[Accountability in Research (August 1993); 3, 1:1-17]
Accountability in Research Using Persons With Mental Illness
Adil E. Shamoo* and Dianne N. Irving**
* Center for Biomedical Ethics, and the Department of Biological Chemistry, School of Medicine, University of Maryland, Baltimore, Maryland 21201
** Department of Philosophy, De Sales School of Theology, Washington, D.C. 20017
Although medical research involving the use of persons with mental illness is critically important, in order for the research to be ethical and legal, there are certain considerations and restrictions which should be immediately readdressed in order to insure that the welfare of these vulnerable research subjects is protected, and their best interests are assured. A brief historical examination of medical research codes, guidelines, recommendations and Federal Regulations reveals the various considerations and restrictions on informed consent and accountability applicable to the use of persons with mental illness in medical research. Several concerns are raised about how these considerations and restrictions have been interpreted, and specific recommendations are offered to improve them immediately by means of representation from consumers and/or their families, and organizations, e.g., NAMI members.
Keywords: mental illness, Federal Research Regulations, informed and proxy consent, accountability, NAMI.
The use of human subjects in biomedical research has been discussed, debated and analyzed for decades now, especially since the findings of the medical research atrocities committed during World War II (e.g., Appelbaum 1987; Barber 1973; Beecher 1966; Frankel 1975; Ivy 1948; Jonas 1984; Katz 1972; Levine 1986; Lifton 1986; Mitscherlich 1962; Muller-Hill 1988; Proctor 1988) -- culminating in the Nuremberg Code (1947). Since then further refinements and protections have been added through the Declaration of Helsinki (1965, amended 1989), and here in the United States, the National Commission (1978), the President's Commission (1979-1983), and a series of regulations at the Federal level.
The Federal Regulations 45CFR46 (1981, 1983, 1989, 1991), with subsequent amendments and additions, clearly required the establishment of research protocol review by institutional review boards (IRB's) at each institution conducting research supported by government funds, or submitted in support of a government-required approval such as that for drug testing. Based on detailed studies of the National Commission and the President's Commission, and promulgated through the series of Federal Regulations, there were specific protections for special groups who were considered "vulnerable", e.g., persons with mental illness, pregnant women and fetuses, prisoners and children.
In the 1983, 1989 and 1991 Federal Regulations, special subparts of these regulations (Subparts B, C and D) were detailed for all of these vulnerable subjects except for those persons with mental illness. This raises the question as to why persons with mental illness -- who are unquestionably a "vulnerable" group -- were not considered as human research subjects needing special protections in a specific subpart of the Federal Regulations?
The official response seems to have been that the Federal Regulations do provide for such protections -- by authorizing local IRB's to determine if any such protections are necessary. But the question remains as to why such extensive, well-debated and publicly-scrutinized recommendations such as those reported by the National Commission and the President's Commission were virtually dropped for only that vulnerable group of human research subjects who were persons with mental illness -- and not, then, also dropped for the other vulnerable groups? Additionally, what was the justification for delegating to local IRB's the essential responsibilities for affording protections for persons with mental illness -- protections which were otherwise provided through the recommendations of the National Commission? What was the justification for allowing the IRB's to in turn redelegate this authority in certain cases to others?
The best description of what happened to the issue of the use of persons with mental illness in research is summarized in a section on "special populations" in a book by Paul Appelbaum, M.D., Charles W. Lidz, Ph.D., and Alan Meisel, J.D.:
Special protections for mentally ill and mentally handicapped persons were the subject of extensive recommendations by the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (53). None of those recommendations were implemented. This outcome was the result in large part of opposition from researchers on mental disorders, who claimed that the populations in question were no more vulnerable than most persons with severe medical disorders and that the suggested limitations would seriously restrict research on mental disorders (54). Mentally impaired subjects continued to be covered under the general regulations, which, in the case of incompetency, allow consent to be obtained from a subject's legally authorized representative, presumably a guardian or equivalent person with the power to consent on the subject's behalf under the laws of the jurisdiction. (Appelbaum, Lidz and Meisel, 1987, p. 228).
Another author noted the express protections afforded persons with mental illness under the earlier recommendations of the National Commission, which he assumed would be carried over into the Federal Regulations. Dennis M. Maloney stated:
Due at least in part to the freeze on government regulations, these proposed rules remain the ones for us to heed at present. It is not currently known whether these rules will be revised or made final by HHS in view of federal policy to limit regulatory activities and reduce government paperwork. In the author's opinion, the proposed rules that do exist, as they appear appended to this chapter, should still serve as a useful guide for researchers in this area. Unless unusually substantial changes are made in the proposed rules, final rules (when published) should draw heavily on the proposed rules as appended to this chapter. Once again, the cautious researcher should consult federal agencies (e.g., the NIH Office for the Protection of Human Subjects). (Maloney 1984, p. 377).
It would appear that the issue of using persons with mental illness as human research subjects has been lost in the shuffle, due in part to the lobbying efforts of some researchers on mental disorders, and in part to the relatively small degree of constituency and advocacy for persons with mental illness when the above developments occurred during the mid to late seventies. Especially given their vulnerability and the reality of their use in on-going research, the situation of these persons with mental illness remains a concern. As one historian has recently and succinctly put it: "...it is the socially powerless and disadvantaged who are most likely to be subjected to unethical research" (Gillespie 1989, p. 13).
The aim of this article is thus three-fold: first, to indicate briefly some of the history relevant to the use of human subjects in medical research in general, and of persons with mental illness in particular; second, to draw attention to two of the most significant inadequacies in the Federal Regulations which deserve further examination and discussion; and third, to propose recommendations in which members of the National Association for the Mentally Ill (NAMI) could be useful in carrying out these regulations on many levels of participation -- including membership on IRB boards, and in the role of the "consent auditor" as proposed under the recommendations of the National Commission Report (National Commission, 1979, p. 6).
INFORMED CONSENT AND ACCOUNTABILITY IN TWO MAJOR HISTORICAL DOCUMENTS
In reviewing briefly two of the most well respected historical documents concerning the use of human subjects in medical research, it should be noted that some restrictions present in early documents have disappeared from the more recent ones. On the other hand, several exceptions included in recent documents were never contained in the earlier ones. This does not, we would argue, presume that what is most recent is thereby what is most acceptable or sophisticated. With respect to readdressing the current Federal Regulations, it would be prudent to reconsider and reevaluate again both old and new restrictions and exceptions.
A. The Nuremberg Code
Although no reference is made to the Nuremberg Code (1947) in the current Federal Regulations, these codes were the basis for later codes and regulations. Given that this multi-national tribunal in particular was so close to the massive empirical evidence presented to it on the scientific fraud and human abuses of recent medical research which used human subjects, we should note the issues which they so clearly judged to be critical, and which they so strongly considered should become incorporated in any country's policies when human beings are allowed to participate in the medical research enterprise.
It is often not realized that the ten principles of the Nuremberg Code were actually part of the "judgment" handed down during the Nuremberg Trial's decision in United States v. Karl Brandt:
...The great weight of the evidence before us is to the effect that certain types of medical experiments on human beings, when kept within reasonably well-defined bounds, conform to the ethics of the medical profession generally. The protagonists of the practice of human experimentation justify their views on the basis that such experiments yield results for the good of society that are unprocurable by other methods or means of study. All agree, however, that certain basic principles must be observed in order to satisfy moral, ethical and legal concepts. (in Katz 1972, p. 305) (emphasis added)
The court proceeded then to list ten basic principles, the core of which includes the following points.
The very first principle relates specifically to the human subjects who might participate in biomedical research, and is by far the lengthiest statement. Two issues are critical. First, only competent subjects may participate; and second, they must have legal capacity to consent. The former calls into question whether or not persons with mental illness may be used as research subjects at all. The latter is emphasized as the means by which the court could determine "criminal culpability and punishment" (in Katz 1972, p. 305; Katz 1992, pp. 227-239)
In addition, this first principle requires that the subject: be able to exercise free power of choice without restraints; have sufficient knowledge and comprehension of the information; be advised of the nature, duration and purpose of the experiments, the method and means used, what risks to expect, and the anticipated effects on his or her "health or person". Interestingly, the duty and responsibility (legal) for the consent process fell to the investigator -- a "personal duty and responsibility which may not be delegated to another with impunity" (in Katz 1972, p. 305). Presumably, the purpose here was for any court (such as the Nuremberg Court) to be able to determine the accountability of any harm caused to a competent legally consenting human subject.
These two points -- competency and accountability -- continue today to be the crux of the issues concerning the use of persons with mental illness in medical research. The debates need to be redirected to the issues of whether or not the mentally infirm may be used as human research subjects; and if so, what are the limits and restrictions. Who is at least legally accountable for insuring proper consent from and for causing harm to these vulnerable human subjects during and after their participation in medical research?
There seems to be no consensus even now on what constitutes "competency", even within the field of psychiatry which is, at least professionally, the ultimate judge on these issues, although the literature abounds with conflicting theories and suggestions. What is often not discussed is the issue of accountability, especially in these days of "consensus" building and committee or team decisions which in effect dilute the responsibility and therefore the accountability for those decisions. These two major issues of competency and accountability need to be settled in order for any regulations or guidelines to carry any credibility -- either ethically, professionally, legally or socially. If no "consensus" on the theoretical level can be attained, then perhaps the issues should be referred to the electorate, if credible research using persons with mental illness is to continue. What seems to have happened is that research has been allowed to progress before these and other such critical issues have been resolved.
To return to the Nuremberg Code, with reference to the experiments themselves, it requires that they: be designed and based on the results of animal experiments as well as on a knowledge of the natural history of the disease or problem under study; not be random or unnecessary in nature; be performed to obtain fruitful results for the good of society which are not procurable by other means (if they first satisfy moral, ethical and legal demands); should avoid all unnecessary physical and mental suffering and injury; should not be conducted if there is a reason to believe death or disabling injury will occur; should not contain degrees of risk greater than those determined by the humanitarian importance of the problem; should be performed with proper preparations and in adequate facilities to protect the subject against even the "remote possibility of injury, disability or death" (in Katz 1972, p. 305).
Clearly, the Code requires that no experiment should be performed on human subjects which are scientifically ill-designed -- or ill-proposed. The most basic of ethical considerations requires that the experiment be designed and executed on scientifically sound bases (Shamoo, 1991). There is also a limit to the amount of physical or mental injury to the human subject which is acceptable, as well as a concern expressed for the quality of the facilities in which these experiments would take place. These points are not to be lost in current debates about research fraud and misinformation (Shamoo and Annau, 1990). No human patient should be subjected to even minimal invasion, injury or risk simply because the scientific investigator did not really know what he was doing, or if he was mishandling his data.
The final issue the Codes were addressing concerns the medical researcher himself. The experiment should be conducted "only by scientifically qualified persons"... with the "highest degree of skill and care" (see Irving 1993; Shamoo, 1991). And the researcher should be prepared to terminate the experiment at any stage if he believes injury to, disability of, or death to the human research subject may result (in Katz 1972, p. 305-306).
It is precisely here that the connection between informed consent and accountability can be justified. It is the medical researcher who is responsible for the design and the execution of the experiment. This is his area of professional expertise -- an expertise not possessed by the human research subject or others who may be tangentially involved. It is precisely because he has this expertise that he is allowed to design and to perform these experiments on human subjects. It is because of this expertise that he is the one charged with informing the human research subject about the facts and details of his scientific experiments. And it is on the basis of the medical researcher properly communicating this scientific information to his patient that the patient can become truly informed and thereby freely choose to participate in the researcher's project. If the responsibility for the design and for the execution of the experiment rests on the scientific expertise of the researcher, as well as the responsibility for the quality of the scientific information which is required to be transmitted to the human research subject before that person can give truly informed consent, then it is the researcher who is accountable for the safety of his human research participant as well as for the quality of the informed consent process.
Therefore, the medical researcher as a research expert is accountable for his own work. But what precisely is a "medical research expert"? Generally, scientists have little if any medical course work; physicians receive little if any course work in basic or clinical research skills or techniques in medical school education. Are their separate goals mutually inclusive? Generally, scientists are seeking basic facts and data about nature; physicians are seeking how to cure or care for their patients. Could these inherent differences be a serious source of conflict of interests (Irving, 1993)? Finally, how to assure the integrity and accountability of the expertise of the medical researcher, the quality of his research, the reliability of his data -- and the physical and mental integrity of the consenting human subjects on whom he performs his experiments are additional issues that need to be readdressed (Shamoo 1991, 1992, 1993).
In sum, the Nuremberg Code of 1947 clearly identifies at least the two critical issues of informed consent and accountability before competent human subjects are permitted to participate in medical research.
B. The Declaration of Helsinki
The Declaration of Helsinki (1964) begins a series of refinements on the Nuremberg Codes. There is more of a recognition that the researcher is not just a "scientific" researcher in medicine, but that he is also a physician, with a special physician-patient relationship and professional responsibility, which must be taken into account when he is doing medical research on any human subjects. Hence, "clinical" research is the focus, and there is a considered effort to distinguish between therapeutic and experimental research. It is here that the potential conflict between "researcher" and "physician" begins to take shape, as well as the issue of the use of incompetent human research subjects.
The Declaration begins by "binding" the doctor with the words, "The health of my patient will be my first consideration." First, it binds the doctor to the International Code of Medical Ethics which states that "any act or advice which could weaken physical or mental resistance of a human being may be used only in his interest" (in Katz 1972, p. 312). The implication is clear: clinical research may only be performed when there is a direct therapeutic benefit to the patient. The Declaration proceeds to make the distinction between therapeutic and experimental (or purely scientific) research, a distinction which is based on the purpose or the aim of the research. The aim of clinical research can be directed to the therapeutic value it may have for that patient; or it can be directed to the acquisition of purely scientific knowledge, with no therapeutic value to that patient (in Katz 1972, p. 312). Although the distinction is an important one, and one which needs to be reconsidered in the current Federal Regulations, the Declaration actually seems to contradict its own statement (above) on this critically important distinction.
For example, the Declaration proceeds to address two broad categories of clinical research: clinical research combined with professional care; and clinical research which is non-therapeutic! In the former category it states that "the doctor can combine clinical research with professional care, the objective being the acquisition of new medical knowledge only to the extent that clinical research is justified by its therapeutic value for the patient." Yet, under the second category it states that "in the purely scientific application of clinical research carried out on a human being, it is the duty of the doctor to remain the protector of the life and health of that person on whom clinical research is being carried out" (in Katz 1972, p. 312).
This distinction made between therapeutic and experimental research is an important one, but confusing and misleading as a guide to clinical researchers. Is the Declaration restricting clinical research to: therapeutic research only for the direct benefit of the patients; or allowing therapeutic research which would only benefit similar "classes" of patients; or for purely scientific knowledge? These are three different aims. Nevertheless, it does afford us an important issue to be grappled with when reconsidering recommendations for the Federal Regulations. This issue has been vigorously debated in the recent past; yet has resulted in no real consensus. The issue needs to be redebated - particularly considering the fact that all three types of research are on-going.
The Declaration also makes another important distinction: between consent for research by competent or by incompetent patients. Under the first category (clinical research combined with professional care), the requirements for competent patients are similar to those of the Nuremberg Code. Again, under the second category (clinical research which is non-therapeutic), the "nature, purpose and the risks of clinical research must be explained to the subject by the doctor." Clinical research "cannot be undertaken without his [the patient's] free consent, after he has been fully informed"; and the subject of clinical research "should be in such a mental, physical, and legal state as to be able to exercise fully his power of choice". Finally, the consent should as a rule be obtained in writing (in Katz 1972, p. 313).
But what is new is the express possibility of the inclusion of incompetent patients in the research enterprise (Katz 1992, pp. 227-239). That is, the Declaration continues that if there is legal incapacity, "consent should also be procured from the legal guardian." Thus with proxy consent by a legal guardian, an incompetent human subject may participate in clinical research. Once again, the issue of consent in the Declaration is tied to the issue of accountability. The issue is not whether or not a patient is "medically" competent, as might be determined by a psychiatrist or by a behavioral psychologist; but whether he is legally competent. And legally, we know that a third party may give consent only for activities which are for the direct benefit of the patient. This would presumably preclude the participation of persons with mental illness in any medical research other than that which would be for his direct benefit.
Given the theoretical turmoil within the psychiatric and psychological professions (whether understandable or not), at least legal accountability can be assigned. Here the patient must be legally determined to be incompetent so that proper legal guardianship may be provided. Presumably this guardian is also accountable to the courts for any decisions he makes on the patient's behalf.
Again, in the Declaration accountability will ultimately rest with the researcher himself. The researcher is to consider the same requirements for scientifically sound, well designed and properly executed experiments as in the Nuremberg Code in order to limit the risk of harm to his patients. Because of his expertise, it is the clinical researcher who is vested with the responsibility to inform his patient of the relevant scientific facts in the consent process, and it is the clinical researcher who has ultimate responsibility for the project: "..the responsibility for clinical research always remains with the research worker; it never falls on the subject, even after consent is obtained." Indeed, here the doctor is even cautioned about the use of experimental drugs: "Special caution should be exercised by the doctor in performing clinical research in which the personality of the subject is liable to be altered by drugs or experimental procedures" (in Katz 1972, p. 312). This "special caution" should not be lost, we would argue, on researchers who use persons with mental illness as research subjects.
Thus, in addition to the initial general issues of informed consent and accountability in the Nuremberg Code, the Declaration begins to open up the controversy by making two new important distinctions: first, between the permissibility of therapeutic and/or experimental research; and second, between the consent requirements for competent as opposed to incompetent research subjects. It is in the recommendations of the National Commission where much greater scrutiny of these and related issues are found.
RECOMMENDATIONS OF THE NATIONAL COMMISSION
The Belmont Report (1979) of the National Commission identified the need for special protections for persons with mental illness when they participate as research subjects, especially with respect to their compromised ability to give informed consent. This report also indicates the need for "other parties in order to protect the subject from harm":
Special provision may need to be made when comprehension is severely limited - for example, by conditions of immaturity or mental disability. Each class of subjects that one might consider as incompetent (e.g., infants and young children, mentally disabled patients, the terminally ill and the comatose) should be considered on its own terms... Respect for persons also requires seeking the permission of other parties in order to protect the subjects from harm. Such persons are thus respected both by acknowledging their own wishes and by the use of third parties to protect them from harm.
The third parties chosen should be those who are most likely to understand the incompetent subject's situation and to act in that person's best interest. The person authorized to act on behalf of the subject should be given an opportunity to observe the research as it proceeds in order to be able to withdraw the subject from the research, if such action appears in the subject's best interest. (National Commission 1979, p. 6) (emphasis added)
These recommendations of the National Commission are quite explicit and clear about the need for vulnerable human subjects, specifically persons with mental illness, to be protected from harm by third parties, who may even observe the research and act in these subjects' "best interests." In current Federal Regulations this language will shift to persons "knowledgeable about" the patients' illness, and to surrogates who make "substituted judgments" about what the patient would decide if he/she were competent.
FEDERAL REGULATIONS (1981-1989)
From 1981-1989, the Federal Regulations were subdivided into four major subparts: A, B, C and D. Generally speaking, Subpart A covered protection for research subjects who were competent and who could therefore give informed consent. The additional subparts provided special protections for several groups of research subjects who were considered "vulnerable", particularly insofar as their ability to give true informed consent was in some way compromised. Thus Subpart B provided extra layers of protections for fetuses, pregnant women and in vitro fetuses fertilized ex utero. Subpart C included prisoners. And Subpart D referred to children.
However, no special subpart was provided for that category of research subjects who are persons with mental illness. The only reference to these human subjects comes in Subpart A, in a discussion concerning IRB membership:
If an IRB regularly reviews research that involves a vulnerable category of subjects, including but not limited to subjects covered by other subparts of this part, the IRB shall include one or more individuals who are primarily concerned with the welfare of these subjects. (Code of Federal Regulations 45 CFR 46, 1981, 1983, 1989, p. 7) (emphasis added)
To our knowledge, there is no strong evidence to indicate that NIH has required IRB's to include either persons with mental illness, members of their family, NAMI or any of the state AMI's to serve on these IRB's on a permanent basis. Our interpretation is that they have construed "individuals who are primarily concerned with the welfare of these subjects" to mean psychiatrists, nurses, ethicists, and lawyers. If this is their interpretation, then we obviously disagree strongly with such an interpretation. Regardless of the interpretation, the fact remains that to our knowledge, no NAMI members are involved at all. Therefore, we suggest that NIH has not appropriately enforced the Federal Regulations as stated above with respect to vulnerable patients. This is probably due in large part to the fact that very few have objected to their policy -- and thus the policy has remained in effect to the present time.
Three things should be pointed out at this point. First, despite the clear concern for the use of persons with mental illness in the National Commission's Report, these series of Federal Regulations do not identify them as requiring any specific protections in a separate subpart as is afforded the other vulnerable groups. Second, supposedly they will be provided sufficient protections through an IRB -- a point to be seriously debated. Third, even when simply under an IRB protection, the IRB is to include one or more individuals who are "primarily concerned with the welfare of these subjects". The language descriptive of these additional IRB members will change -- we think significantly -- in the present 1991 Federal Regulations.
THE PRESIDENT'S COMMISSION REPORT
In 1983, the President's Commission issued its final recommendations in their volume, Summing Up. The Commission expressed obvious concern for the lack of attention to protections for certain vulnerable human subjects in its section on "Protecting Human Subjects":
(5) The National Commission's recommendations on research involving children and the mentally disabled should be acted upon promptly. Ethical concerns about these individuals revolve around the issue of informed consent. In order for research on the causes, treatment, and prevention of pediatric diseases and of emotional and cognitive disorders to proceed in an ethically acceptable manner, the National Commission had urged the adoption of special protections for children and the mentally disabled. Although the Secretary, HEW (now HHS), was to respond promptly to that Commission's recommendations, it has been four years since those recommendations have been submitted...
The Department has partially responded to Recommendation (5) regarding rules to protect children and the mentally disabled who are asked to participate in research. After further encouragement from the Commission during 1982, HHS on March 8, 1983, published in the Federal Register regulations for research involving children. Although the Commission has continued to urge HHS to act expeditiously to remove regulatory ambiguities and impediments that may exist to research with mentally disabled subjects, Secretary Schweiker has informed the Commission that no regulations will be issued by HHS research involving persons institutionalized as mentally disabled. (President's Commission 1983, p. 54, 56) (emphasis added)
Thus, despite these strong recommendations of both the National Commission and the President's Commission, under present Federal Regulations persons with mental illness may participate in both therapeutic and experimental research as human subjects with only those protections afforded them by an IRB. Furthermore, even those protections apply only to those who are institutionalized, and neglects those persons with mental illness who are not institutionalized but participate in research as human subjects in out-patient care, physicians offices, etc.
It is clear to any objective observer that those inflicted with serious diseases of the brain (which affect one's ability to think -- and therefore affect their ability to give true informed consent) are more vulnerable than those with serious diseases of other organs of the body. It is also clear that all other special groups had strong advocacy groups at the time. The mentally ill cannot advocate for themselves, and thus this responsibility was then, in part, assumed by the researchers in psychiatry -- those who would have benefited themselves by using the mentally ill in their research protocols. Is this not a conflict of interests? It is also important to note that the National Alliance for the Mentally Ill (NAMI) was only started in 1980, and thus not available for discussions on these issues. Thomas Jefferson's comments on the lack of efficiency inherent in a democratic form of government could be applicable here. The lack of an advocacy group for persons with mental illness at the time of the hearings for the proposed Federal Regulations contributed to the present situation, where persons with mental illness are not even recognized as a vulnerable group at all. Additionally, all of these vulnerable groups may be used in research with "certain exemptions" from the standard protections! This unfortunate situation is captured by the remarks of Professor Jean-Louis Baudouin, QC, Professor of Law at the University of Montreal, Canada. She summarized the American position on the use of persons with mental illness in an article on medical ethics in the journal Medicine and Law (1990): "One is surprised, however, to find no specific provisions relating to mentally handicapped persons."
PRESENT NIH CLINICAL CENTER POLICY
In 1986 the Clinical Center at NIH approved a new policy for the consent process in clinical research with patients who are or will become "cognitively impaired". It is certainly laudable that the Clinical Center, on its own initiative, took extra steps to afford more protection for cognitively impaired research subjects, yet several concerns remain.
Aside from obviating the current debate as to what constitutes "cognitively impaired" as opposed to "organically impaired", the policy, which was adopted in 1987, included a new effort to assure that the consent process is appropriate. This new protection was afforded by a Durable Power of Attorney (DPA). They state that their use of the DPA was modeled after a Maryland State law which, they claimed, extended the use of the DPA (originally used for purposes of property management) to allow competent Maryland residents to select a surrogate in advance who could later consent to procedures in health care if that resident should become incompetent to make these decisions in the future (Clinical Center 1987; Fletcher 1985, p. 1).
Under this new Clinical Center policy (which applies only to intramural research in the Clinical Center), the Institute Clinical Research Subpanel (ICRS) -- or the CC's IRB -- may give approval for a subject who is "not seriously impaired" to give informed consent to select a surrogate decision maker. The DPA is the record of the impaired subject's choice of a surrogate. That is, human subjects who are or will become cognitively impaired may appoint a surrogate to make decisions for the subject about his or her participation in medical research at the NIH. Thus, the surrogate would provide the "best substituted judgment that the subject would consent to the research if he or she were not impaired" (Clinical Center 1987, p. 2; Fletcher 1985, p. 1-3).
We have several concerns with this present Clinical Center policy. First, in 1987, when this policy was written, the Maryland State DPA statute was generally applied to the management of financial affairs only. (a) It was unclear if it could apply to health care decisions. In 1987 one Maryland attorney general stated in an opinion that the generic DPA could possibly apply to health care decisions if the document specified that intent. (b) It's application to medical research decisions had not really been addressed, although it could possibly have applied if executed by a competent person and so specified in the document. (c) It could only be executed by competent persons, not by never or presently incompetent persons -- i.e., the category of research subjects participating in medical research at the Clinical Center. In October of 1993, a new Health Care Decisions Act will go into effect which specifically authorizes the use of the DPA for health care decision makers. But to use it for never competent persons, or persons who lack decision making capacity at the time the document is executed, would not be consistent with either the current Maryland law or the new (1993) Health Care Decisions Act (D. Hoffmann 1993).
Second, there is already a growing concern about the acceptance of the use of the concept of "substituted judgment" in many health care contexts (e.g., Case Studies 1992; Nelson 1992; King 1992; Truog 1992; Capron 1992; Santurri 1982; Dresser 1984). Third, this policy applies only to research with cognitively impaired human subjects, and does not include persons with any of the other mental illnesses that are being studied.
Fourth, one has to wonder how a subject who always has been and who is still cognitively impaired can competently give informed consent to choose a surrogate who will make such important decisions and "substituted" judgments for him/her. There are some who would argue that:
...It may seem anomalous to say that a subject is, at the time of appointment, incapable of consenting for herself. However, it is not unusual to see situations in which a subject may be capable of understanding that someone may act for her in making a decision and of naming a trusted person to perform that function, even though the subject is not capable of understanding the risks and benefits associated with a complicated research protocol (Fletcher, 1985, p. 2).
This analogy between a subject's ability to choose who to trust, and the ability to know that that person would -- or could -- know if the subject would choose to participate in research if the subject were competent, would still remain an "anomaly" to many.
This rationale is similar to that used in determining the capacity of the research subject. For example, in such determinations, the assessment of a patient's actual functioning in decision-making situations is given precedence over either the outcome of a patient's decision or the patient's status (Fletcher, 1985, p. 4). The relevant issue, they claim, is whether the individual is capable of making a particular decision. However convincing this may seem to some, it is clearly very possible that there is a distinct difference between the mere ability to make a decision, and the kind of decision that one is making. Arguments about imposing other's preferences on these subjects aside, there is at least a broad area of decision making which is considered in most other contexts of life as "competent" and "rational", and past that point the kind of decision that is made is definitely relevant to the determination of competency (unless one is willing to entertain the notion that a person is "competent to be irrational"). In the genuine concern to restore and protect the individual rights and dignity of incompetent persons, perhaps care is needed not to extend their "autonomy" so far that these persons are instead actually injured or neglected (see Shamoo and Irving, 1993; Ganzini, Lee and Bloom 1993).
Another determination that is used is that, except as noted, individuals should be assumed to possess decisional capacity unless otherwise demonstrated; incapacity should be found to exist only when the individual lacks the ability to make decisions that promote his/her well-being in conformity with his/her own previously expressed values and preferences. However, it is not unusual for an incompetent person to appear to be competent when they are not; many learn how to play that game. In addition, it does not follow that if a person's expressed values are to slam his head against the floor every five minutes, and that therefore if he expresses that he now desires not to do so, that that person is incompetent. These directives as to the determination of competency or incompetency need to be seriously addressed by as large a body of responsible experts as possible.
Fifth, of equal concern is the provision in the Clinical Center policy that when the subject is so seriously impaired as to be incapable of understanding the intent or meaning of the DPA process, then a next-of-kin surrogate may be chosen by the physician, with an additional consultation with the bioethicist from the CC Bioethics Program -- who may delegate this responsibility to other consultants approved by the Director of the Clinical Center. This includes participation in therapeutic research. This NIH practice would probably not be consonant with most past or present state laws in this area which provide a definite ranking of family members for the use of substituted consent. Under the current substituted consent law in Maryland (Health Gen. Sec. 20-107) family members in a certain order could make health care treatment decisions for an incapacitated patient, but it is "highly unlikely that medical research would be considered such treatment under the statute". The law will be repealed in October 1993; and the new Health Care Decisions Act would not likely be interpreted to allow an agent, including family members, to consent to the use of incapacitated patients in medical research unless the patients were competent at the time the document was executed and unless specified explicitly in the document. For never or presently incompetent patients, even family members probably could not consent for such purposes. In any event, under most state laws, the choice of who the surrogate would be would not depend on the physician, or the CC bioethicist, or the Director of the Clinical Center, etc. -- but on the family ranking as already specified in the law. (D. Hoffmann 1993). Certainly some clear rationale openly discussed and accessible to the public should precede a policy which would use institutionalized seriously cognitively impaired human subjects in any kind of medical research.
It would appear that what has transpired up to the present, then, is a constant shifting of the accountability involved in the use of persons with mental illness in research. No Federal Regulations have incorporated the earlier calls for special protections for this vulnerable group. And even the purported protections afforded by an IRB have now been transferred in part to surrogate decision-makers under the "legal protection" of a dubiously valid DPA -- a surrogate who may be appointed by a cognitively impaired person, or his/her physician to participate in medical research, with the approval of the CC bioethicist, who may delegate this authority to "other consultants". Who, then, is clearly legally responsible and accountable for any harm which the research subject sustains as a result of the experiment? And who is responsible, now, for the "welfare" and the "protection" of the research subject?
THE PRESENT FEDERAL REGULATIONS (1991)
In 1991 the Federal Regulations were amended to incorporate the relevant provisions required in the Common Rule (1991). These present regulations make no changes in reference to any special protections for persons with mental illness, other than, as before, an IRB. However, there is a subtle change in the language concerning IRB membership when the use of vulnerable subjects are to be considered:
...If an IRB regularly reviews research that involves a vulnerable category of subjects, such as children, prisoners, pregnant women, or handicapped or mentally disabled persons, consideration shall be given to the inclusion of one or more individuals who are knowledgeable about and experienced in working with these subjects. (Code of Federal Regulations 45 CFR 46, 1991, p. 7) (emphasis added)
Our concern is the shift from the language "shall include one or more individuals who are primarily concerned with the welfare of these subjects" as used in the 1981-1989 Federal Regulations, to "the inclusion of one or more individuals who are knowledgeable about and experienced in working with these subjects."
The primary goal, as stipulated in a long series of recommendations and regulations, is the respect, protection, and welfare of these vulnerable human subjects. Such protection may not necessarily be afforded if the IRB member is now simply someone who is "knowledgeable about and experienced with these subjects." Clearly, if this role is now played by a physician researcher, for example, who has personal interests in the outcome of the experiment, there is a serious concern for a conflict of interests.
So the question is not only why were persons with mental illness dropped as a special category of vulnerable subjects in the Federal Regulations, but also who is accountable for these vulnerable patients, and who is protecting them and truly looking out for their welfare? At present, no regulations cover research subjects with mental illness who are not institutionalized; those who are institutionalized are only covered by an IRB, with a board member who only has to be "knowledgeable about and experienced with" these kinds of human subjects; only an IRB policy at the NIH Clinical Center covers in more detail such protections, but those protections apply only to research subjects who are cognitively impaired, involves the use of a dubious DPA which has been authorized by a cognitively impaired subject, and if they are severely impaired may be authorized by the physician researcher himself.
It is time, we think, for a full reexamination of several of the concepts and rationales which have guided the use of persons with mental illness in both therapeutic and experimental research for many years. Such a reexamination should include at least such questions as:
1. Are vulnerable groups of human subjects, e.g. persons with mental illness, really being exploited because of: their institutionalization; the impaired status of their ability to give true informed consent; sometimes the absence of any family members to protect them, e.g., the homeless, elderly; undue pressure from other members of families with vulnerable subjects to "eradicate" the illness for future generations; bench or physician researchers who benefit professionally from the research, etc.?
2. What should be the criteria by which to determine if a patient is "competent" or "incompetent"; and who should determine these criteria? Who should be allowed to make the determination that a particular patient is competent or incompetent? How accountable should they be for their diagnosis?
3. How should diseases which afflict persons with mental illness be classified, and who should decide? How accountable should they be for their classifications?
4. Is it true that persons with mental illness must be used in therapeutic research for their class of illnesses because such information is "unprocurable by other means"? Is "unprocurable by other means" an absolute societal value; or are there possibly some sorts of experiments that shouldn't be done or information which must remain "unprocured" simply because to obtain it would involve the unethical or illegal use of human subjects?
5. Should persons with mental illness be allowed to participate in medical research which is: only for their direct benefit, and/or for the benefit for classes of individuals with that same disease, and/or for the benefit of purely scientific knowledge and the greater good of society? Who decides these issues? Who is accountable for these conceptual decisions?
6. Is it medically, ethically or legally accurate or realistic to claim that persons with mental illness are competent to choose a surrogate to make a "substituted judgment" as to whether or not they [the patients] would freely participate in therapeutic or experimental research if they [the patients] were competent?
7. Should the appropriate "third party" role be to give a substituted judgment about what the surrogate thinks the patient would want to do, or to give an informed judgment about the best interests of these subjects?
8. Can a Maryland DPA law, based on the protection of the property of a citizen, and now even questionably applicable to only presently competent persons to choose a surrogate to make health care decisions after they have become incompetent, be appealed to and be the justification for the on-going practice at the Clinical Center of allowing cognitively impaired patients to choose someone to make a substituted judgment for them to participate in medical research?
9. Precisely who is legally responsible for injuries sustained not only during the course of experiments, but for symptoms which may appear some time afterwards (the entire IRB, the CC Director, the researcher, the attending physician, the surrogate holding the DPA, the next-of-kin, the "other consultants" appointed by the CC bioethicist)?
10. Should the role of some one who is knowledgeable about these vulnerable patients be separated from the role of some one who is concerned about the welfare of these patients? Perhaps there should be both kinds of members represented on the IRB boards.
1. Public and academic comment should begin again on many of the issues surrounding the use of persons with mental illness in therapeutic and experimental research.
2. We strongly urge the NIH OPRR to immediately enforce and monitor compliance by placing individuals concerned with the welfare of persons with mental illness on all IRB committees dealing with research using these vulnerable subjects. Moreover, all such individual IRB guidelines should immediately reflect this change. NAMI members should be tapped for this function. This could include the use of NAMI members as "consent auditors", as monitors of the experiments themselves, and as possibly legally approved surrogates to advocate for the welfare and best interests of persons with mental illness who have no reliable family members, next of kin or friends.
3. NIH should immediately incorporate the relevant recommendations of the National Commission with its original language into all of its intramural and extramural research programs.
4. The Common Rule of the Federal Regulations (45 CFR 46) should be amended to include the recommendations of the National Commission. We further recommend that the rules be applied to all persons with mental illness -- without separating them into institutionalized and non-institutionalized groups. Both groups are vulnerable and need protection -- whether in an institution, a regular health care facility, an out-patient clinic or a physician's office.
The authors strongly support most research efforts, and especially research that improves the treatment of persons with mental illness. Our call for recognizing them as a vulnerable group requiring special protection from potential abuse in no way diminishes that support, nor is it incompatible with our strong support for research in general. But we all have a moral obligation to ourselves and to our society to ensure proper, ethical and humane treatment for persons with mental illness when they are used as subjects in any kind of research.
1. Appelbaum, P., Lidz, C. and Meisel, A. (1987) Informed Consent - Legal Theory and Clinical Practice. Oxford: Oxford University Press.
2. Barber, Bernard. (1973) Research on Human Subjects. New York: Russell Sage Foundation.
3. Baudouin, J. (1990) Medicine and Law 9:1052-1061.
4. Beecher, H. E. (1966) Ethics and Clinical Research. New England Journal of Medicine 274:1354-60.
5. Capron, A. M. (1992) Where is the sure interpreter? Hastings Center Report 22:26-27.
6. Case Studies. (1992) Substituting our judgment. Hastings Center Report 22.
7. Dresser, R. (1984) Bound to treatment: the Ulysses contract. Hastings Center Report 14, 3:13-16.
8. Fletcher, J., Dommel, F. and Cowell, D. (1985) A trial policy for the intramural programs of the National Institutes of Health: consent to research with impaired human subjects. IRB 7, 6:1-6.
9. Frankel. (1975) The development of policy guidelines covering human experimentation in the United States. Ethics in Science and Medicine 2:43-59.
10. Ganzini, L, Lee, M. A., Heintz, R. T. and Bloom, J. D. (1993) Is the Patient Self-Determination Act appropriate for elderly persons hospitalized for depression? Journal of Clinical Ethics 4, 1:46-50.
11. Gillespie, R. (1989) Research on human beings: an historical overview. Bioethics News 8, 2:4-15.
12. Hoffmann, D. (1993), Associate Professor of Law, University of Maryland, Baltimore, MD (personal communication, May 13, 1993).
13. Irving, D. (1993) The impact of scientific misinformation on other fields: philosophy, theology, biomedical ethics, public policy. Accountability in Research 2, 4:243-272.