Panel Report Update: David Shore, Ph.D., National Institute of Mental Health
DR. SHAPIRO: ... take up the item that we skipped this morning in Alexís draft proposal and then move on to next steps in the agenda. So let me first turn to Dr. Shore. Dr. Shore, welcome, itís a great pleasure to have you here. Thank you very much for coming. He will of course be addressing the issue regarding reporting to us in the health update, and I think we all have some information to share on that. Dr. Shore.
DR. SHORE: Thank you very much. I appreciate the opportunity to present to the White House Advisory Commission. I attempted in Los Angeles to give you the very quick version; I donít think I did a very good job of it and I appreciate your giving me another opportunity and some time to walk you through some of the points that we considered and some of the recommendations that came out of that meeting. So that people...
Five and twenty yearsí service on their local IRB. Therefore, they have been wrestling with these risk-benefit, informed consent issues on a fairly regular basis. And we thought that they had some valuable experience to share and perhaps to help guide other IRBs as they seek to address some of these controversial issues.
The group in particular I think was helped by Gary Ellis early in the meeting, reminding us that the current Federal regulations already provide for the provision of additional safeguards by IRBs when potentially vulnerable subject groups are involved in the study. I was asked what was the difference between our approach and your approach and the way I would express it would be the difference between strategy, long-term changing laws, and tactics which was more our approach. Given the system as it exists today, how can we make it function better? How can we not depend on the Congress to change a law or the Department, etc.? Where do we have leverage right now?
These are the members of the expert panel. Ned Cassem is a Jesuit priest as well as a psychiatrist. We have experts on substance abuse, disorders of aging, psychopharmacology, representatives of the Alliance for the Mentally Ill and the Alzheimerís Association, and our representatives of Arena and Primmer. Laura Roberts, to my knowledge, is the only individual whoís formerly trained as a bioethicist and a psychiatrist. So, I think we had a very good group again, virtually all with extensive IRB experience. The presenters are I think well known to this group since you have had on your agenda in recent months several of these individuals. Youíve heard from Gary. Laurie is, of course, on the panel. Abel Shamoo presented as did Alice Wichman of the NIH Clinical Center.
Let me start out with what I would consider some of our shared values--call them biases if you will. Weíve believed research on the etiology of these disorders is greatly needed. We do not know the etiology of autism, Alzheimerís disease, schizophrenia. And, there are no satisfactory animal models, so such research must involve human participants. Were we to confine such research to the least ill individuals, we would I think be doing a disservice to those who are most severely impaired because they are precisely the individuals who have the most to gain and have the greatest need for the benefits of research that will uncover etiology and new treatment targets.
We also felt that we needed to engage experts on the plentiful disorders and the symptomologies. Again, this was a matter of tactics. We thought that changing some of the laws involved might take a decade, and as youíve seen with the sheep cloning, events have a way of overtaking our well-planned schedule, and our concern was that rather than try and focus on what the law should say in ten years, we needed to think about how we can better inform the IRBs that are considering these issues this year and next year.
And, finally, we needed to be very clear that serious abuses of human subjects have occurred, whether Tuskegee or Willowbrook, and we need, as the National Institutes of Health, to recommend steps that will decrease the likelihood that we will see a recurrence of such problems. In the past, as in Willowbrook, people have been subjected to abuses simply because they were compliant. In other cases, the informed consent process was not adequate; people were not properly informed of the risks or alternatives to research. So, the panel asked, essentially, how the local IRBs might best fulfill their responsibilities to protect research participants. Again, we looked at IRB discretion. We considered that, as the biochemist would say, the rate-limiting step because at NIH we do not fund studies that have not been previously approved by an IRB that either has a multiple-project assurance or a single-project assurance. Our process is somewhat different, and that you heard about this morning.
There were attempts to describe what we call points to consider when involving individuals with questionable capacity to provide informed consent. And the four topics on which we focused were the roles and responsibilities of IRBs, surrogacy, and advance directives. The assessment of capacity to consent, which you heard about through Dr. Appelbaum here previously, and the issue of conflicts of interest, and Iím going to touch on each of these in turn.
In terms of the recommendations that came out of this group--this is actually fairly similar to a recommendation that the National Alliance for the Mentally Ill has been talking about for the last several years. It was fairly clear that while the Federal regulations mandate five individuals on an IRB, one of whom should be from outside the institution, if we have an IRB with 20 members and we still only have one individual from outside the institution who represents the views of the community, the views of the consumers, the views of the families, that that outlook is at risk of being diluted and we felt that additional membership of individuals who can represent the views of the community, the views of the subject population, the views of the families would be very beneficial and could head off some of the potential problems that have already occurred. IRBs should be hearing from consumers, advocates, and others who are independent of the institution.
The second, in effect, follows I think from the first, which that, as was pointed out, IRBs already, as youíll see in the handout thatís outside with 45 CFR 46 can already require additional safeguards. We recommended--the experts to NIH recommended a sliding scale such that the higher the risks, the lower the benefits, and the greater the impairment of capacity to consent the more safeguards should be considered. Waiting periods, monitoring, frequency of reporting--in particular, the IRBs should consider the potential value of an independent outside monitor either representing the family, the clinician, or the IRB in cases where there seems to be relatively high risk and individuals may be relatively impaired.
It was also pointed out that IRBs may or may not realize they already have the right to observe virtually all of these processes from recruitment through the informed consent process through the debriefing of subjects after a protocol, including the debriefing of family members. So, again, we recommended that greater scrutiny be given to studies that have relatively higher risk. We opted against the rows and columns and figuring in which category a particular project would be and then automatically applying a certain set of procedures. We decided that if the local IRB is really representative of the broader community, it should be able to judge those factors.
The issue, then, of surrogates--we believe that in many cases if an individual cannot understand all aspects of the protocol, a family member may be quite helpful in translating, if you will, the document; and whenever possible a surrogate should make research decisions reflecting the views of that individual when decisionally capable based on prior decisions. I think many people have recommended the best interest standard, but what that would do would simply be to outlaw research on people with disorders such as Alzheimerís disease since arguable the best interest is to leave that individual alone. And, therefore, in many cases we see individuals who expressed a prior desire to participate in research studies to see that their generation is the last that will have to suffer through this disease. But it seems that there is some risk that individuals will have a right to say no but will not have a right to say yes.
In terms of assessment of capacity to consent, we are a long way from understanding how to properly address this. Youíve heard four different criteria. We came down in favor of a rather more strict criteria than is, I think, currently being used. We were concerned that some individuals can recite back a list of side effects or a list of facts about a given protocol, yet may be delusional about the study, delusional about their part in the study and how it might affect them, and therefore we believe that there should be priority given to developing instruments that can detect an individualís appreciation. Unfortunately, as weíve heard in Los Angeles, that is arguably the most difficult of Dr. Appelbaumís criteria to formally address and we believe that NIH should prioritize this and the group recommended that the institutes do so.
It was also pointed out that comprehension for a person being capable of informed consent is not necessarily a yes-or-no, black-and-white decision. IRBs should also consider ways to enhance the individualís understanding and appreciation. You heard me talking all about giving small amounts of information repeatedly over time, about giving out a single sheet summary of the key aspects of the study every time the individual comes in to participate in part of the trial, and that may be considerably more valuable than having a six-page, single-spaced document in the individualís research file or lost in their home somewhere. Questions should obviously be encouraged. If we donít communicate better, weíre just asking for additional problems in the future.
Conflicting roles, potential conflicts of interest--here weíre talking about everything from financial to concern over mixing clinical roles and research roles, and I think these need to be addressed by the investigators and carefully considered by the IRBs since individuals with severe cognitive impairments may be more vulnerable to the therapeutic misconception.
Last, but not least, there was a view that this group felt that the common rule should apply to all human subject research involving individuals with questionable capacity for consent.
The implications--obviously the state of our knowledge concerning etiology is relatively primitive for severe mental disorders. Such research must involve individuals who are affected by these illnesses directly. Ethical research hinges upon informed consent, confidentiality protections, and adequate review.
We briefly considered issues such as placebo controls from washouts, etc., and decided that in order to really do that justice we would need another meeting and we are now planning such a meeting for the future to talk about research design and perhaps the talk following mine will lead into that.
In general it seemed that the advocacy groups understood the need for a better understanding of the underlying causes of these diseases, and they also understood that the biological vulnerabilities, as revealed by genetic studies for biochemical measures concerning who might best respond to a given treatment, might well not have direct benefits for that particular individual and in my attendance at these meetings it seems that research without the prospect of direct benefit of greater than minimal risk is of primary concern to some of the people who have spoken here in past sessions. We try to make clear that if we outlaw research that does not present the prospect of direct benefit and exceeds minimal risk, we may be making it extremely difficult to help the next generation of patients who will be admitted and the next generation of medications that will be used for their treatment.
Okay, I have three slides left, so I think Iím doing all right as time. Again, promising genetic paradigms. Designs intended to predict which individuals will benefit the most from which treatment generally involve either taking blood from an individual, finding a gene five years later which may or may not lead to understanding the etiology. We know that some individuals respond very well to certain treatments; others are troubled by side effects and donít respond very well. We would like to go back and understand the biological differences between people who respond very well to treatments and those who donít. Such studies would not be a direct benefit to those individuals, but they would potentially be of great benefit to the next group of patients who will come into the clinic with that same disorder. So, the concern was that we need to know why some people respond to treatment and others donít, and what genetic factors cause people to be vulnerable to the development of these severe disorders so that future generations of the same families do not continue to be at risk.
There was also a concern about--of the need to compare people who suffer from one particular type of mental disorder with individuals who have a different disorder or a different form. A childhood onset obsessive compulsive disorder vs. adult onset. There seems to be a trend toward narrowing the focus of distributive justice such that if we were studying a comparison population, that some people seem to be saying that research should be outlawed. If we go back to research of 30 years ago in which we compare chronically ill patients with healthy college students, weíre not going to learn a great deal. Weíve discovered that until we can disentangle the effects of treatment of chronic illness, of hospitalization from the effects of the underlying illness, we are not going to be of the notion to understand the path of physiology.
And my final slide. Different degrees of research, risk, and decisional impairment should, the panel argued, lead to different levels of scrutiny and different levels of safeguards. Additional protections, everything from involving the family surrogates and independent monitoring may be quite advisable in high-risk protocols. But if we treat everyone who has a cognitive deficit as incapable of understanding research, thatís really not accurate, and thatís really not respectful of those individuals. Scientific neglect of mental illness and substance abuse is not exactly beneficent and just and we, again, are trying to strike a balance maximizing the potential benefits, minimizing the risks, and comparing the risks of the experimental interventions with those that are intrinsic to the source being studied.
So, that is all I wanted to say. I left some time for questions. Oh, thank you very much.
DR. SHAPIRO: First of all, thank you for being here today and preparing this presentation for us, and thank you also for mobilizing the panel in the first place to consider these issues, which as you know we consider really quite important and very relevant to our work. So, weíre very much the beneficiary of some of this and thank you very much for that. But let me move, now, to questions and let me turn to Dr. Childress so I know--I saw his arm up early in this discussion.
DR. CHILDRESS: First of all, I just want to echo the expression of gratitude to Dr. Shore and his colleagues for arranging the conferences our Chair has indicated, and also for giving NBAC members a chance to be at the conference to listen carefully to what was being said, and also to have an opportunity at the end of the conference to actually try to incorporate some of the suggestions into the evolving draft. So, weíre very grateful for all of that, and also for the written statement now because some of the formulations here are illuminating in different ways, even than the earlier discussion.
One clarification, though, and then one question. The clarification has to do with the way in which some of the discussion in the work and some of your comments today seem to suggest that NBACís drafts along the way and the discussion and the meetings have really put greater-than- minimal-risk research without the chance of direct benefit at risk of being outlawed. But really what weíve done is actually said in most of the drafts that informed consent is required and if, for instance, as the document suggests, that you can have altruism combined with cognitive impairment, that is certainly true. But, weíre suggesting that at least in the draft--we havenít voted on all this yet--that it may be very important to think about altruism in relation to people who can actually give voluntary informed consent. So thatís a clarification, because as our drafts evolve weíve not made efforts toward outlawing this, but rather setting certain kinds of limits on it.
The question that I have is this: I seem to hear in your comments today a little difference in tone from the conference, at least on the part of several speakers in the conference, regarding the question of possible regulations. There seemed to be at the conference a tendency to say no regulation unless theyíre mentioned in the guidelines. And yet in your remarks today there was much a more tentative approach and less distinguished strategy and tactics so it would be difficult to bring about regulatory change, so letís go with guidelines. And then the recommendations we are considering--some would be for regulations; some would be for guidelines. But I guess what I was wondering is whether this documentís maybe in contrast to some of the specialty conference, isnít now as opposed in principle the regulation, but rather as a more practical matter itíd be difficult, take time, etc. Itíd more effective to go the direction of guidelines. Is that a shift in views?
DR. SHORE: The only one of our recommendations that I believe would require legislation is the last one, actuallyóof the application of the common rule to all research involving the cognitively impaired. None of the other seven, the other six recommendations would require regulation and then in effect all could be implemented now by IRBs around the country, and thatís why we took that approach. Again, weíre given a system which was intentionally decentralized and the ultimate authority was given to the local institution to citizens who live in that area to people who feel strongly about issues in that area, and we thought that in this current political trend brief
that it was relatively unlikely that a big-government solution was likely to soon be implemented and therefore we focused on what could be done now. I think that IRBs are having difficulty. The same protocol reviewed by three different IRBs might be considered acceptable, unacceptable, or in need of revision, and I think IRBs have been asking us questions, have been asking each other: How do you deal with this problem. And, thatís why we brought together IRB members who have--weíll, when we tried this, this is what happened and this is what worked and when we tried this it was a disaster. So, again, weíre trying to really provide more points to consider; that is, consider the risk, consider the degree of impairment, consider the prospect of benefit, and then bring in family members if you feel that the subjects are at risk or misunderstanding. Have an IRB representative present. You donít need legislation to require that. And again, the process can and should be visible throughout.
DR. CHILDRESS: I guess Iím unclear, then, would you have any objection if, for instance, through a change in regulations, IRBs would be required to include at least one voting member? I would say one could offer this guidance but one might also seek a change in regulation. DR. SHORE: IRBs are already required to include one voting member from outside the institution. It was our concern that of the five required members there is only one who is required to be from outside the institution at the present, and we now we see IRBs that consider a broad range of protocols and can take 20 or 30 members. And it was our viewpoint that the expression of views from outside the institution needs not to be deluded. One person has a hard time speaking and convincing 30 others and that people should be brought in with specific expertise when such protocols are considered. IRBs already have the authority to do that and they already have the requirement to have at least one person from outside. We emphasize there be at least one.
DR. SHAPIRO: Thank you. Iím just conscious of the time, so with no disrespect to anyone I think we could try to keep questions and answers within three minutes or so; otherwise, theyíll be considered speeches, which are out of order at this point in time. Mr. Capron?
PROF. CAPRON: I have several questions I wanted to ask you within the three-minute limit. Let me ask this one. The problems with certain kinds of psychiatric research in terms of design and informed consent have gotten a good deal of attention for a number of years. The UCLA study, the court opinions, and so forth. Has this led to changes in the protocols conducted or approved at NIMH? And if so, how do these factors that have drawn concern get flagged and taken into account in the decisionmaking process about which studies to conduct or fund? And how many protocols if any have been affected and disapproved because of such concerns?
DR. SHORE: Okay, I actually think I can answer that in three minutes. The answer is that yes, since 1994 we have changed things a great deal. In fact, we sent out between 200 and 300 copies of the OPRR report to all of our investigators who were doing research that might have might have anything to do with the kinds of designs that were used at UCLA. That was our first step. We have seen considerably more IRB comments and concerns coded on the peer review process as part of the outside scientific merit review. Studies cannot be funded with an IRB concern until thatís resolved between NIMH staff and OPRR. IRB comments--we routinely require they be addressed if a studyís going to be funded. The third part was how have we changed the way we do business. Since I guess Iím the person responsible for doing that, I can answer that fairly clearly. I think what we have tried to do in the last couple of years has been to identify protocols that raise this kind of issue: Taking people off medications to which they have previously been responsive; prolonged medication-free intervals, especially in outpatient settings. And we have in many cases required the consent documents. In some cases, frankly required changes in the protocol. Iíve gone over to Executive Boulevard on many occasions. And we wanted to make sure that the consent documents explain clearly the risks and the alternatives so that if people wanted to participate they could do so, but that those risks were adequately described. And, in some cases we wrote paragraphs describing certain kinds of risks for certain kinds of studies and they were adopted, so weíve been a lot more active. I would say I look at probably one protocol a month of new studies in that way.
PROF. CAPRON: If I could just understand the process a little more clearly. There was a time when what I understood from OPRR was that study sections (if I am using the right term) didnít see this as part of their review of consent documents or the human subjects protection issues as such, but occasionally those issues emerged from the discussion but werenít really seen as an assigned responsibility. So, if I could just ask: Have the changes that you described in NIMH been accompanied by some formal directive to study sections about their responsibility in this regard, or has this been more or less what you are regarding as a heightened sensitivity in the community; and, secondly, is it only in those cases in which you see something fall into the categories of research that you mentioned that you requested consent documents, or are those now routinely reviewed either by the study section or by your office for either all experiments or all categories that youíve set out someplace where we could see a listing of them rather than just hearing it orally here?
DR. SHORE: I think I can answer those also. I think it is a misconception that IRBs, IRGs, or study sections--either term applies--did not look at human subject issues. At least for the last ten years as far as Iím aware since Iíve been sitting in on study sections, thereís been a requirement and informed--pink sheetsóat the end of a section that refers to human subjects and therefore the reviewers are required to address that. Often reviewers ask to see consent documents--not always, in certain kinds of studies more often than others. And, so, if the reviewers want to see the consent documents, they can ask for them. They can take them into account and decide whether theyíre satisfactory or not. And then they can decide whether to code an IRG comment or concern regarding human subjects. Thatís been the case for at least 10 years as far as I know.
In terms of routinely reviewing consent, we donít maintain a repository of consent documents because very often they change each year. The consent documents--well, these regulations havenít changed in 15 years. I think itís--most of us who have been involved in research are aware that the standards by which consent documents are judged have changed dramatically over the past 15 years and thatís why we specifically tell people not to take an old consent document that was approved by your IRB and assume that itíll be okay now, because thereís a very good chance that it would be considered inadequate now. But we do not routinely look at all informed consent documents. What we do is we look at specific kinds of what we consider to be potentially higher risk studies.
DR. SHAPIRO: Thank you. Steve, then weíre going to have to go on to our next--.
MR. HOLTZMAN: As Jim Childress said, the draft in its current state does not recommend prohibiting nontherapeutic research involving greater than minimal risk per se. It only effectively prohibits it in the case in which the individual subject is incapable of giving informed consent. My question to you is: If you were to look at, on average, a 100 protocols involving nontherapeutic research involving greater than minimal risk, how many of them require individuals who are incapable of giving informed consent?
DR. SHORE: Well, let me give you a good example which will probably--.
MR. HOLTZMAN: Iím trying to get my arms around the magnitude of what is potentially being prohibited.
DR. SHORE: Let me give you what I think is one very accurate example: research on the genetics of Alzheimerís disease. The studies that have been conducted thus far that have identified or applied different genes that seem to confer a substantial vulnerability to Alzheimerís disease have typically involved the study of affected sibling pairs--an older brother, a younger sister, etc. In that case, given the progression of Alzheimerís disease, it is almost inevitable that by the time the younger sibling develops diagnosable Alzheimerís disease, the older sibling is probably not capable of providing informed consent.
PROF. CHARO: Iím sorry--excuse me, Dr. Shore, but since all you would need to do is venipuncture, which is minimal risk--.
MR. HOLTZMAN: Most people have said genetic studies involve more than greater than minimal risk.
PROF. CHARO: This is--no, this is exactly where the discussions are, but in this particular case you have the disease already. Youíre already symptomatic, youíve already been diagnosed, itís not new information. So, itís just the venipuncture.
DR. SHORE: I would agree with you that in my view, as a former IRB member, I would see a study in which looking at medical records and taking a tube of blood as long as alarming information is not passed on to the individual or the family might meet minimal risk criteria, might qualify for a waiver of informed consent, but not all IRBs feel that way and many of the people whoíve testified before this group seem to oppose waivers of informed consent. And, it was my concern that certainly earlier versions of your document seem to make it very difficult for studies in which, for instance, insurance discrimination, employability discrimination based on confidentiality issues might cause this, as Mr. Holtzman was suggesting, to be greater than minimal risk.
PROF. CHARO: Iím sorry. Weíre going to be getting into this tomorrow, obviously, so it has slop-over to the definitions, what--the content of the meeting of minimal risk, but to the extent that people have argued that genetics research frequently is nonminimal risk, itís in the context of fishing expeditions, in which youíre looking at asymptomatic individuals and looking for markers for things that you think might predispose. Itís not in a situation in which somebodyís already symptomatic where theyíve been diagnosed. Anything thatís going to be found is already evident, so that the research is not creating a new risk except to the extent that youíre doing the blood draw. So Iím still--I find Steveís question about the magnitude of experimentation that might be somehow cut off by the recommendations of this report to be a very good question--I would love to understand the magnitude of whatís at stake, and yet I find this particular example perhaps not the most persuasive one for getting at that category of experimentation.
DR. SHORE: Iíll give you another example: Consider individuals with certain forms of rapid cycling bipolar disorder. Some people respond very well to lithium; some people respond very well to anticonvulsants; some people donít respond very well to any of those treatments. It may well be that they differ in receptor subtypes and that a spinal tap might be necessary to detect such differences to predict which individuals in the future would respond to which treatment, that spinal tap is not likely to help that individual because that individual has already either responded or not responded. It was my concern that a blanket outlawing or making it so difficult that one would have to go to the Secretary of Health and Human Services to get a waiver, would in effect make it impossible to do such studies, and presumably the next cohort of patients could benefit from that information.
MALE VOICE: All right, letís--if there are some questions--Iíll turn to you, Eric, but letís not try to settle all the issues now because we donít have time.
DR. CASSELL: Itís a 10-second response. The way the report is coming out, those would be permissible studies given certain protections that donít presently exist, but they would be permissible studies so it would not be ruling such a study out.
DR. SHORE: Iím glad to hear that. I was very alarmed by some of the earlier views--.
DR. CASSELL: Particularly in the example you used where the question of capacity of consent in that particular example of manic depressive or--bipolar rapidly cycling, the issue of impaired capacity of consent is not as great as it is, for example, in Alzheimerís or in schizophrenics and so forth.
DR. SHORE: A person with rapid cycling bipolar disorder can be quite psychotic.
DR. SHAPIRO: We will have an opportunity to return to this question.
DR. SHAPIRO: Weíll return to this question.
DR. CASSELL: We appreciate the input and weíll be discussing it.