implement. And when we have short funding cycles such as for grants, it is almost impossible to sustain something like this.
DR. FRANCIS: Thank you both very much. The floor is open to members of the committee and to others.
MR. BURKE: Thank you both for your presentations and from another New Englander welcome. My question is for Linda. I was intrigued by your remarks on involving community participants in design collection use of data. Can you share some techniques or how that was done well or places where you saw it wasn’t done well?
DR. SILKA: This study focused on all the towns and cities in Maine. We actually sent out a survey that went to every city in town at least five people who had various kinds of positions. It was based on a survey — a variation of a survey that was used in book called Swimming Upstream which was looking at community partnerships in a completely different area than health. I can get that reference to you as well.
And what we did is we modified their table and their survey that they did to ask about different steps in the collection of information that could be used by a community. We asked about a number of different areas the community was interested in. And then we did a number of different regression analyses to see who used what. But the thing that was very interesting to us and we did a similar survey with faculty that were involved with these projects. The difference between what the faculty wanted and what the community wanted was very large. Some faculty thought obviously what the community wants is for us to be involved, for them to be involved and for us to be involved in everything.
Across a lot of the communities that wasn’t the case. It was very eye opening to see that there were very nuanced views of how people want universities and faculty to be involved and how it can be helpful. And I would be happy to share the technical report from that. We are now starting to do it with some other groups too, but that was a major focus on municipalities and people have assumed for a long time that all of the cities and towns in Maine are similar to each other. They are not.
One other quick result we got from that is we asked people whether they had worked with university researchers and asked them questions about level of trust. We found that people varied a great deal in terms of whether they had worked with researchers, varied a great deal in terms of whether they trusted them. The trust factor wasn’t a really strong predictor not in the way that we thought it would be.
MS. MILAM: We heard from each of you about the important differences in cultures that you both experienced across your states in part due to geography and in part due to differences in origin and that sort of thing. I guess I am wondering in addition to having no similarities you both are in states that have more advanced health information exchange than we have in other parts in the country or at least development of policy. One of the pushes that might be argued as seen with health information exchange is overcoming “obstacles” or challenges to free flow of information, i.e., sometimes privacy or legal barriers.
I am wondering if either of you have had any interaction with your health information exchanges from a policy standpoint because I am hearing from you of all of the specific unique needs that need to be taken into consideration that seems to be sort of intention with easy quick flow of health information. And I was wondering if you have any experience with that.
DR. PETERMAN: I don’t. It is a really interesting question, but I don’t.
DR. SILKA: This doesn’t directly answer your question. Working with the Muskie School which is another group in Maine, we were asked by the state health department to look at mental health information needs in Central and Northern Maine. What are the issues? What are the concerns? We did a lot of focus groups in a number of different communities and talked to a lot of providers. And I don’t have an answer that says that people had strategies. But I do know they were very concerned because they felt on the one hand a great need for information and to be able to share information. Because these are such sparse light populated areas, everybody knew everybody else. There was on the other hand a great concern about sharing information. They were struggling with the same kinds of questions and probably would likely be very interested in things that come out of this committee because they really were worried about that.
The same kinds of issues are coming up with some of the sustainability issue because the surveys that had been done in that area have been in Maine have involved working with state government to collect information and there is a real battle going on now with the state agencies saying we don’t think this information that we have gathered about the environmental problems in Maine should be shared, should be disseminated. And the researchers from some of the universities saying this is some of the most important information that should available to people. Again, there is a tension there that has not been answered except that some of the people in the state environmental departments have said if you push this, we will not do anymore research with you. It is important enough that people are saying we are going to make an issue out of it.
MS. BERNSTEIN: Is that a stigma issue for the state or what is the reason why they —
DR. SILKA: We have had some meetings at the university to try to figure out why that is the case. We think it may be because it is possible that those state agencies — they have both an information function and a regulatory function. The data do not look good. And people could say if you were better regulators, we wouldn’t have these kinds of problems. Just conjecture.
DR. FRANCIS: That does raise the general question. If we are giving guidance — both of you have — and it has been a theme all the way along today that sometimes communities either want the data only given back to them privately so they can benefit, but not others. Or they really don’t want dissemination at all. Clearly the theme — it is very hard for you as a researcher coming along if others have done it but you haven’t been able to get access to it.
DR. PETERMAN: And as a practitioner too. And that is where the real implementation happens.
DR. FRANCIS: I guess the question is how should we be — what principles do you see as important principles or processes as a way of mediating that tension. One of the strongest views was well if the tribe doesn’t want the data shared, the data don’t get shared. Their word is the final word. There are actually governments you can look to there. They are official, tribal authorities which is not the case for even say a group of Cambodian or Laotian immigrants even though they might have very close cultural ties. I wonder if you could speak to first of all —
PARTICIPANT: — collection of towns.
DR. FRANCIS: Yes. Do you see it as a balance and if so how do we — help us think through more about how we draw those balances.
DR. SILKA: Again, this is another hopefully quick story that maybe suggests some of the challenges. One of the communities I worked closely with is Lawrence, Mass. And Lawrence is the most Latino community on the East Coast north of Miami. It is close to most of the nature research universities. What was happening in Lawrence is they were getting inundated with us as researchers coming in and wanting to do research. If the past research hadn’t been published, the people who were coming in didn’t know that the community had already said we have been asked these questions and we didn’t see any result coming from that. If someone in the community said okay, we don’t want this published that then also comes back and in a sense hurts the community because researchers keep coming in and asking the same kinds of questions.
One of the things that people in Lawrence did was to create a Mayor’s Health Task Force where they started to think about those kinds of things and to try to understand how a community like Lawrence differs from a tribal nation. In a tribal nation you have to have permission to come on the land. But in a community researchers can shop around and find another community group that will allow them to do research.
What Lawrence did as a community is struggle with these kinds of issues. I actually brought along and I can pass this around. We wrote a paper together that is called Creating Community-Based Participatory Research in a Diverse Community, a case study, where they looked at all the different IRB regulations and how communities could understand it and when they wanted to allow researchers in and how they could grab hold of the research process so that the end result would be the problems, the kind of health problems they were seeing in their community, for example, among their youth. Data would be collected that ultimately would help and that their commitment to other communities like theirs could be acknowledged and responded to. But at the same time this community wouldn’t be in the position of having lots of researchers come in and do research and do it in a way that isn’t helpful.
They tried to come up with their own strategy of something that would balance. They wanted to know what are the IRB-type terms. They didn’t want it just because they happened not to have gone to college and didn’t know those words. They didn’t want those words not to be used. They want to be prepared to go head on with researchers in terms of using the ideas. But that meant that there needed to be lots of ways to learn and think about these things. They tried to get most of the major access points in the town where people came in to try to engage in research to get everybody up to speed in terms of understanding this is what an IRB is. This is how people might come in to ask you about research. Here are the opportunities. Here is why we should stay involved because it is really important.
They told over and over again a story about EPA actually coming in and saying we know that your children have serious environmental health problems and we are going to help solve them. And the way we are going to do it is we have all these little saliva kits and we want you to go out and collect the saliva from your children. And that has become an important event in the community because they understand. The EPA didn’t ask is this a good way to do the research. Are you comfortable? How can we think about this together? And here is how the data in the end we think that can’t be used. It became a galvanizing experience for that community.
Every community isn’t Lawrence, but the question is how do we as a group think of strategies so that other communities can learn from Lawrence and that is why they wanted to write an article and get it published.
And then another step that they took is they developed basically a shop talk that they take to all the major universities in the Boston area and give talks about how you do research with communities, how you engage with them because they saw that as — they looked for what are the access points for thinking about these kinds of issues and making sure that the next generation of researchers understand this kind of partnership approach. And they are very committed to rigorous research. They want to have research that really matters, but they want to have it done in a way that doesn’t demean communities.
DR. PETERMAN: I see this happening too. I was at a meeting recently with refugees in Central Massachusetts trying to look at ways to address some refugee issues. One of the representatives there from a community agency said we have been interviewed. We want programs. We do not want to be interviewed again. We don’t want another study.
This kind of for me gets to the core issues here which is that if we don’t invest in what things that were mentioned in the community as a learning system for health report which is infrastructure and partnerships that what we will end up doing is a series of small steps that don’t look at long-term solutions. I know from a recommendation point of view this isn’t necessarily a great comment to have, but what we need is money to support prevention in the sense that if we have partnerships and we have infrastructure from the beginning that are identifying the problems and then have institutional memory for the research that is happening that is what is going to make the results happen in the end. Otherwise we are just little gnats flying against the windshield.
MS. KLOSS: Thank you. I couldn’t have said that more clear. But I was struggling with wanting to probe a little bit on your experience in supporting communities that have identified through whatever means; a need that they have found will make their health or their life or their wellness help improve that. I had a sense that that was what we were hearing from the communities who came forward and were interviewed for the CHIP report that these were more spontaneous projects that came out of an identified need rather than a researcher coming in with their own agenda. And how do we nurture that latter because that was kind of the thread of the learning community where we are taking the clues from the community and then adapting the resources that might be available to help whoever is the sponsor achieve those goals.
DR. SILKA: When we were looking at environmental issues, this happened to be at Lowell and I brought together a group of Cambodian leaders that I had worked with and a group of environmental health researchers from the university to talk about a grant that we could write on some environmental health issues. Jeri has heard me talk before about there have been a lot in the community about lead poisoning from in houses because the housing is old in Lowell and a lot of it hasn’t been — the lead hasn’t been dealt with in some of the homes. When we got people together, we started talking about this. The community leaders — some of them said to me but Linda lead poisoning isn’t our issue. That is not what we are worried about. I said, what is it. And we talked some more. And it turned out what they were worried about is they said we are really worried, something in our houses is killing our children. It is really affecting our children. It was lead. We weren’t using the same language.
But if we hadn’t been talking about lead, the faculty that I had invited to be there wouldn’t have been the right faculty. This goes back to your question of how to — yes, we need to do the things that are really responding to the community’s needs. But in terms of getting the right research partners there, how do we do that in a way that is going to work so that it is the people who are studying the things that are going to be helpful and how do we broker that and make that happen?
And one of the things when I was a faculty member at UMass Lowell, one of my roles was as the special assistant to the provost for community outreach and partnership. Part of my role was to bring faculty to these kinds of places and to help make these relationships happen. And what I would say to some of the partners after they had worked for a while with an untenured faculty member, I would say did you like working with this person. Did this work for you? Was this helpful? Tell me a little bit about how it was helpful? And when they did that then I would say well the only way this person is going to be around to continue to help is if they person gets tenure. Let’s figure out together how we can meet their needs in terms of the system they are in in publishing and sharing this information and meet your needs. If you helped to build this partner who has the skills that you want now, let’s make sure that they are able to stay here and how can we do this.
It is trying to figure out how to make this something that really can continue so that it is embedded like Jeri was saying and it meets both sets of needs. We get good knowledge that is really going to help solve problems in lots of areas, but the person, the faculty member isn’t doing it at cost to their career and it is not being done in a way that the community is doing it at cost to the health of their children and those are the puzzles that I think we are all kind of at the cusp of figuring out, but maybe we haven’t quite figured out yet.
DR. PETER MAN: In the UMass system we have two different things that I have been involved in and one has been a community engagement component that has really worked to identify needs in the community and matched them with needs of the researchpartners and that is how I ended up with working with this community partner in Springfield, Mass and that has been a great thing.
We have also had a system where — kind of going along the lines of what Linda has said is that institutions are having their feet held to the fire to need to get involved in community. They bring together all the faculty and they get all the community partners that they have and they say you guys need to work together. I won’t tell you which institution did which. And that has been a lot less successful than having a person who is like the dating master and helping to find your match and you have that partnership that is built. What Linda was saying even in terms of talking to community partners about what they can do to help academic folk get tenure, that is another investment.
I think maybe it is because some of my background is in economics and thinking about how can we put this into a currency that makes sense to people so that they want to put the investment in so that it makes money. I think we really need to be thinking in terms of cost benefit analyses, maybe creative cost benefit analyses, but what kind of a benefit do we get from a 2-year NIH grant or versus what kind of a benefit do we get from a tenure investment that supports the whole system? I think maybe it might be something good to bring in economists to help us make these decisions or these analyses.
DR. MAYS: I want to thank you both for a great presentation. My question is actually addressed to Professor Silka and that is you talked about this issue of the co-production of knowledge and that in doing that sometimes we see very different desires on both sides. And I am wondering if you could talk about the community side so that we get a better sense of not just being good stewards of the data, but making sure that we start to encourage people to develop data in ways in which they pay attention to that other side of what the community wants because we pay a lot of attention in making sure we have the right ends and everything else so that we can publish. But I am wondering if you might share a little bit about what are some of the other things to pay attention to.
DR. SILKA: One of the things that I have found really interesting in terms of working with communities is the rigor with which people approach things that they care about. If we can tap into that rigor, it could be very interesting.
If I ask a community partner something about research, people will sometimes shrug and say — certainly I have worked with groups where they say why do you need control conditions. Isn’t that a waste of money? But if we start the conversation at a different place and I try to think with my community partners about something where they really think that things might not work and they know they need comparisons in trying to figure out what they see as important areas where there are comparisons. This sounds like a long answer, but it might be that I ask people, for example, about when you grow stuff in your garden because people are doing a lot of that in Lowell — when you grow things in your garden, how do you decide where is the best place to do it and what is going to work and what hasn’t worked? And people will have really strong opinions about the care with which you need to do that.
Then what we try and do is find the connector so we are able to say to people you know that same worry you have about gardening and you are not sure if you know if you plant the so and so pods here as opposed to here that is going to work. That is what we worry about with research is we worry about whether we actually know if we spend all this money to try to do dat dit dat and you feel like your child is not getting better and the money could have been spent another way, is that what you want? What do you want? And we get into these discussions about rigor that are on their terms that then get really translated to the research so that people say I don’t want to have — I might not use the word intervention, but I don’t want to something that is done if we don’t know if it works. Yes, let’s figure out together how that would work. But it is connecting to the rigor and the places in people’s lives where they have that rigor and then helping people see that research is very much like that. As one of the people this morning was saying, if we already knew the answers, we wouldn’t do the research and really getting that kind of spirit.
We did a project it was to look at smart growth so looking at sprawl and things like that. Again, that is not directly related to health. But one of the things that Massachusetts was worried about is they were not getting a lot of the diverse leaders involved. We created a whole set of focus groups and then a festival that was about best practices people bring from their home country on transportation on housing. And it wasn’t another way to get at the rigor that people have and the best practices that they are thinking about and it opened up a lot of interesting kinds of conversations. Again, that is another a long answer, but saying that it is this intriguing thing of entering into the culture and then saying where are the points of rigor and then how do you tap into that in terms of the research and then what will a community say if you say what do you see as the most important thing here in terms of the really tough things we could do to make a difference.
One other quick story. We did some work with the Buddhist Temple around looking at heavy metals in fish. And we are thinking one of the things that the community is being told is, don’t eat the fish. It does not work. People are going to eat fish. It is an important part of the diet. Is there a way to engage the community thinking about what do we need to know about the fish, where the heavy metals concentrate? I know something about how the Buddhist nuns when they are cutting up the fish which parts then get cooked for the monks since the monks are high status. If you can show that the monks are eating this part and not this then you are more likely to get that impact.
How do you enter into the rigor in the community in a way that helps people see that research is just like other things that we do, but we need to do it well just like we do the cooking well, just like we do the housing well, just like we do the gardening well. How do we as researchers build those kinds of relationships in a way that the community can say two or three things to us?
I am not uncomfortable with research. I feel like I know what it is about. I want good research done in my community and I want other communities to know that my community is one that really celebrates research and sees it as important and that does happen.
Vicki, that is a really long answer, but it is truly stepping out of who we are as researchers but always staying researchers. One of the things that when we were having some conflict in the communities, some of the community people said to me Linda, what we need you to do in this case is we need you to be the researcher because then we can stand behind you. We can say this work is coming with this researcher. They were saying acknowledge your role, but don’t act like you are in that role. It was a very interesting kind of paradoxical thing.
DR. FRANCIS: Others with questions? We have been focusing on collecting data for research mostly. The model you have been talking about are models in which the community is engaged in generating new data. And I wonder whether you have any thoughts about engaging the community or whether it looks different if what is being used are clinical, medical records or linkages? We have a community attitude survey, but we also want to know whether the admissions to emergency rooms have gone up or down. We have to look at clinical information.
Part of what is a motivation behind this set of discussions that we are having is that there are new powers to use kinds of data that were originally collected for one purpose for another. On the one hand it is great because we have all this data, but on the other hand it is scary because people may lose trust.
Related to this you both have experiences with the use of non-health data. And some of the really interesting linkage possibilities that we saw in the CHIP report are things like where the grocery store is located. You were referring to environmental data and so on. I wonder whether you have any thoughts about those kinds of linkages.
I remember somebody telling — tell this as a little bit of a story. One of the people involved in the earlier discussions for the CHIP report about dissemination. They actually figured out from clinical records combining it with addresses and environmental information that a particular building had really high lead levels. Then the question was, how do you disseminate that information when you have used a variety of data sources that people might not have anticipated that you would come out with this result. The person who told this story wasn’t primarily thinking about protecting the landlord. But what he was thinking about was protecting — other people in the building might learn things that are tremendously beneficial, but also shocking.
PARTICIPANT: Your property values tank or something.
DR. FRANCIS: Or the school teacher figures out that your kid might have had significant lead exposures, stereotyping, and so on.
DR. PETERMAN: To that final issue that you brought up I just have a brief comment which is there is definitely a place for tiered release of data results as in determining who should have the information first and given an opportunity to act and then moving beyond that. I think that is something that could be made into a standard that maybe with a series of decisions about when to make tiered data results. That could come over a whole lot of ethical dilemmas.
Then to some of your earlier points of that question as far as linking other data with health data, the Economic Research Service has done some very remarkable work linking GIS data to locations of food availability with food deserts and is looking at nutritional outcomes. Michele Ver Ploeg — she is an analyst at ERS and she does really wonderful work and she is also very approachable. You should just give her a call for looking at that because she has great methodology for how she does that. She is a consumer economist.
I have another comment about the clinical records that I don’t think is directly related to what you were saying. It is really disturbing to me I guess maybe because I am a researcher or maybe even it is a discussion we were having at lunch about how so much of our data has become so public in so many ways. It is a really disturbing to think of being able to access data that people think are private and being able to link them up. I guess we will leave that for our ethicists who are coming later this afternoon.
I do know that in a research proposal that I am putting together with this community agency that we are putting that we are linking clinical records with specific people that will be looking for motivations and outcomes. That is different than looking on a population basis. But we are definitely saying we are looking on behavioral side, but we would also like to link to clinical records and we will be asking for explicit permission. Because the community health center is the driving partner on this with the embedded researcher that is a very — it is what they want and then it is a very easy thing for us to justify to an IRB.
DR. SILKA: One of the research teams at UMaine — these are people who are focused primarily — their primary interest is on use of JAS type monitors that kids can wear and it can track where the kids are going. This group of researchers was contacted by the Passamaquoddy community because the Passamaquoddy community in Northern Maine was very concerned again about things like food deserts and activity levels.
This group was developing a plan to collect data, but also use extant data for that community and connect it in order to look at the youth patterns of activity and sort of what was going in the community and link it to various health indicators.
But to me I don’t have an answer to your question, but to me what was surprising about being in on those discussions is that the researchers didn’t think that there was any possible issue there in terms of the connection. I think that is valuable for all of us to know because these were primarily non-health researchers, people who were GIS specialists and saw this as what a great opportunity. Just what a wonderful opportunity to try their new toys that they had in a way to collect some data that would otherwise be challenged to get. And because the Passamaquoddy have lived there for generations, they also had gotten a group that was interested in genetic information involved. And throughout all those conversations people really weren’t raising questions about the ethics of the issue. To me that is an example of the kind of group that we would want to be able to reach and understand what would they need to know in order to be more sensitive about the way they are going about the research.
A second example is I am on the board for the migrant health. It is a federally qualified health agency in Maine where we just have vans where we go out to the various migrant health worker areas to do health. And we have been contacted by a number of researchers who are interested in this population and connecting some of the data that could be collected and also comparing it to some of the health data in Maine.
The board of the migrant health center is really struggling with figuring out what to do and how to think about this and I think is another example of a group that would eagerly await guidance and suggestions and examples of what to do.
DR. FRANCIS: I will ask you one more which is about your experience with groups like Laotians and Cambodians and so on and identifying community leaders. We had a fairly lengthy conversation this morning about who speaks for the group and the question of whether tribes are different. I would just love to hear you talk about refugee communities and what you know about that.
DR. PETERMAN: I think I would like to hear what Linda has to say because she has more experience in this. It is such a complicated issue. I haven’t settled on a satisfactory way of dealing with it.
DR. SILKA: As more and more refugee immigrant communities gets settled in the United States, that is going to be a major kind of issue because I think it is exploding all over the place.
One of my colleagues who has worked a lot with refugee and immigrant communities — she said that the way for her, she knows when a community has in a sense, established itself is that the community can speak up and say that person isn’t speaking for us that that is the case.
I have seen this in a number of different communities that I have consulted with is when it is a fairly large population and community there is often one organization that captures a lot of the funding and as seen as the leading refugee and immigrant agency in that city for a while. And they are able to capture the resources and then like all of us get corrupted by those resources and no longer speak for the community. And over and over again in some of the communities I have worked with I have watched these really difficult kinds of situations where people become — certain organizations become the go-to place if you want to do research or you want to do so and so. Instead of becoming networkers they become gatekeepers, in ways that are really very challenging.
At one point I was on one of the major agencies and Lowell fired its executive director. There were three of us on a meditation team to try to figure out what to do about this situation, two Buddhist elders in the community, and then I was there to talk about what happens when you take federal funding and you are doing funny stuff with the federal funding. It was all in Kumei(?), and I was really struggling to try to figure out what is going on, who is doing what. And the reason why I tell that story is because I found it so helpful.
I still don’t know exactly how or what I learned except that I learned that by watching how people were responding to this person who had been fired and watching over and over again I really started to get a feel for the kind of people who almost get so pulled into the mainstream culture and the status and the mainstream culture that they really struggle to continue to hear what their community is saying to them.
We had many meetings, not just those, but others where the gender differences came out where it was really clear that in some communities that I was working with that really a kind of sexism was going on that to me was really worrisome, but it wouldn’t work for me to speak out until somebody else was able to say — somebody in the community say sexism is wrong because it hurts our children.
Again, that is a wandering answer except to say that I don’t think that there are rules, but it is another one of those amazing things where somehow in each place there needs to be a lot of context so you get this kind of nuanced sense. I felt like a lot of the community people were trying to figure out the university that I was at and figure out who they could trust there like who were reasonable spokespeople from that university and who were jerks and what they started to get a feel for, how you work with a large institution.
As we go forward in terms of building these kinds of community partnerships that rather than seeing this process of learning who to trust as something that we can have a set of rules that maybe we need to have a set of almost stories and metaphors and things that actually capture the process of how we enter into each others’ world so that we can actually get things done together.
Just one other quick example. A student came to me in my class and he was Cambodian and he looked very troubled. I asked him what was bothering him. He said that one of his teachers, one of his professors had said I want to study the families in your community because I know how just troubled Cambodian families are. And my student said I couldn’t tell her no because she is my teacher. That became a really good story then. It was an awful thing to happen and we could talk through it.
But it became a kind of story that we could use within the university to say let’s think about how we engage with people and what kind of status we are carrying that may make it difficult for people to give in effect informed consent. And how do we pay attention to who we are and what we are doing so that the end result is health gets better in the community, these relationships exist. We become stronger, better, more nuanced researchers and that we learn, but we don’t say that we are not researchers and we are not faculty. We have a lot of interesting things to think about as we become a more diverse country and the diversity could be along all kinds of different sorts of dimensions.
Now that I am living in Maine people say you are not a Mainer. You don’t understand Mainers. It is true that you have been around for 30 years, but that doesn’t make you a Mainer. There is a culture. How do we understand each other’s culture and how do you set up rules and regulations that don’t exclude opportunity, but also understand the complexity that is there? That is why you have a wonderful task ahead of you.
DR. PETERMAN: I actually have something to add. It adds complexity though. In thinking about identifying leaders, leaders for whom? And the reason I ask it like that is because some of my more recent research has been with adolescence and emerging adults. These are young women who are themselves children of refugees or were born maybe in refugee camps, but grew up in the United States. And, Linda, you can correct me or talk to this. Especially when there are groups of people where adults move to the United States then community leaders have an importance that they don’t have for these younger generations who are very individualistic and who really don’t want older people even in their own communities speaking for them. One of the reasons being is that these are the children who came to the United States, picked up English quickly, acted as the ambassadors for their family between the US system and between their family and their culture and to have somebody from a senior generation appointed or recognized as speaking for them can be really problematic.
MS. BERNSTEIN: Just talking about whom to trust made me think about two different ways we were thinking about it. This morning we heard more than once about how Indian communities don’t trust the federal government. Other people don’t necessarily trust the federal government, those who live in Rochester, Minnesota apparently, but they trust the Mayo Clinic. We keep asking the question of how are Indian communities maybe different than other communities. There is a long history of Indian communities in the federal government in this country. You can imagine why maybe they don’t trust the federal government. But with refugee communities they are much newer, much less experienced with the government of the United States. I am wondering if that is a different way to look at it in terms of what institutions they trust or don’t trust.
But also you mentioned a different dimension which is who is a figure of authority. Maybe those are related but not exactly the same. I am wondering if you had more to say about the first example and also whether there are other dimensions that perhaps different communities that you worked with that we should be thinking about in terms of who people trust that aren’t either of those dimensions.
DR. PETERMAN: An experience I had and I am just going to give a very personal experience is I shared an office with two Cambodian refugees at this agency. We would over our breaks and at lunch just kind of talk about the experiences they had been through. I asked them very explicitly about trusting the federal government. I can’t remember that context. But one of my office mates said we are amazed by this government that would take a people in and welcome us and then give us support. In that one context I don’t think he was speaking for the entire community, but it is different than having a history of the federal government essentially doing terrible things to your people. I think that there are some differences.
MS. MILAM: I would like to share some information. We did a randomized telephone survey of attitudes on privacy and security in West Virginia. We had the same conclusions though. We found that people generally don’t trust the government, state or federal, don’t like health plans and really trust their providers. If you have a message that needs to get to those people, you do it through the providers because they are the trusted entities.
DR. SILKA: A lot of times we focus on sharing information in equal power kinds of things. One of the struggles that we encountered with work in the schools is that sometimes refugee and immigrant parents would say to us we feel like the teacher should be in charge. For us to give people advice or to come to something like parent/teacher conference or to enter in that is not respectful. So kind of these multiple ways to think about the relationships that make it really challenging because partly it is who you trust. It is also how do you show respect, and that there are so many complexities.
DR. FRANCIS: Thank you both enormously. This was just absolutely fascinating. We are going to reconvene at quarter after three.
MS. KLOSS: We want to thank in advance Dr. Spellecy and Dr. Kaufman for being with us on Panel IV today, entitled Consumer Attitudes about Community Health Initiatives. We know that within that broad title we will be able to have the kind of far-reaching discussion that each of our panels has had today.
So welcome. I believe everybody who is in the room has introduced themselves except you. So as you take up the podium please tell us a little bit about the work you do, and your background. We will dive right in.
DR. KAUFMAN: Thanks so much for having me here. It is a real pleasure to share some thoughts with you guys. Just so you have a little idea of my background and the kind of work I am doing, my training is in genetic epidemiology.
I did a little bit of that work and just sort of by happenstance, in some select populations, Ashkenazy Jewish families and then Alaskan natives and American Indians, and I think some of the things I learned most from that work were the significance of community work.
I didn’t intend on that, but it has really opened my eyes to those sorts of issues. And now at the Genetics and Public Policy Center, which is part of Hopkins — we are down in Washington — we have spent the last five years doing quite a bit of public engagement work, mostly with two populations. One of them is the American population. We have been doing a nation-wide public engagement, and then also with the VA patients.
And in both instances we have been talking to people about their attitudes and concerns about participation in large-scale gene environment research. We have started pretty much all these studies out by giving an example of a longitudinal cohort study that would be nationwide, and ask people to donate blood samples for genetic analysis, medical records, fill out questionnaires and be updated along the way with broad research goals.
We have asked questions about just general willingness to participate, and fears and expectations, why people will participate, what they would like to get back. We have talked a lot about return of research results. I would be happy to talk to you about that.
We have talked about privacy concerns, ideas about consent, and some methods of how they would prefer such a study to be carried out. So the stuff I put in the statement, I thought you guys were interested in privacy so I thought I would give you my general gestalt about that. I am happy to talk about any of those things if they are interesting to you guys.
And I am also happy to talk about our public engagement methods. I will just say there are lots of different ways to do engagement. There is talking and there is listening. There are different times for both. When you are planning a study you may really want to sit and listen to what people have to say. And when you are ready to do a study, you may really want to just be talking and making sure that your community understands what it is you are going to be doing.
So with all that said, in a lot of studies about the conduct of research including work that we have done, we have seen that people strongly value their privacy as well as the chance to participate in biomedical research. So there is sort of broad, public support for biomedical research that is balanced against people’s concerns about maintaining the confidentiality of individuals as well as the reputation and the wellbeing of communities that are involved in research.
Many studies have observed that despite ubiquitous concerns — and when I am saying studies, I am talking about the sort of social science research that we have been doing, and many others are doing this as well. So a lot of folks have observed that despite these ubiquitous concerns about protection of privacy, majorities would be willing to share their clinical data and samples for research as long as their permission is sought beforehand or the data is de-identified to protect their privacy.
So in the study we conducted about willingness to participate, 90 percent of respondents in the nationwide survey we did, about 4,000 people, 90 percent of the people said they were concerned about their privacy. We asked why.
Less than half of them said that they were afraid that their day would be used against them. About two-thirds of the people say that they just don’t want people to know the icky things about them, what their health is. We were sort of surprised by that. We thought it would be more identity theft and insurance discrimination. But it turns out it seems that people’s concerns are more just actual privacy about their own self.
In addition to worries about discrimination and you knowing things about me that I don’t want you to know, we have seen that concerns about maintaining privacy are strongly related to issues of control over information about oneself and about one’s community.
Providing individuals and communities with a sense of control over how data and research findings are communicated to the world outside may be as important as minimizing the harms that can actually accrue from studies inadvertently sharing personal or community information. We have observed that we don’t think communities want to constrain or restrain how research data and research results are used and shared. They want to be asked.
Our observations suggest that if community members are given a seat at the table where the decisions about the uses of research data are made, and if communities are told in advance by study leaders about studies, data sharing plans and privacy protections, it is this transparency that is most important, more important than what the researchers are actually going to do in many cases.
This transparency about what will happen will often lead to trust and acceptance of researchers’ goals. Participants and communities want to be given the opportunity to understand and assent to the research process. The respect that researchers show communities by communicating their intentions and soliciting input before they act, is critical.
A proactive, honest approach about the realities of a given research endeavor will go a long, long way. This isn’t to say that community-based participatory research isn’t worthwhile or important. Not all studies are going to be able to do that. But many studies can engage their populations before and during the process. And that may go a long way, especially if you have got a longitudinal study or a study where you are recruiting people that previously have been difficult to get into research.
So researchers don’t have to make everyone happy, they don’t have to give research subjects everything that is desired. But if they can explain what they will and won’t do and why, that may be enough in many cases. One example from our survey — it is not privacy related.
We asked people if they were going to participate in this sort of national study, would they prefer a one-time broad consent for use of their samples, a consent each time a research idea came up or have a sort of menu of diseases that could or couldn’t be studied, and about 45 percent said broad, 45 said narrow and about 10 percent liked the menu idea.
So then we asked people, which of these things would you allow your samples to be used for, if it was a sort of study by study. And we gave a big list of conditions, starting with cardiovascular disease, cancer, mental health, diabetes. And then we got into some sort of touchier things — IQ, I think we put sexual orientation, ability to learn a foreign language.
There were probably 12, 15 conditions, and about 10 of them were mainstream. You would think that maybe the people who said they would prefer a narrow study by study consent would be less willing to support some things. For all the mainstream conditions the level of support was exactly the same, willingness to support research for those studies was exactly the same for people who would give broad or narrow consent. It is just a question of wanting to be asked. So we thought that was interesting.
Another example that we found more related to privacy is that large majorities of people don’t believe that researchers can absolutely guarantee the privacy of data that is collected for research. However, most people are willing to accept this reality and assume that research projects will act in good faith and do their utmost to protect the privacy of participants.
If researchers can meet these expectations to do their best and be explicit and transparent about the protections that they offer and the remaining risks, then communities and individuals are likely to assume those risks and participate in mainstream biomedical research. This sort of transparency has to begin at the very latest at the time of consent.
During consent, potential research participants should be told about the different levels of de-identification of data that are possible, the fact that de-identification might be possible, especially if you are dealing with DNA, the explicit protections that are offered by the study protocol and the risks that remain.
In addition to providing participants with forthright explanations of primacy risks, consent should also detail what data can be gathered through study protocols, who the data will be shared with — academics, industry, government. Broad categories are very, very important to people. Whether you are going to share data overseas is very important to people.
How the data will be analyzed, what formats the data are likely to be published in, et cetera, and what sort of information is going to be returned — aggregate results, individual results, no results. If you are going to return no results, you just need to say that upfront.
The desire of participants to know what they face up front will be satisfied — well, it may be satisfied — based on their honest assessment of the risks and protections that researchers are offering, and establishing that trust, again, will carry you a long way, I think, if you are doing community-based research.
When longitudinal community-based studies are being designed that require acceptance from a broad range of participants, researchers may do well to engage communities in advance of recruitment and consent, in the sort of way we have done. The VA has begun a study, taking a good deal of what we found into account, especially with respect to the consent process. And that has become, in a way, standard practice.
Also, transparency about privacy risks must continue, and other aspects of the study as they change. And studies are changing all the time now, protocols are continually being revised as technology grows, as people think of new uses for data. As the study goes on, transparency needs to continue. If risks or protections or data sharing change, that has to be communicated in advance.
If it is communicated in advance, it is likely to be met with acceptance. When an unexpected situation, like a privacy breach, occurs, the researchers must be direct and communicate as quickly as possible to let the community know. The worst thing you can do in any case is not to tell people what you are doing.
If you are not familiar with what happened with the Texas blood supply registry, research that had been going on for years and years without consent on newborn blood spots was abolished, simply because people were not told it was going on. People didn’t actually object to the research, they objected to not being asked. And as a result millions of blood spots were destroyed and that resource is gone.
In addition to clear communication of outlined privacy risks that accompany participation and research, the research enterprise must work to fortify the protections that it offers participants. Policies about the publication and public access to de-identified data that include genetic sequences should be reviewed by parties that share or publish such data.
Researchers need to be encouraged to use Certificates of Confidentiality to protect participants from forced disclosure of their identities for use in civil, criminal, administrative or legislative proceedings. The NIH needs to consider adopting a different model of certificate confidentiality, such as the one used by the researchers at the US Department of Justice.
That Certificate of Confidentiality actually forbids researchers from turning over data. The NIH’s certificate just says researcher, you don’t have to if you don’t want to, you can’t be compelled to. But you still can if you want to.
There is no reason to give researchers that out. It just softens the protection and can put researchers in a spot they don’t want to be in. So it seems it will be worth changing that. It may also be worthwhile to examine what practices researchers and data access committees are using to maintain privacy, especially when data is being shared.
Finally, people and research studies are now using social media to connect and organize themselves into new communities of shared interests that transcend traditional geographic and sociodemographic boundaries associated with the term community. These new forms of community have unprecedented ability both to gather and disseminate information about their interests and their concerns.
I think large scale research needs to be increasingly mindful of keeping these new facets of community involved in discussions about privacy and the release of study information that may impact both the communities and individuals. Thanks.
DR. SPELLECY: My name is Ryan Spellecy. I am an Associate Professor of Bioethics at the Medical College of Wisconsin. I am trained as a philosopher but I tend to work in bioethics. I am also an IRB chair. Thank you for not throwing things at me.
I will approach this from the perspective of an ethicist and an IRB chair, and I want to focus my comments on the use of data for research because that is where my area of expertise lies, although obviously this will have import to areas other than research, all sorts of secondary data use.
And I want to begin by highlighting some main areas of ethical concern and briefly talk about a fundamental conflict in the ethics of research that the more I think of it, the more I am concerned that it can’t be resolved. And so I want to end by proposing an alternative way to conceptualize how we might best uphold some of the ethical principles of research.
It might not completely resolve the issue, but it might be the most ethically defensible way forward. And so when we talk about research ethics and the use of data in research as I have heard from other panelists today and other speakers, it seems obviously that privacy, confidentiality and informed consent forms the crux of the ethical issues when we talk about issues related to this research.
If patients give their consent on the front end to use their data for research, there is really no controversy. Where we start to get concerned is as is often the case or was the case, patients do not give their consent for the use of their data for research, and then their data is used for research nonetheless.
As Dr. Kaufman just sort of said, they want to be asked. And when we don’t ask, and we never ask, at the end of the day we violate their privacy and their confidentiality. And let me give you a commonplace example.
A woman goes into her physician at an academic medical center for her yearly checkup, fills out the check box things of do you wear your seatbelt, night sweats, anything like that, headaches, and then does a history and physical. If we don’t get her consent, she offers up this information, sometimes very private, personal information, to her physician, with the understanding that it is given in confidence and that it will be used for her clinical care. Because we never told her it was going to be used for research.
It goes into her electronic medical record, which is wonderful because it is so accessible and so secure. It is not those paper things any more. But it is also so accessible for researchers. So a researcher comes along who doesn’t even know this woman, and this woman has never met him, applies through the IRB to do a chart review study.
We do these all the time. Consent is waived. He accesses her data for research, to which she never consented. And in a very real sense we have violated her privacy and her confidentiality. Because at the front end of that encounter, the understanding was not in place that when I am giving you this information you will be using it for reasons other than my personal clinical care, other than reasons intended to directly benefit me.
And I say that we violate the privacy and the confidentiality, not to sensationalize things or not to suggest that chart review research is unethical, but to highlight what is at stake here and to encourage us to be cautious as we move forward with this kind of research. So that we are careful about things like data breaches and breaches of confidentiality, and we see the chilling effect when there are breaches of confidentiality.
Even if we do get consent, the chilling effect that has on research when somebody leaves a laptop in a car and it is breached. Trust is shattered, it is expensive because now we have to pay for credit monitoring and all that, for anyone whose data might have been on that laptop.
So when we talk about confidentiality, privacy and informed consent, consent is important because it is the primary means by which we uphold the Belmont Principle of respect for persons, or as it is sometimes called respect for autonomy. And respect for persons requires that we ask someone when we use their data.
At the end of the data it is their data, and they ought to be in control over who has access to their data and who doesn’t. That is the gist of respect for persons. Of course no Belmont principle is absolute, and the Belmont Report exhorts us to identify the relevant ethical principles and balance them. And in this case and in your standard chart review we are balancing the good, the beneficence of the research that it might hold, versus the affront of respect for persons.
But this is the problem that I am concerned about here, that when we think about informed consent as the placeholder for respect for persons, it is an issue of rights. And rights don’t admit of degrees. Your rights are either upheld or they are violated. If your right to vote is violated, your right to vote is not partially violated, it is either violated or it is not.
Rights admit of absolutes, whereas things like beneficence, going good, we can do more or less of that. We can’t more or less violate your rights. And so as we are trying to balance these principles, you have one principle, this right to informed consent, that really doesn’t admit of any balancing act against another one that does it. There are almost apples and oranges.
Once solutions have been offered, still in this framework of informed consent have been well let’s just consent to everybody on the front end. This is enormously expensive, and if we really want to consent them — I do some pediatric ethics stuff, too, and one of the things we learn is, when we engage children in the consent process or the assent process, we have to take seriously their refusal.
We know this about kids. We can’t present them with the opportunity to assent to their research participation or their clinical participation, and then when they make the answer that we didn’t want them to make, we can’t say oh sorry, and yank it out. It is very bad. It teaches them that their opinion doesn’t matter.
The same is true for adults. If we are going to seek their consent, they have to be able to decline to participate. And if they can decline to participate, all of a sudden that data set is open to bias, because we have all the decliners that we can’t access.
So if we seek consent it is expensive and it creates real problems with the integrity of the data set. Other opportunities have been notifications, which are usually buried in HIPAA notifications or the general consent to treat, which quite honestly people don’t read very closely, if they read them at all. And so it is hard to see how these notifications really do anything in the name of respect for persons.
So a better alternative — I want to explore briefly a better alternative that I think re-frames this conflict between rights and beneficence. And it is not a question of do we consent them or not, or do we notify them or not. But rather, it seeks to engage communities, to uphold respect for persons.
And additionally, the nice thing is the data shows us that these approaches enjoy widespread support among communities. And I think you have heard that from the other people who have offered comments today about community engaged research and such approaches, and the widespread support, when done right, that the community affords these approaches.
And I think that NCVHS already recognizes this, of course, when they spoke in the report. The community as a learning system, using local data to improve local health, seeks to build that infrastructure to involve local communities in health care decisions about data use in fostering a sense of data ownership. And so hopefully, I am preaching to the choir here.
The challenge is how to engage key stakeholders and communities in the process in a manner that is not ad hoc or after the fact, but one that does so in a true spirit of respect for persons. And there are two models that I want to mention briefly, that might be useful for considering the ethical use of data.
And that is, emergency research that proceeds with an Exception From Informed Consent, which is called EFIC research. And what I will call, Community Engaged Research, or CER. Dr. Kaufman mentioned community based participatory research. This goes by different names. It means very different things. But for our present purposes the gist of it will do. So I will call it community engaged research, and I know the previous discussants referred to this sort of research as well.
So those are two models, EFIC research and CER, that offer lessons both in what works well and what doesn’t work well to engage the community. So first, EFIC research. EFIC research is unique in that it invokes a special regulatory framework that allows researchers to conduct emergency medicine research out in the field without informed consent, as long as certain protections and guidelines are in place.
This research, it is not possible to conduct this research without consent because it is often out of hospital emergency medicine research like heart attack or public access to fibrillators, where we can’t get their consent because their heart is not beating and they are in a technical sense dead. So there is no time to consent them.
And we can’t really prospectively consent them because it is very difficult to recognize who is going to have a heart attack at any given time, unless we just pass out consents at the drive through of McDonald’s, which has its own challenges. So we recognize that we can’t do this kind of research, yet the research is essential for advancing health care knowledge.
I’ll bet some of us around this table know someone whose life has been saved by public access to fibrillators, those things you see at the airport and such. Or we know a story of that. And that would not be possible without this kind of research. And one of the safeguards that researchers in the IRB have to conduct is community consultation.
And this is interesting because the point of community consultation under this research is not proxy consent. It is not supposed to be well we can’t consent the individuals so we will consent the community. Rather, the point is to consult with the community, learn what they think about the research, if there is any changes that should be made to make it more appropriate for the community.
And then see if there is any real genuine community concern about doing this sort of research in the community. And this provides an ethical framework for thinking about doing research in which consent is not feasible, but still adhering to respect for persons without consent.
Additionally, in the published research around EFIC research, the public attitudes towards the research has been both positive and supportive. And I want to highlight one important finding in some of the research in the surveys that they do, to gauge public support for this kind of research.
The public respondents — and this was in random digit dialing surveys — showed a really interesting, sophisticated understanding of the research. They indicated, when a greater proportion of survey respondents supported the conduct of this research in this community than did want to be enrolled in the research themselves.
I know that came out awkwardly, but basically — I will just make up numbers but I know they are partially accurate. Say 67 percent of the respondents said I would want to be enrolled in this research. But then maybe 89 percent said, I think this research should be conducted in our community.
And so they recognized that, you know what, I don’t want to be enrolled in this research and I wouldn’t enroll in it, but I think it is important research and I think it should be done in my community nonetheless.
The other model I want to talk about briefly is Community Engaged Research. And I will close with that. This offers a method for engaging communities. We talked about it in this group at length and there are great summaries out there so I won’t explain it.
But the basic idea is, unlike this EFIC research in which researchers come to the community with a study design already planned out and they wish to implement it, Community Engaged Research engages with the community to identify research priorities and build the study up in a bi-directional manner with the community.
They are bi-directional, they encourage respect for all parties, and it acknowledges that the researcher has as much to learn from the community as the community has to learn from the researcher. I believe that one of the previous commentators mentioned the importance of sustainable research enterprises in Community Engaged Research.
And I really can’t emphasize this enough. Communities, especially the ones we work with in the Milwaukee area, become pretty savvy about research pretty quickly. And when they don’t see the benefit from the research, they are less likely to want to engage in the future. And one representative from the Milwaukee Public School District, it is an urban school district, it has all the challenges you would expect a poor, urban school district to have, shared with us.
She said that MPS, Milwaukee Public Schools, is littered with the skeletons of hundreds of helicopter research projects. And what she means there is, you come in, you do your research and you get out. And you never tell us what the results were, there is no benefit to us that we know of. We don’t even know what happened. And when these community groups see this, they become reticent to participate and to collaborate in the future because they see no benefit. The researcher gets the benefit, but they don’t see the benefit.
So in conclusion, the use of data for purposes other than the reason for which it was initially collected poses unique ethical challenges. Fortunately, there are some examples such as EFIC research and Community Engaged Research that provide a model for conducting this important research in a manner that adheres to the highest ethical standards, that upholds respect for persons but perhaps does so in a model that waives consent. Thank you.
MS. KLOSS: Thank you very much. Questions? I will start out. Have you had experience with the Community Engaged Research model for projects other than genetic projects?
DR. KAUFMAN: I have seen it. I think a lot of the work that I saw in especially the American Indian communities that I worked in, genetics was being folded into that. It’s sort of like Framingham. It had been going on for a long time and now genetics was a possibility. But the community aspects of it had been put in place long before.
I would say just generally speaking, I think that bio banks and genetic researchers are wrestling with a lot of the problems that have already been thought about and in many ways addressed by researchers and American Indian communities and Alaskan Native communities participating together. There is a lot of work that has already been done there.
A lot of wheels that may not fit exactly on large-scale genetic research, but they have thought a lot of this through. And I think to the extent, if you feel like you haven’t heard enough from those folks I would encourage you to bring them in, because there is an enormous amount of work about how do we do this that they have already solved.
DR. SPELLECY: Most of my experience with Community Engaged Research is actually not in genomics research, both from an oversight perspective as an IRB chair, and as a researcher. It is things like biomes prevention in the schools to healthy eating, to — oh gosh, we did a project using foto-novelas, which is almost like the telenovelas in Spanish, to teach healthy eating habits to Latino seniors on the South Side of Milwaukee. I can send you the little thing. It is really funny.
MS. KLOSS: So really examples of those kinds of potentially high impact projects that are very community based and were brought to life in our CHIP report.
DR. SPELLECY: Yes.
DR. MAYS: Thank you to both of you. It is nice to have an IRB Chair here. We don’t throw things, we instead just ask questions. One of the questions I want to ask you in particular is, we have been talking a lot about the importance of these relationships.
I am trying to get a handle on, as we move around this issue of community data and also the IRB has a role in terms of protecting it and making sure it gets out, what are the things that in the IRB you think you should be thinking about in terms of relationships? Either from the point of view of protecting the committee, or from the point of view of protecting the researcher.
DR. SPELLECY: So I will say two things, and the one is probably not what you want to hear. It is a little controversial in IRB circles. Most IRB professionals, OHRP, take the view that — so there is a regulatory prohibition against IRBs considering long-term effects of research. And a lot of the time the negative impacts on the community are considered as part of that prohibition.
So it is something that IRBs are actually by the regs not supposed to consider. It is a big problem. There is an article by Lainie Friedman Ross and some others that I believe I cited in my written comments, that proposes a really neat, informative model by which researchers, the IRB and the community can collaborate, because she really lays out the gaps in human subjects protection, in community-engaged research.
And one of them that she highlights is the fact that IRBs are not supposed to by the regulations consider most of these downstream effects. So there is that. So your question was, what should IRBs be thinking about, and then what should they be thinking about in terms of protecting investigators.
DR. MAYS: Yes. We have to get community boards, et cetera. So what kinds of things should we be thinking about in terms of the nature of that relationship as to who those community individuals are, and what they are allowed to do and not do.
DR. SPELLECY: So one of the major challenges is, when you engage in community engaged research, the community partners often become members of the research study team. And all of a sudden they are now, because they are conducting research regulatorily, they meet the criteria for investigators. And so all of those regulatory aspects apply to them.
And what they would have normally done, without blinking an eye in their role as a community help nurse or something, they can’t do as a PI, as an investigator and a part of this study. So one important thing is adequate training for community members in the role of the IRB in the regulations.
And I know that the University of Michigan does a lot of Community Engaged Research and they are piloting a new education and training program for community based research, especially in an academic and community partnership model.
Most researchers and IRBs, we do this thing called CITI — I don’t even know what the acronym stands for, but you sit in front of a computer screen and —
MS. KLOSS: I was going to ask what that acronym was.
DR. SPELLECY: I don’t know.
DR. KAUFMAN: No one knows.
DR. SPELLECY: Yes, no one knows. It stands for something we have to do. But CITI is a real problem for Community Engaged Research, because community partners aren’t just computer — this is sitting in front of a laptop, and they are not as comfortable.
MS. BERNSTEIN: Can you just explain in two sentences what that is?
DR. SPELLECY: The regulations require some sort of training in the ethical and regulatory conduct of research. And most institutions use CITI, which is an online program, because it is easy to use. You just pop yourself in front of a computer, you go through, you see some slides. If you are me, you make notes about where you disagree with what they say the right answer is, but you do it anyway, because you have to pass it. And then at the end if you get 80 percent right you pass, and you have met your regulatory criteria for three years or so.
It doesn’t work so well with community partners, because they don’t have laptops or they are not as computer savvy and not comfortable. And so at MCW we just basically took the contents of CITI and tweaked it for the community-based members, emphasizing things like the dual roles that they find themselves in. Here is what you can’t do when you are a member of a research team that you would normally do without even thinking about it.
We did that in a face to face interactive training session. It is not perfect. And Michigan is doing this as well. It is not perfect because then you have to have people doing this. The nice thing about CITI is, you just log on whenever you want to log on and it is easy. But it simply doesn’t work for community partners.
And so that is something that they need to think about to protect the investigators. Because if a community partner does something that violates the regulations, the buck stops with the PI.
MS. MILAM: I was thinking about the example that occurs across the country, Ryan, that you have that you talked about. And you have in your information about how a patient who goes to an academic medical center is really there for a clinical visit and may have gotten her Notice of Privacy Practices years before, or may have just been asked to sign an acknowledgment and never really handed the notice, which is the typical practice in a lot of places today.
And she could be surprised to find that her information is part of research, but it is done so possibly legally because of the IRB approval. Consent is out of the mix. But we still have notice in the mix, really. It is a legal requirement. I am wondering, has any work been done around, since notice is the component that is still required, is there a better way to reach consumers when they come for health care? Is there a better way to share this information?
Is anybody doing it in such a way that it is meaningful, and patients understand what really is going to happen with their information?
DR. SPELLECY: There has been some research on using multimedia, using computers and things like that. And all the research with which I am familiar shows that is not terribly effective. I think it is important to start looking at other models of engaging communities that work in other arenas like community engaged research.
There are numerous ways that people do this, with EFIC research, you can do it with random digit dialing, or town halls. The bioethics center at the University of Minnesota, I don’t know if they still do this but they have a booth at the state fair where you can talk with people about issues in bioethics or research, and stuff like that, as a function of community outreach that you can reach people with a corn dog and a question about their HIPAA notification.
DR. KAUFMAN: We did a little bit of work. The VA was trying to weigh doing opt in or opt out consent, opt out being if you don’t tell me you don’t want to be in the bio bank you are going to be in it. And so if you do that, obviously you have to notify everyone. And we asked people — and actually people weren’t that put off about opt out compared to traditional opt in consent. Which doesn’t mean it is ethical. There are a lot of issues with it, I think.
But we asked people about methods. When would you want to be given that opportunity and where? And people did not like e-mail, did not like even being given the opportunity to call in, did not particularly like getting a letter at their house, although that was more popular than e-mail and telephone.
They thought that should happen when they make an appointment or when they see their doctor, or when they are there giving a sample. The VA works a little bit differently, is my understanding. So sometimes there is a lot of sample giving without seeing doctors. That is a common interface. But face to face, which is obviously not the most efficient, at least from a researcher’s perspective.
MR. BURKE: I looked up CITI, because I am the institutional official at Harvard. So if anyone asks, I knew that spontaneously. But it is the Collaborative Institutional Training Initiative. And it allows IRB members to, no matter which facility they are affiliated with, to take a common training with the vignettes and the scenarios that you described. It is torture, as I remember it.
So my question is mostly for David, but for Ryan as well. In this standard we might have of actively engaging a community member with the option to develop a reliable informed consent, of asking them, in your experience what is the depth of understanding they want before they answer the question?
How much do you think they need to hear or know before they feel comfortable saying I want to participate? Is it a shallow level of a description of the initiative? Or can it go deeper?
DR. KAUFMAN: I think it is pretty shallow. But in our focus groups we made a video describing the study, and it is, I think, maybe four minutes long, pretty shallow, surface level. And we started the focus groups with two questions, just sort of overall what do you think, would you be interested. And then second, what questions would you have.
Overall reaction was, a lot of people thought — not everyone, but a lot of people thought this is cool, or this should be done, exactly what Ryan said. More people think it should be done than would participate, although maybe 80 percent it should be done and 50 percent I would do it.
There are a lot of questions, people have a lot of questions after we get their initial read. And it depends on — people want to know what is going to happen to their data and what they are going to get back, what it is going to be used for, what are the limits, and who is taking care of it.
There are a panoply of other questions that some individuals have and others don’t. What am I going to learn is a big question. Everyone wants to get results back. So that is a big one. And who is the data being shared with, that is a big one. I don’t know if that really answered your question.
MR. BURKE: It does, thank you, in particular the pledge to return the result.
DR. KAUFMAN: And I can talk about that ad nauseum, if you want to feel nauseous.
DR. FRANCIS: I want to pick up on — I have got a bunch of questions, but the first one, I want to explore a little more the thing that both of you observed from the data, which is the higher percentage of people who think it is a good thing to have it done, than are themselves willing, or at least to say they are willing to participate.
Actually, I think the University of Washington data, or the data on the Just Want to be Asked study with Trinidad and others. Their experience actually was that the people actually did participate at pretty high levels. So, it may be that people, when they actually get asked whether to participate, do participate, even though they express their unwillingness. I don’t know.
It is sort of a paradox because we have all been talking here in the beginning about it is really important to get this data used. And so sharing the sentiment that it is a good thing to have good public health work done, good research done, but how do we negotiate the fact that it seems that from those two examples of data, more people like the idea of it than the doing of it.
MS. BERNSTEIN: I have kind of got to say, if I have to have a heart attack to participate, I don’t’ really want to participate either. Maybe that is why.
MS. KLOSS: I don’t know what the research says, but that makes intuitive sense to me. If I were just hearing about a project, I would be very interested in the project but not yet willing to commit. So it may be a sequential thing, that I have got to go through step one before I get to step two. And that is just natural human nature or a certain level of skepticism.
DR. KAUFMAN: Yes, what is going to be asked of them, how convenient will it be, how many times will I have to go to the doctor.
MS. KLOSS: Or when you pose that question. I just had a recent experience with a friend of mine relaying a story of being asked to participate in a research study on breast cancer the day she was visiting her surgeon for the first time, trying to work out a course of treatment. She said no. It just couldn’t have been a worse emotional time to be put that question. So I think it is sequential. You have to have gotten to the right place.
DR. KAUFMAN: One little thing I would note is that, when we ask people would you be willing to participate, yes, somewhere around 50 percent, let’s say, that number is pretty consistent, give or take four or five percent across age, race and ethnic group, income, gender, rural-urban. Everyone is interested.
And in the focus groups, when we asked about if there was a result — we did some focus groups on Hispanics and then on African-Americans. And in those focus groups we asked if there was a result that was particular to, for example, African-Americans, would you want that result released? Unequivocal, yes. Absolutely. Why wouldn’t you tell us? Of course we want that information. Don’t withhold that.
There was a little bit of echo of, well, it depends, maybe there would be some stigma. Yes, we absolutely want to know if there is something going on that we can use, or we need to know. So I thought that was interesting.
DR. SPELLECY: So with the emergency research that I am talking about, the description that they get is abbreviated. It is not full, informed consent. It is not supposed to be. And so sure, maybe part of that might be explained by, I think this research is a good idea, heart attacks are a leading killer. But I don’t know if I would want to participate yet.
We do get some qualitative data from that research is well, where they fill in the blanks on why did they say yes, no, or maybe. And there is a myriad of responses to that question. But I tend to view this as a — the parallel I see between emergency medicine research with the exception from informed consent, and the secondary use of data, is that the presumption is that consent has to be waived because the research cannot be done otherwise. We cannot get consent for this.
With these data sets, if we allow people to opt out there is bias in the data. And so then the question becomes, okay, so what do we do about it now. If we can’t get consent how do we uphold respect for persons. And that is where I think going to the community and engaging them in conversations, as well as notification, is important.
MS. KLOSS: Other questions?
MS. BERNSTEIN: Could you talk a little more about what you just said about the second kind of research, that when you let people opt out — in every study where we don’t get a whole, you let people opt out, you are calling people on the phone, randomly dialing, some people don’t answer, some people refuse. And in almost every study that you have like that, you are going to have some kind of sampling bias like that, right? Is it different, what you are talking about?
DR. SPELLECY: Well actually you are an epidemiologist so maybe you can answer this question better than I, but my understanding of this research is that when you are doing population-wide database research that you need access to all of the data, or almost all the data.
DR. KAUFMAN: I think it depends on what your study is aiming to do and how many people you need. You take bio bank as said, we are not going to get a representative sample and it doesn’t matter. As long as we get enough people in the different groups we are interested in, we don’t care if it is representative. There is no such thing as a perfect study. And as long as we get enough people we will have power to do what we want to do.
That is one take. If you are not getting people with — obviously, ascertainment bias, case bias, people participating because they are not healthy or they have strong family histories, those can all bias your data. But there is no perfect study, and there is no study 100 percent of the people will take part in. I don’t know.
DR. FRANCIS: So another question I have is that there is an important difference between being asked and being told. And it sounds to me as though some of what the data that you are aware of suggests the importance of both, that the being asked is not necessarily that I need to be asked specifically for this, but that I need to be asked in some way.
Now the sort of blanket, can you use my data for anything, is pretty meaningless. But maybe it is less meaningless if it gets linked with ongoing sharing of information about what is being done and what is being learned, which is the toll thing.
So as I have been listening all day, coming down, I was pretty much thinking that being asked was just really hard. But I am coming down more and more to thinking that there is real need for interaction between asked and told. And this is me thinking out loud, but I wonder if that resonates with the two of you.
DR. KAUFMAN: That seems like what Ryan and I are both saying. You may not be able to get the perfect ask, but if you explain it up front, if you show good faith, and tell everything you know, then this is what we are doing, and you keep telling, and maybe give people the opportunity to weigh in, opt out, that can buy you a lot.
DR. SPELLECY: I would also throw in the possibility of, instead of asking, to say maybe we can’t ask everyone in the sense of getting their consent. But perhaps we can consult with the community and look at it that way, and get broad community buy-in as opposed to individual consent for some of these projects. Not for all of them. I am not suggesting that we do randomized clinical trials of phase one drugs with community consultation. But for data, yes.
MS. KLOSS: Have you seen any characteristics of a community that enables these kinds of community health initiatives to sort of germinate and grow? I know for example Wisconsin has done a lot with community data sharing, and has some years of work at this. Do you see that this is paying dividends in a more engaged community? And if so, what lessons are there that this committee can pick up on, and move forward?
DR. SPELLECY: So just from my own experience, when there is an established community organization, you have to develop them with research. But the ones that have been around and have been engaged in research are obviously better equipped to do this. They have the infrastructure and the people that know how to navigate this.
But we build new partnerships every day with people, and it just takes longer. But when that relationship is sustained and ongoing, and we have to meet a lot more, I have to meet with people a lot more when I do community engaged research than when I do other kinds of research. When I meet with other academic partners, we meet, here is what we are doing, here is how we are dividing the work, let’s go.
With community engaged partnerships, we meet ongoing, we have conversations, we have coffee, you really are building much more of a relationship. I think in part because when I am collaborating with other people at the college, we are all there for the same reason. When I am collaborating with members of the community, we are not all there for the same reason. We have similar interests and so we need to hash that out.
But there are certain things that they do that don’t dovetail with my interests, and vice versa. So we have to build that common ground.
DR. KAUFMAN: The only thing I would add to that is observing basically what Paul and Malia said earlier, having community members participating in the work, on especially recruiting. That is the way that the Strong Heart and the Goga Dan(?) studies that I have — earth and — that is the way those studies were able to function, was to train the people who knew the people, and have them train us about how to go into those communities and how to stay there.
And it is, I think, really hard to do this kind of work on a two-year grant. Without the guarantee of renewal, longer term stuff is unfortunately important.
MS. KLOSS: Any of the committee members have any more questions? I actually have a follow on.
DR. MAYS: I will ask another question about the community engaged approach you were talking about, the community based participatory research is one thing and community engaged is another.
Again, I kind of want to go back to the notion of the IRB. In terms of those types, when we are trying to get data, is there any — what was the word I want? Concern, or is there anything that is within your charge that can say to the researcher about what they should do.
If they are going to invoke a method, and the method is, let’s say, community based participatory research, is there anything in the charge in which you can say, as part of this, you left out what you are going to give back to the researcher? You think about methods quite a bit in terms of very traditional things. So I am wondering if this falls under your charge to discuss give-back of data or some explanation at the end of the findings.
DR. SPELLECY: That is a great question. My answer is yes and no. I don’t think it is the role of the IRB to be the scientific review committee and say, look, in your protocol you say you are engaged in — so community based participatory research, I don’t tend to use that word because community based participatory research is a very well-defined set of engagement procedures along the way. And it is very easy to do wrong.
Whereas, community engaged research is sort of a well, you know, you did a little of this, a little of that, but it has kind of got the flavor.
DR. MAYS: But I am actually asking about community based participatory research. That is what NIH is funding, and that is what is going to get brought to you as an IRB. And it has those kind of implicit assumptions that those things at the end are kind of supposed to happen, but a lot of times we run out of money and run out of steam. So that is why I am asking whether the IRB has a role in this.
DR. SPELLECY: NIH is funding CBPR, but NIH is funding broad spectrums of things that don’t quite rise to the level of full CBPR. I don’t think it is the role of the IRB to say you are not doing this, you are not adhering to all of the tenets of CBPR. So, not in the sense of a scientific review committee.
However, it is the role of the IRB to weigh the risks and benefits of the research. And if one of the benefits of the research potential is return of information and you don’t do that, it is difficult. Because if the funding runs out and you close the study, well, yes, you deviated from your protocol, but you closed the study. So it is not entirely clear what we can do.
One thing we could do is ultimately, if it happened a lot, we would be forced to report it to OHRP as continuing noncompliance. You mentioned NIH, though, which throws a wrench in it because IRBs are charged with reviewing the funding proposal and the protocol that gets submitted to the IRB for consistency. So we do have that.
If it is NIH funded we can say, because NIH asks us to look at what the researcher told NIH they are going to do and what they end up doing, and there is always slip in there, especially with CBPR, because you submit the funding and then you find out later through the formative phase, well this is what the community wants to do.
But if it is dramatically different — and I have not yet seen an NIH funded study where it is dramatically different so I don’t know what I would do if the funding didn’t match up — I would probably call my institutional official and ask him what to do.
MR. BURKE: You are right. You have more degrees of freedom, depending upon who is funding. So it doesn’t rise to scientific misconduct. But it does rise to continuing violation.
DR. KAUFMAN: What we are seeing in the clinical genetics meetings is talk now of a minimum set of genetic results that possibly should be returned to anyone in the study, where genotyping, whole genome, or genome-wide sequencing is going on. I think there is some movement towards obligation for return.
DR. SPELLECY: In adults?
DR. KAUFMAN: Yes.
DR. SPELLECY: The reality is, with whole genome sequencing, we are not seeing a whole lot of it anyway. It is so new. But in what we are seeing with adults, we are seeing that they are wanting to return some, or all. With kids it is different, because with kids, for example, when you sequence their whole genome you could find markers that will not result in anything that you can treat while they are still a kid.
Huntington’s Disease — there is nothing you can do to prevent it. There is nothing you can do to treat it when they are still a kid. And so do you have an obligation to tell the parents, when the kid is under 18, how you handle that? The way we handle it is, when they are 18 we re-consent them.
It is a good question. I think that you are seeing at the meetings, because fortunately, they are thinking ahead.
MS. BERNSTEIN: One of you mentioned about the DOJ process for Certificates of Confidentiality. And I wanted to know if you could say a little more about how they differ from what we are doing at HHS. And how it is exactly they are enforcing that. I am trying to imagine how they are enforcing that, whether they just get their guys not to ask you for the subpoena in the first place. They are the ones making the subpoenas, right?
DR. KAUFMAN: My understanding is researchers can still be subpoenaed, but their response to that subpoena is the certificate.
MS. BERNSTEIN: But the difference is that they are required to respond, I am not going to give it to you, rather than that they have a choice?
DR. KAUFMAN: DOJ researchers with a certificate cannot divulge that information for those purposes. You can see, if you are doing criminal justice research, it is going to be pretty hard to recruit a population of, for example, prisoners.
MS. BERNSTEIN: We have the same problem with substance abuse patients, who are violating the law. It may be that HHS just couldn’t imagine any researcher that would want to give up their data for that kind of purpose.
DR. KAUFMAN: It has happened. The Northern Atlanta case, a researcher gave up information. I don’t remember all the specifics.
MS. BERNSTEIN: That was covered by a certificate?
DR. KAUFMAN: A violent crime had been committed and schizophrenia was involved. A researcher had this data and it was subpoenaed, and the researcher gave the data up, and —
MS. BERNSTEIN: When it was protected by a certificate?
DR. KAUFMAN: Yes, I think so.
MS. BERNSTEIN: It could happen. It makes me want to go look into that some
DR. MAYS: You reminded me it was one of the things I wanted to come back to. My understanding about the certificates, the HHS certificates, is that what they will tell us is that we are not required to give up our data but what it is, it has not been tested to know if we are going to be protected. Meaning that if I am subpoenaed and I refuse to give up my data, what can happen is, I might end up in jail.
It hasn’t been tested for HHS. I have studies with certificates, so part of what the university tells us is that there is a way in which the data is protected. There is a way in which we are not protected, because there hasn’t been a court case yet that says if we don’t give up our data we won’t get in trouble.
My understanding is that for DOJ, that you don’t have to give up your data, and that the subpoena does not — there is some language they use — supersede. Anyway, it is like with the Department of Justice I don’t think you go to jail. With the other, I think that we are liable to —
MS. BERNSTEIN: Are you suggesting that is because the DOJ one has been tested in court, and the HHS one has not?
DR. MAYS: I don’t know the difference. I just know the advice that we get from university counsel. I did have a challenge for some data that I had. And I refused, and I was told the grounds upon which what could happen. But it went away, so it was okay.
MS. KLOSS: I remembered my question, and it relates to infrastructure. One of the recommendations in our community health information report was a conclusion that community health projects or initiatives could be aided with infrastructure. And earlier this afternoon, Dr. Silka suggested that universities could be a source for that infrastructure for community, both in terms of data skills and research skills, and some continuity within the community.
You both are sitting here representing universities. Do you agree with that as a potential source of infrastructure for communities? And are you living examples of that?
DR. SPELLECY: I am not a living example, but my institution is in that we have partnered with community organizations and set up things like health clinics in the community center. So for example, one I am familiar with, the south side of Milwaukee has a large Latino population. There is a large group called the United Community Center, which serves that population. We opened a health clinic on site there, because the people far above me in the institution recognized that if we want to be able to provide care to the members of this community, we have to do it in the community.
This was fascinating. We and other health care organizations opened up clinics in the community, and nobody went, and they were unsuccessful. When we partnered with the community organization and opened the clinic in the community organization, in their building, it was successful. I could only speculate why that happened.
All I know is, when we opened it and put our sign on the front door, it didn’t work. When we had their logo next to our logo at their building, it worked, or it is working.
DR. KAUFMAN: I think it can go both ways. I think universities can be extremely well trusted community members. And I would not say — there are plenty of examples at Hopkins where that is the case, a lot of good partnerships. There is also a huge divide between the medical campus and the community, the Kennedy Krieger stuff with the lead — a huge disaster.
I think the potential is certainly there. When we do these focus groups around the country, we ask some questions about who would you want to be a steward of the data. People see us and we tell them, how about Johns Hopkins? Johns Hopkins would be good, better than NIH. So I think there is some general trust, that we are good actors. But obviously it has to be done right. Just putting your sign up is probably not enough.
DR. FRANCIS: I would just ask the last two questions that I have asked other people, too. I am just curious about your thoughts on them. One of them is, in terms of dissemination, telling people about what you have learned. It is clear there is a lot of — we heard in the last panel and you all have emphasized it too, the helicopter image.
That if communities don’t learn what has been learned — and I am now thinking not about returning research results to individuals, but conveying to this community, or to communities — what has been learned from this community, that it is really important to do that.
At the same time, several people this morning really held the strong view that if the community wants the information shared only within the community, or if the community doesn’t want it shared at all, it doesn’t go public. Now that was with respect to the example of Alaskan natives and Native Americans specifically.
Maybe that is just an exceptional case, but I would like, if you have any thoughts about the role of — Jerusha Peterman made the kind of intermediate point of tiered disclosure, which is very helpful. I take it what she was saying there, you let the community know first, so they can deal with the fact that potentially stigmatizing information, or whatever.
But if the community really opposes, if you go to the first tier and the community says no, if you have any thoughts about how we should think through that problem, that is one. The other is — I will just go back to the question I asked at the end of the first session, which is, is it different if the federal government has the data?
Should the feds be more protective about letting others use it, when it is data that the federal government has because of the trust issues? Or for any other kind of reason, should the federal government be — should we as advising HHS, should we say, if it is HHS data, you have got to involve communities or it isn’t ethical?
MS. BERNSTEIN: I am assuming you mean data that is collected by the federal government voluntarily, and not regulatory information that you are required to turn in, tax data, census data.
DR. FRANCIS: We could ask it about both.
MS. BERNSTEIN: And there are different reasons why.
DR. FRANCIS: Yes, but you don’t have a choice if it is Medicare. I am thinking about CMS data in particular. Or VA data.
DR. SPELLECY: So what if the community says no? And then the follow up is, should the feds be more protective?
DR. FRANCIS: It is two separate questions.
DR. SPELLECY: In response to the first question, what if the community says no, I am of the persuasion that if the research is done, it needs to be disseminated. But I want to say that if we are asking that question at the end and they say no, we have messed up somewhere along the road.
So if we are engaging the community, especially these Native American communities — I am sure you are familiar with the Havasupai. That wasn’t community engaged research. And out of the blue, the woman, Tilousi, saw that they had used their blood samples for schizophrenia research and the Bering Strait Theory and all this stuff.
I presented this at my own institution at an ethics ground rounds thing. And a researcher said, yes, but this is just research. Do we have veto power over everything? But the community engaged folks in the audience said yes, but there is still a way to do this in a respectful way.
And so I think that if we come to the table and respect communities and respect individuals and work collaboratively about what we are going to use this research data for, we can head off a lot of those problems. The Havasupai case with ASU just didn’t need to happen. That is the extreme case. And I don’t know that it would happen again. But it sure could have been prevented with some better community engagement.
DR. KAUFMAN: I agree with that. I think exactly what you said. If you get to that point, you probably screwed up and you should not release the results. You need to be talking about what those results are well in advance. Obviously you cannot predict everything you are going to find. But you know at least what your hypotheses are, setting out.
Communicate those. It is the kind of thing to come out. We don’t want to do that research. That is good to know, we won’t do it, then. I do think that it is a very unique situation, but that said, if you ran up against something that serious, it would probably be worth considering.
DR. SPELLECY: I think you made a very good point. Do you want to do this kind of research? No. Okay, then we are not going to do it. I think it is better to figure out we are not going to do the research in the first place, than to do the research without telling people what you might do and then have the data in the results later.
DR. KAUFMAN: What was the second question?
DR. SPELLECY: Should the feds be more protective of data, especially with the issue of the trust to the feds.
DR. FRANCIS: Yes, should we be more worried about community engagement when it is the feds that are the stewards of the data, or the states, that have some sort of governmental responsibility for the data? And maybe, to go to Maya’s point, some of the data that the states and that the feds get, they don’t get it, it’s not voluntary. This is regulatory.
And so we are not talking about people who voluntarily shared. And with states often, if we are thinking on the public health side, it is data that they got pursuant to public health powers, and when Utah Medicaid left its computer unprotected in the disastrous way it did recently, there it is a HIPAA covered entity, because it is a payer.
But if it had gotten exactly the same data from the prospective of the public health powers, well, you know, it is just state law that we are looking at there. And so I guess, do you see more in the way of responsibilities when the data were collected either voluntarily or, even more importantly, non-voluntarily in the hands of the government?
DR. KAUFMAN: Leslie, I don’t know if the responsibilities are different or more, but I hold up that Texas blood spot thing all the time. Do you need to do focus groups and town halls all over Texas in order to do that? Maybe not. But some measure of — it may just be CYA, honestly, but some measure of — maybe you do want to do a couple focus groups and see how people feel about it.
And if people are really opposed to it, then you need to think about it. And if they are not, then you need to still let people know that you are doing it. What does it cost to let people know? It is actually a neat thing. Oh, I am participating in research. My baby is participating. Well I don’t want my baby too. Fine. We won’t use your newborn’s blood. What does that cost? It costs something.
But I think some minimal common sense letting people know, that is engagement. It is not full on engagement, but it is probably worth doing, especially for anything large scale, state, national.
DR. SPELLECY: So there is an interesting parallel here with newborn screening in Wisconsin. There was a scandal year’s back involving the University of Wisconsin Madison Medical School — not my med school but the other med school, which is a state med school. And they were sued because, as part of the newborn screening, they were doing some research on cystic fibrosis.
It is complicated, but they didn’t return the results to the parents, and the parents found out later that their kids had CF, but that the researchers knew all along. Ultimately they were fined because of sovereign immunity, apparently. But this was something where there was technically a pamphlet in the myriad of pamphlets that moms were given at birth, where they could call this number to opt out.
But the general sense was, it never really went anywhere, because they were acting as university employees. But they did end up suing the medical college because we were tangentially involved in the research, although that got thrown out too.
But the general sense was that this sort of vague opt out was not enough. This was newborn screening that was done via public health on all newborns. And so they obtained it really without consent. But the opt out in that case wasn’t enough. But this was a situation in which you had a result that could have been useful to parents.
If it is data that is just sort of Medicare data and CMS data, and things like that, there are two questions. What do people want to know, or do they want to be asked. There is that end of the equation. And then there is the, what is the obligation of the federal government. And I don’t know that I have a great answer for you other than those are different questions.
The federal government might have certain obligations, given how they got this data. But at the same time it is important to know whether people care at all in the first place, or how we can help them care less, by asking them.
MS. KLOSS: Any other questions? Anything you would like to add? Thank you very much. In theory we are supposed to ask for public comment, but I don’t see any members of the public here who aren’t staff.
MS. BERNSTEIN: We don’t have any people on the phone, so I am guessing no one wants to make public comments at this time.
MS. KLOSS: So our hearing is adjourned. Again, we thank you very much. And thank you to the members of the committee and the staff, who organized the day.
MS. BERNSTEIN: The panel is adjourned for the moment.
MS. KLOSS: Let’s spend five minutes talking about how we are going to —
DR. FRANCIS: Our agenda has us staying for a few minutes to talk about how we will organize our day tomorrow.
MS. BERNSTEIN: We have a few more minutes to think about how we eat the elephant tomorrow. In small bites.
DR. FRANCIS: I think what we should do, we should identify families of problems. We should identify families of issues, and talk about what recommendations we would want to make for each. My strategic suggestion for that is actually, each member of the subcommittee and member of the staff, if they are willing, but members of the subcommittee — may way of thinking about this is, we have heard this incredibly rich discussion.
Linda and I could decide after dinner that we are going to sit down and make a list of issues. Or we could try something quite different, which is we could ask each member of the committee or member of the staff to identify one issue that they think, assuming we are writing a letter that is going to look at this general question of next steps for community data use, what would be one area and possible recommendations.
Which could just be, this is an issue we need lots more thinking about, that you think ought to be part of those recommendations? So have a sort of bubble-up strategy. You see what I am suggesting? Things to think about for tomorrow. Anything you want people to think about before tomorrow.
MS. KLOSS: I thought it would be useful to go back to the CHIP report and look at this section where we teased out the issues that needed to be advanced, and perhaps at least began using those as frameworks. Use those to frame our discussion.
DR. MAYS: One of the things that I was thinking about as we were going through the hearing is those things that for the community are in the hands of the federal government, who we are the advisors to, and those things that may just be out in the universe that we can’t do a lot about. For me, two came up that I thought, oh, we actually have teeth to be able to move ahead on.
And one is, what happens relative to IRBs. Remember, the IRB is under a multiple assurance that really comes from the federal government. That is why I started asking some of those questions. I started thinking about are there things that we could suggest that they reconsider.
Like I really thought it was fascinating to know that they aren’t to consider the long term. So the question is, can we start talking about that? I don’t know exactly if we are ready for a recommendation, because I don’t know how that changes it. But that to me seemed to be a big thing.
The other that came up for me — and I am always going to have this — is the acceptability of the CITI training, and the recommendation of making changes to that relative to thinking about these more public health community issues also is something that is what is accepted by OHRP relative to the university saying that we have met our partner education.
So to me, thinking about those things that we actually have some power behind us to do, it is harder to me. But this is, we could get some potential change.
MS. KLOSS: The CHIP report does tease out conclusions in two ways. The more general kind of guidance for communities who are undertaking this, and then the areas where there is a specific recommendation that might be advanced to HHS. And I think we probably want to do both.
Because I can certainly see going through what we have heard today and drawing out principles that advance trust in data, or trust in how data is being used based on all of the testimony we heard today. I think it would be wonderful if we came away from tomorrow morning with a list of those principles. I don’t know that we even need to get much more detail than that.
We have got stories that have been brought forward to us, and testimonies that elaborate on those principles. For example, community engaged research, certainly I would have to go back to this morning, there were just so many of them. So I see taking two paths. Certainly the one you are suggesting, Vickie. But also doing what CHIP did, which was give some specific recommendations that go beyond what the initial report did with regard to trust of data.
And probably the third is, to see where we have got some gaps. Because there were certainly some avenues that you were looking for testimony on that perhaps we didn’t have. And there may be a couple of gaps that could be filled in through interviews, or through conversation.
MS. BERNSTEIN: The subcommittee could consider whether they are not ready to make those yet, and would want to continue down that line, depending on available resources. I am looking over there at Maya, at whether you want to try and get those people that were at the other meeting or try and get some other group of people where there is a gap to talk to us.
That is up to you. I am not sure whether you feel ready, or you feel like there are some gaps, or we could talk about that some more tomorrow.
DR. FRANCIS: We may be ready on some things and not on other things.
MS. GREENBERG: As the exec sec, one of my first questions always is, do you feel you need to hear from some additional people, given what the questions are that you are asking? You can always hear from additional people, and there are always more questions. But because one hearing — although we have been a little bit more going from — we used to have a series of hearings. And maybe we had more money in those days, or more time, I don’t know.
But it seems now, it is not atypical, if we have been able to define our questions well enough and get enough testimony that we can go from a single hearing to maybe some recommendations. But that’s always the first question in my mind, from a planning point of view, as the executive secretary.
There are some other ways. I think our budgeting is typically only one stand alone hearing per subcommittee per year. I just don’t think we have — I think that is pretty much what our budget is for. But then we have hearings sometimes associated with a full committee meeting. I think the standards subcommittee is meeting in June.
And then we also can do other, if you want to hear from one or two people you can sometimes do that through a break out session during a full committee meeting. Or you can do some kind of webinars as a possibility too. We haven’t done that as much as we probably should and can in the future. So that is one set of questions.
You mentioned going back to the CHIP report, which obviously this came out of. But also, I couldn’t help but think about the stewardship primer and whether we might, there might be some who want to update that or add a codicil or something, particularly on community engagement. I am sure that mentioned community engagement to some degree, but I don’t think very much.
MS. BERNSTEIN: As long as you don’t mention privacy. Let’s just say that is a sore point for the subcommittee.
MS. GREENBERG: All right, I won’t go there. I would certainly think we should go back and look at that. And then also, I was just wondering, Maia, if you knew or if anyone from the department, like maybe Amy — but we commented on that Advanced Notice of Proposed Rulemaking. When was it, in the fall that we commented?
MS. GREENBERG: I haven’t heard anything about it since. So I wondered, are they still processing the results?
MS. BERNSTEIN: I will be careful about what I say because it is a rule making process. What I know is that they had over 1,100 comments. The average length of the comments was 30 pages each. A lot of comments. There is a Notice of Proposed Rulemaking, the next step, that is in process.
MS. GREENBERG: So there probably will be a next step.
MS. BERNSTEIN: I am probably not supposed to say that, but it has. The typical — they are going through the typical steps that happen. It is a very, very complex rule, and as I say, there were many, many comments on it, much of which dealt with multiple topics in the rule, and were not the cookie cutter responses that you got. Many institutions that are covered by the thing responded each with individual comments. It is a lot for them to process.
MS. GREENBERG: I can only imagine, yes. There was some testimony, I think it was this morning, that related to some of the issues that were in that ANPRM.
MS. BERNSTEIN: One can imagine that the Administration would have maybe liked to get this out before the end of this Administration. But the timing for something that complex, I would, if I were a betting person, would not bet that they would get through both the proposed with a comment period, which is likely going to draw many more complex comments, and a final rule, by what would have to be about September, October, before they set the rulemaking process down before the election.
That is just sort of realistically about how the process goes in an election year. I don’t know that I can say more than that.
MS. GREENBERG: That is fine. And I guess, although I know you have pointed out that although the comment period for that is long gone, but the committee is always in a position to comment on things. And so I don’t know whether there might be some timeliness to —
MS. BERNSTEIN: A couple of questions, which I don’t know if they came up, whether those, for us, whether they come up in the comments. I don’t know if you read all the comments, but somebody may have, said you should have been moving in this direction and you haven’t proposed anything to change this thing that Vickie proposed.
And the committee could at any time make recommendations to the Secretary. And knowing that that rulemaking process is in process —
MS. GREENBERG: Right, then it would be good to get that out sooner rather than later.
MS. BERNSTEIN: If you wanted to make recommendations about that, I think you could make a narrow letter, just dealing with those things. We would have to get it through the process of a committee, and we have done meetings by phone that are public meetings before, if it doesn’t happen at the time that we are having the meeting in person. We are just processing. But you could make a narrower thing just to focus on those, and then deal with the other issues at some later time, if you wanted.
DR. FRANCIS: The ANPRM did not raise the question of long term community risks. The ANPRM did not raise the question in weighing the risks and benefits, IRBs are not supposed to take into account the long-term community risks.
MS. GREENBERG: I wasn’t aware where that came from, exactly.
MS. MILAM: Not following on the ANPRM discussion but back to the other discussion, and really sort of building on what Linda and Marjorie were talking about, I am on page 30 of the CHIP report. One of the needed elements of infrastructure is a privacy and security framework to guide communities on using local data. And the framework we have available to us, certainly there are a lot of frameworks available.
We referenced the stewardship primer, but there was an original letter, I think in 2007, that laid out stewardship more fully. And it, I think, noted that the population health application could be done at a later time. So there is already a reference within it to pick it up later.
And in terms of principles, what we are hearing from communities is that they are challenged in the application of this broad framework to the population health scenario. We understand by regulation how it is applied in other situations, but we heard today a lot of suggestions of how certain of those principles might be applied. We may find we need more information, or we need to hear from more actors in that space.
But we certainly did hear about how a number of those principles might be shaped in different criteria and components. And we could use that framework as a way to sort of figure out what have we spoken to, what is there clear guidance on, and where are there gaps.
DR. FRANCIS: Could I add to that, that I was also looking at envisioning a federal role, the next page of the CHIP report, page 31. There are a whole list of things that we thought the federal government can do. Just to illustrate the first one, facilitate and provide resources to strengthen communities’ capacity to collect level data.
Well one of the points we heard was that short-term funding cycles lie in serious tension with that. I think we heard plenty today to support that claim. So we could go — maybe a nice little structural thing would be to go back through the bullet points on page 30, and then again, the ones on page 31, and look and see what we have, and maybe take the first hour and do that.
And then I will re-read the stewardship report tonight too, and see if there are bullet points there that we could pull out.
MS. MILAM: I don’t know if it is feasible, but it would be great if we could have hard copy or have it on the screen for people tomorrow. It is a nice existing structure. It is a framework already in place that we have already spoken to.
DR. FRANCIS: Which one? The stewardship?
MS. MILAM: Yes, from the earlier letter, not the primer. The earlier letter, I think, has more information.
DR. FRANCIS: I will get it downloaded from the website tonight.
MS. GREENBERG: The original report that started off as a secondary uses report and then kind of evolved.
DR. FRANCIS: The other homework was, I really would like people to think about something that is an area they think we — I think one of the things that happened with the NHIN series of letters and that series of hearings was that there was a kind of mega letter about areas, and then there were much more specific letters about particular areas.
We could think about a mega letter based on this. And then where we see a gap we could say, this is an area that requires addressing, and we are going to hear more.
MS. KLOSS: So we have two blocks of time in the morning, 9:15 to 11:00 and then a break, and 11:15 to 12:30. It might be a facilitation of our work to focus the first part, start with that framework and see how well that is taking us along, not worry about recommendation 30, and then after break think about next steps, how far are we going to get.
MS. BERNSTEIN: We had to publish an agenda, Marjorie, but I don’t know how much we have to stick to it tomorrow. Depending on how the flow of the conversation goes, I think we can adjust as we go along, depending on what the co-chairs want to do.
MS. KLOSS: Paul, I have a question for you. I wasn’t at the quality hearings, and you had a half-day session following that. We modeled our agenda after what you had done. How did you use that second day?
DR. TANG: Basically, essentially we outlined the findings, the conclusions and the recommendations. We are about to approve our close to final letter. It is a handful of pages, and it is approximately that outline. It is what did we hear, what do we conclude from that, and what are we recommending.
MS. GREENBERG: I know that Matt was keeping a running record of what people were saying on the PC. Didn’t you and Blackford offline — then we had the full committee meeting, so you were around for a few days. Although Blackford had to leave.
DR. TANG: Blackford left, so I did my lumping, and reported that out, actually, at the full committee. And I think it resonated with the rest of the members who were there.
MS. KLOSS: We thought we might want to start by looking at the conclusions from the CHIP report with regard to where it needs to advance trust. That is outlined pretty well, and put that out, what we learned, that helps us move in that direction. And then also look at the framework of the stewardship letter.
MS. GREENBERG: Are you talking about the guiding principles in that letter? This is a December 2007 letter.
DR. TANG: It seems like the issue is trust in relationship are there. In fact the one we had, it was at NCHS, I remember the table. This whole involving the community has come up a number of times. I think what was impressive is the clarity and the examples.
So actually this Indian tribe really crystallized the notion of how do you identify community. They have a really identified community, and they have the benefit, it turns out in this case, of real governance, mandated governance and their representative. And it turned out to be a model. And then the notion of small. We are all actually small tribes. We just don’t know it.
I think it helps us restate or perhaps consolidate and state clearly what would need to be done. You could see a number of recommendations that I think are more clearly delineated following this.
MS. GREENBERG: I remember a hearing. I guess it was the second. That was the one when we were at NCHS, the second one for the CHIP report, which was the privacy. That was at NCHS. I don’t know if it was that hearing or it was a different one that the subcommittee had, where it just became so clear.
This has been going on now ever since HIPAA really, because I hear various testimony in my head, going way back. But when you try to avoid really telling people what is happening, it comes back and bites you. For a long time, I think that was really the — not out of malice, but just well, we will never be able to do the work, or people won’t understand, or whatever.
And I think I have been hearing increasingly, and certainly I heard that today, that it may be quicker initially to just try to get on with it and not really — but you will fail. That upfront work, that communication. And I think in some ways, — and someone was asking about incentivizing — it goes against a lot of the incentives that are out there, to publish, to get stuff out, the idea of doing a lot of research and work, and then the community says we don’t want this published and then not publishing it.
They were saying, if you get to that point you have failed along the way. But I still think that we are hearing the cream of the crop here. But there is a lot to think about.
MS. KLOSS: I think we would be well advised also to approach this discussion from the consumer perspective, from people who are being served by this data use, and get their heart and be very anchored in that.
MS. GREENBERG: And there are some linkages there with the quality.
MS. KLOSS: There are. I heard a number of those.
MS. BERNSTEIN: We specifically did not ask for individual consumer advocates, for example, this time. We wanted to see what the more empirical evidence was specifically. But we have heard in the past at this committee where we are permitted to look back at other things. Not maybe specifically on community based research, but on other kinds of uses of research, uses of data that are price related, from consumer advocates over the years, certainly since I have been staffing the subcommittee for seven years. And we can certainly look to some of that stuff if we find something that is useful in there.
MS. KLOSS: We heard a lot about people’s preferences. That is what I heard.
MS. BERNSTEIN: I understand.
MS. GREENBERG: Are you here tomorrow?
DR. TANG: Yes.
MS. GREENBERG: They haven’t seen the quality letter, but you might —
DR. TANG: We could share it.
MS. GREENBERG: We could share the current draft of it. Would that be helpful to people?
MS. BERNSTEIN: Sure. I just meant if you were identifying a gap that we should —
MS. GREENBERG: I could e-mail it to people.
DR. TANG: I think part of our news is that this turns out to be ahead of the game again. This is ahead of the curve.
MS. GREENBERG: That is what I was trying to say.
DR. TANG: It’s not even on the landscape. We are now just trying to get into health information exchange. But we haven’t gotten into touching community, dealing with things that affects us tribes. And I think that hasn’t been described that way. Yet the science is advancing so that we will become tribes.
MS. BERNSTEIN: That is interesting. That is true.
MS. GREENBERG: Multiple tribes.
DR. TANG: That might be the news. That is the news that is the leading news. And actually we have been anchored in real live tribes, and explain how that does it there, and what is incumbent on us to deal with the provisions.
MS. GREENBERG: I think you were the last person to comment on that quality
DR. TANG: What do you mean I am the last person?
MS. GREENBERG: If I was going to share with them the status —
DR. TANG: I was the last one, yes. I responded yesterday.
MS. GREENBERG: Monday. So should I sent that to them?
DR. TANG: As long as you don’t redistribute.
MS. GREENBERG: No, just it’s a public meeting.
DR. TANG: It is our draft.
MS. GREENBERG: Okay, it’s just a draft. I will just send this to people around the table.
MS. KLOSS: Okay. So we are ready to adjourn, and we will reconvene at 9:00 a.m.
(Whereupon, at 5:13 pm the meeting was recessed to convene the following day at 9:00 a.m.)