Anthropology, Genetic Diversity, and Ethics 
A workshop at the Center for Twentieth Century Studies 
University of Wisconsin-Milwaukee  
Henry Greely
[Participant Information]

Good morning.  My name's Hank Greely.  I appear to be your chair for the entire morning.  With me on the panel today are two other speakers, and I'd like to start by introducing all the speakers, and go into the talks.  Charmaine Royal is a geneticist who is a postdoctoral fellow in the Office of Bioethics and Special Population Research at the National Human Genome Research Institute at NIH.  She's a Research Associate in the Division of Epidemiology and Behavioral Medicine at Howard [University], and she's right there.  Morris Foster is an Associate Professor of Anthropology at the University of Oklahoma.  He's the author of Being Comanche: A Social History of an American Indian Community.  In recent years, he's done substantial research on the ethical, legal and social implications of genetic research with Native American communities.  I'm Hank Greely; I was surprised to see in the pamphlet that I am an attorney for Stanford University.  [laughter]  I don't know whether I would be more surprised, or whether the University's legal counsel would be more surprised!  [laughter]  I know they would be more appalled at the thought!  I'm actually a law professor, not a practicing lawyer, and a professor by courtesy of genetics and co-director of Stanford's Program in Genomics, Ethics and Society.  I haven't practiced law for 14 years, and am very happy about that!  [laughter] 

So, I guess in virtue of being chair, I get to go first.  I'd like to talk about issues in group consent.  Group consent is a concept that has intrigued me for about five to six years now, one that I think is quite important; but also, like most important things, not simple.  Let me first talk about why I think group consent is essential, and then secondly talk about the very real problems with group consent in many contexts.  The idea behind informed consent really has two branches.  One branch is protection of research subjects from harm.  The idea is if people know fully what they're getting into, they will be more likely to make decisions that will protect themselves, or that will adequately weigh for themselves the balance of benefit and detriment, rather than relying solely on the judgments of an IRB or the researchers or someone else about how the balance fits.  The second branch of the justification for informed consent is one of respect for individual choice in the [inaudible]; that, by asking someone their permission, you are respecting their rights as a person.  In my view, in many sorts of research done with pre-existing, culturally significant groupings, that 'research subject' should be viewed not just as the individuals who take part in the research, but as the group itself.  If you're looking at the ancient migrational patterns of the Irish, your research subjects may be individual people from Ireland, but the research subject more broadly is the Irish, the people who participate in the study, and people who don't participate in the study.  They are all potentially affected by that research; they all could face some harms or some benefits from that research.  They all, I think, should have a right to a say, if feasible, if possible, in whether the research should go forward or not, for both reasons.  They are all potentially harmed; harmed as individuals and harmed as groups, and potentially [for benefit?].  And just as individuals should have the right and self-protection to draw the balance between costs and benefits, so should the groups.  And also, the groups, I think, socially defined human groups, have an identity that demands respect in terms of some level of [autonomy?]; that the group itself should be viewed as the entity that proper respect, for which proper respect requires asking permission of them as a group, and not just as certain individuals drawn from the group. 

The justification, I think, for group informed consent is strong.  It gains, I believe, for science in a further more pragmatic representation, in that science, and particularly science dealing with human subjects, depends on the willingness of human subjects to participate.  [It] depends on the willingness of human subjects -- their friends, relatives, neighbors, and political representatives -- to both funding for science, and that the embitterment and resentment of human subjects who have been ill-treated is in the long run a very bad thing for science.  It's not only, I think, unethical, but it's very short-sighted.  The Tuskegee research, that continued for over thirty years and forty years in Alabama, continues to this day to have some implications for the willingness of African-Americans to participate in research.  It's not the entire explanation, but it's certainly a part of it.  Unethical research, or research that leaves research subjects feeling embittered, cheated, mistreated, is not only unethical, but it's bad in the long run, I think, for being good scientists.  Research with groups, or research that purports to be about groups, that is done over the opposition of groups themselves, I think has a strong potential for a similar negative reaction; a similar sense of resentment and bitterness coming from it, some of which I think we already see.  With Native American groups, reacting in part to the legacy of several hundred years of exploitation by scientists and non-scientists, without the notion of [inaudible].  So I view the argument for group consent as quite a strong [argument].  On principle ground, on pragmatic ground, it seems to me essential for researchers operating with groups to try to make sure that they operate with the permission of the groups. 

Having said that, there are difficulties.  I think the difficulties are usually surmountable, but they're only surmountable if you know what they are before you start.  And I want to talk about four different difficulties.  The first is one we touched on yesterday, defining what the group is.  If you're doing research with a population that speaks a [Nadevi?] language in the southwestern part of the United States, is the appropriate group the tribe whose reservation you're doing the research on?  Is it a larger tribal group, so that if you're doing research with one Apache tribe, is the appropriate group to consent from that tribe?  Is it other Apaches?  Is it all [Nanmey?] speakers, [inaudible] among others, in the American Southwest?  Is it all non [Nanmey?] speakers including those in Canada and those in Alaska?  Is it all Native Americans?  All are to some extent potentially -- and this is Eric Juengst's [nested form?] -- are potentially, to some extent, affected by the research [inaudible].  Where do you decide where you [inaudible]. 

Related to that is another point [inaudible].  The groups don't overlap with what are called the demes, [inaudible] congruent with the demes.  There is no genetic population that in any meaningful sense can be identified as genetically Irish or [Zulu?] or Apache or anything else.  There are overlapping -- there's a large spectrum of frequencies of different alleles, but overlapping a number of different dimensions.  But both of these problems about defining the group are, I think, [inaudible].  The issue of the demes not overlapping perfectly or not being perfectly congruent with the socially defined groups is, I think, clearly true, but there is some overlap.  And I would argue that, although you need to be careful to avoid reinforcing the idea, the cultural groups have a [inaudible], that the cultural groups are your only starting place, the only place to look for someone to ask for permission from.  If you don't know the breadth of the deme, or even whether a deme meaningfully can be defined to exist, you can't ask the deme for permission.  You're left with, for want of anything else, the culturally defined group.  How far up the group that, the nested hierarchy of the group is, [inaudible].  You go to the village, you go to the county, you go to the country, you go to the region...  But that, I think, the answer [inaudible] pragmatically and ethically appropriate: you should talk to the people you're doing the research with [inaudible].  You consider yourselves primarily this group ofHutterites, you consider yourselves primarily Hutterites in general; you consider yourselves primarily Anabaptists.  What do you think the appropriate grouping is, at which decisions could take place? 

Second problem: Even if you decided who the group is, who do you talk to to get informed consent?  Who is in a position of sufficient authority to give that consent?  And that also is going to be a serious problem in many, in many situations.  In some situations, it's so serious as to render the whole idea impossible.  For example, Irish-Americans, which I guess I'm one-fourth blooded.  There are roughly 40 million Irish-Americans.  There's no authority structure that exists that speaks for an Irish-American.  There's no "there" there, [inaudible], so seeking to get consent of Irish-Americans, just seems on logistical grounds, implausible if not impossible, finding or [inaudible].  On the other hand, with smaller groups, there will often be some authority structure.  Then you've got a problem: what if there's more than one authority?  How do you figure out which authority structure should be, which authority structure's permission should be sought?  And again, I go back to discussing the issues with the people you're working with.  Find out who people with whom you are doing the research, who they view as appropriate, meaningful authorities.  Now, in a Native American context, and that is for legal reasons, the tribes have, at least federally recognized tribes, are sovereign political entities.  That will necessarily include tribal governments.  It may include groups other than the tribal governments as well; it may include culturally or religiously significant or socially significant groupings, that are recognized not by the federal government, but by the people with whom you're dealing, as having significant authority in this [inaudible] group.  In the Hutterite example [you?] discussed last night, the appropriate authority was the minister of the Hutterites.  Most folks [inaudible] the political authority he had, but recognized from within the group as having authority to make decisions for the group.  Now, if there is no group authority, in case you're in a context where there isn't any authority structure you can find, but you're working with a relatively small group, or if you're working with a small part of a large group, it seems to me that the last fallback position should be to make sure that you've talked with and tried to achieve consensus from the people you're directly dealing with.  So to go back to the Irish-American example, I think it's completely infeasible to think about getting group permission or group consent from Irish-Americans in general.  But if you're working with, let's say, a Roman Catholic parish in Seattle that is overwhelmingly Irish, and that's the source of the Irish-Americans that you are using for your research, talk to the parish.  Hold group meetings.  Be available to answer questions.  Seek to provide the circumstances under which that group, which doesn't have a formal authority structure, can reach its own consensus in favor of or opposed to [this research?].  And if you don't get some sort of general agreement -- and I'm being somewhat vague here as a necessity -- some sort of general agreement that, you know, 100%, 80%, 50%, 2/3 vote, I don't know, but some sort of general agreement coming from the group of people you're talking with, you shouldn't do anything.  [inaudible] 

How do you go about getting the consent?  This one, I think, is more easily approached, kind of thinking about how you don't go about getting consent.  You don't go about getting the consent by helicoptering in to a region, wearing a white coat and saying, "I'm a scientist and I'm here to help; give me some blood."  In order to get a true informed consent, I think that certainly on an individual and a group level, you need to build a relationship, or have a relationship, one where where you answer lots of questions, you ask lots of questions, you've interacted with the population, where you've got a sense of partnership, and the population has grown to trust you.  And yo'uve given them good reason to trust you.  That's not an easy prescription!  Nor is it one that the NSF or the NIH is likely to fund.  "I'm sorry, I need twelve months worth of support to go get enough, gain enough trust from the population that they'll agree to let me do this research."  But I think it's that kind of long-term relationship that's going to prove essential [inaudible] basis in order to make sure that a truly informed consent or decision -- or lack of consent [inaudible].  And you need to answer it [inaudible] a large, possibly interminable number of questions, if you need to spend a lot of time, have a lot of patience in order to make sure to do that [inaudible].  Not that, that won't be true in every circumstance; in some circumstances, it may be a population that is eager to participate in research [inaudible] that doesn't require very much time to do that.  I was talking about these issues with [inaudible] in Europe [inaudible], and they couldn't understand why [they ought to be?] concerned about this.  And they cited the example of some work they'd been doing with the Cornish people, in Cornwall, the southwest part of England.  They started to explain their project with the Cornish Historical Society, which had presented [inaudible], "Oh, this sounds wonderful, we've got doctors and nurses to do all the blood, [all the blood training?].  For them, a lengthy consent process wasn't needed; for other groups, a very lengthy consent process may well be needed.  Like almost everything else here, it varies from place to place. 

Fourth set of issues:  What about individual informed consent?  An emphasis on group consent, from my perspective, does not imply any lessening of the importance of individual consent.  That's true both for ethical reasons and for any research coming within federal regulations for American legal reasons.  Individual informed consent is still essential.  Group consent just adds another layer on top of it.  The group consent alone is not enough; you have to get group consent, plus the consent of the individual people who participate in the research.  There's a problem here as well.  It may that the group consent taints the individual consent, that once the population as a whole, the leaders -- once the Hutterite minister has said, "This is a good project; we're going to participate in it" -- it may become very difficult for individual members of the group to say, "No, I don't want to participate in this."  This, I think, is one way in which the whole ongoing saga of Iceland fits into this.  Something that I, too, have given [inaudible] several times yesterday, I wish we had more time to talk about it, more time to talk about Iceland.  Iceland, in a way, has met the group consent requirement.  It's a country of 270,000 people; it's culturally quite homogeneous; it has a functioning democracy; it's got the oldest [inaudible] Parliament in the world, and a Parliament [inaudible] by party-line votes, voted by a majority to approve the database.  I would have to say, as much as I disapprove of that decision, [inaudible].  But the Icelandic scheme doesn't provide for meaningful individual consent.  You cannot, you are not asked whether you'd like to participate or not; you are given an opportunity, a window of several months, in which to file a form saying you don't want to participate.  Should you fail to file the form during that period, you're never allowed to withdraw your record from the database.  And apparently, if the form is not filed on your behalf, and you are a child or someone that [inaudible], you participate in this database.  It presumes consent, and I think that's the largest example we've got right now of group consent being used to override individual consent, in ways that I think are inappropriate. 

For other kinds of research, I think the lesson with respect to the interaction between group consent and individual consent is that the individual consent process has to be structured carefully, again in a way that will vary, depending on circumstances, but in a way that allows the individual a truly meaningful opportunity to say no.  That may mean, ifit's culturally possible to do it, that you interview and ask consent from individuals in a private, isolated position, where no one will know or hear whether they say yes or no.  It may also even mean that you slap a bandage on the arm of somebody who hasn't participated, so that they can look to the rest of the group as if they participated.  What options are necessary are, again, going to vary depending on the circumstances and the people with whom you are working.  What's important, and what's important in all these issues, is acknowledgment of and respect for the interests, the cultures, the wishes, of the people you're working with.  The group consent issue, then, has problems.  The problems cannot be perfectly solved.  Solutions to them can be approached, and it's my quite strongly held view that even with the inevitable shortcomings of those solutions, even though group consent cannot be a perfect process, it is substantially better than the absence of group consent; so much better [since originally adopted?].  Not just as a practical response, but as an ethical and a legal requirement for certain forms of research. 

I'll close by pointing out that this is not about whether we should [inaudible].  Many sorts of researchers have long lived, as a pragmatic matter, with group consent.  Epidemiologists and ethnographers in particular have known that you're not going to get anywhere without the cooperation of the population.  And the cooperation of the population means having meetings, talking with people, talking to the leadership, having, gaining the trust of the people you're working with.  I see this issue coming up in this context as a result of a collision of cultures, not so much the cultures of the scientists and populations being studied, but the cultures of different sciences.  Where geneticists, who are largely accustomed to dealing with fruit flies, yeast, and C. elegans, or dealing with data, manipulating data that others have collected, are now in a much larger and more complicated data collection role than they were before.  And dealing with that role has taken a substantial adjustment.  I think that recognition of group consent as a requirement is an appropriate way for that adjustment period to end.