The ethical basis of psychiatric research: Conceptual issues and empirical findings

Comprehensive Psychiatry (Official Journal of the American Psychopathological Association)

VOL. 39, NO. 3

MAY/JUNE 1998

The Ethical Basis of Psychiatric Research: Conceptual Issues and Empirical Findings Laura Weiss Roberts The study objective was to clarify the ethical basis of psychiatric research by outlining conceptual issues and empirical findings related to the ethics of human experimentation. A systematic review of scholarly and empirical literature covering three decades was performed. The ethics of human research has come to be understood in relation to the principles of respect for persons, beneficence, justice, and the ideal of informed consent. Subjects who cannot fully participate in informed-consent processes are especially vulnerable to exploitation. Current dilemmas stem from insufficient attention to the vulnerabilities of mentally ill research participants, problems in applying ethical concepts and guidelines to psychiatric research, and claims of research misconduct. Empirical studies indicate that (1) psychiatric symptoms significantly affect informed consent, (2) psychiatric patients may possess certain strengths with respect to research involvement, (3) proxy decision-making is problematic, (4)

informed consent is also difficult to attain with the medically ill and others, (5) patients are motivated to participate in research by the hope of personal benefit, (6) ethical aspects of research are poorly documented, and (7) institutional review processes may not be adequate to protect vulnerable subjects, Psychiatric research can be performed ethically, according to standards set throughout the biomedical and behavioral sciences, so long as researchers and institutions are respectful of special ethical issues in human experimentation and strive to include vulnerable study participants fully in research decisions. However, many gaps in the empirical literature exist regarding the specific nature and implementation of ethics principles in psychiatric research. Efforts to advance both science and ethics, including the study of ethical dimensions of human research itself, are essential for the future of psychiatry.

MMENSE CONTROVERSY has surrounded the question of whether it is ethical for individuals with mental illness to participate in research.~-~4 Those who believe that clinical psychiatric research is unethical raise concerns about poor informed-consent procedures, the dangers of human experimentation, researchers' lack of training and skill related to ethics, and the potential exploitation of individuals with severe mental illnesses. On the other hand, advocates of clinical research argue that "protecting" psychiatric patients from protocol participation needlessly usurps their autonomy and deprives them of opportunities for personal benefit. They further suggest that scientific neglect of psychiatric disorders represents a great injustice toward millions of individuals now and in the future. As this debate has unfolded over the past several years,l 1-23inadvertent and intentional unethical practices by psychiatric researchers have come to light, some scientific endeavors in the area of mental illness have been curtailed, and trust in the field of

psychiatry has been jeopardized. Moreover, the research and educational missions of academic medical centers have become more tenuous as the rate of increase of overall federal funding has leveled and private, for-profit funding of experimentation has significantly increased. 15.24-29This transition has led to unexpectedly difficult conflicts of interest, broad changes in research priorities, and new ethical dilemmas. 15,24-29 Given this unprecedented context, growing tension exists between our need to protect society's most vulnerable

I

Copyright03 1998by W.B. Saunders Company

From the Department of Psychiatry, University of New Mexico School of Medicine, Albuquerque; and Consultation Psychiatry Service, University Hospital, Albuquerque, NM. Supported by a Young Investigator Grant from the National Alliance for Research on Schizophrenia and Depression. Address reprint requests to Laura Weiss Roberts, M.D., Department of Psychiatry, UNM Health Sciences Center, 2400 Tucker NE, Albuquerque, NM 87131-5326. Copyright 03 1998 by WB. Saunders Company O0l 0-440X/98/3903-0009503.00/0

Comprehensive Psychiatry,Vol. 39, No. 3 (May/June), 1998:pp 99-110

99

100

members from exploitation and our ability to lessen their suffering in the years to come. To help clarify the ethical basis of psychiatric research, this report describes the ideas and events that have shaped our understanding of ethical issues in human experimentation, including key ethical dilemmas currently faced in psychiatric research. It then outlines the empirical literature on informed consent and vulnerability in mental and relevant medical illnesses and related areas. On this foundation, it is suggested that psychiatric research can be performed within the ethical standards of the biomedical and behavioral sciences so long as investigators conduct their studies in a manner that reflects knowledge, attention, and respectfulness regarding the special issues inherent to experimentation with vulnerable mentally ill individuals. Moreover, institutions in which psychiatric research endeavors are pursued should meet or exceed professional ethical guidelines, for instance, with regard to institutional review board (IRB) functioning and implementation. However, to perform ethical research in the future, we must also endeavor to anticipate, investigate, and respond constructively to the ethical dilemmas newly emerging in our profession. CONCEPTUAL ISSUES: EMERGENCE OF A "NEW ETHIC" FOR HUMAN RESEARCH

Prompted largely by the medical profession's past mistakes, an increasingly sophisticated conceptual understanding of the ethical aspects of human research has emerged since World War 11.30-40 Tragic studies were conducted in our country, most clearly documented as beginning in the 1940s. In the Willowbrook Study, for example, profoundly mentally impaired children were intentionally infected with hepatitis to observe medical outcomes, and similarly in the Tuskegee Study, 200 black men who had contracted syphilis were purposefully deceived and not treated. 31-33In other experiments, several hundred retarded and imprisoned individuals were given unproven influenza vaccines and then later "challenged" with the infection by researchers at an eastern university, and 500 inmates at Statesville's penitentiary were infected with malaria by researchers at a medical school in the Midwest. 32 In a lesser-known example, 76 poor Hispanic women in San Antonio who had requested contraceptives were instead given placebos and not informed. 31 For the most part, these activities were

LAURA WEISS ROBERTS

not hidden by clinical investigators except, ironically, from their research participants. 3°-33 The predominant value of acquiring scientific knowledge was so strong during this postwar era that it simply eclipsed the ethical considerations related to human experimentation in the United States. This lack of recognition changed abruptly in 1966 with the publication of a landmark article in the New England Journal of Medicine. 31-34In this report, Dr. Henry Beecher described unethical practices in 22 experiments and made an urgent plea to halt the use of humans in nontherapeutic clinical studies. 34 He cited, for instance, protocols in which penicillin was withheld from soldiers infected with streptococcus, new cardiac catheterization procedures were performed on patients who had not been told, and live cancer cells were injected into patients without full consent. Beecher's report riveted the attention of the public and the medical and scientific communities. For a country quietly troubled by the use of human beings in experimentation since the Nuremberg Medical Trial, it became evident that a new ethic was needed for human research. Intensive efforts were then undertaken to develop federal guidelines for ethical conduct of human research in the United States. 30,35-36 IRBs were mandated by federal regulation in 1974 to assess the scientific merit of research protocols, to protect human research participants, and to ensure compliance with regional and national regulations related to experimentation. Federal guidelines for research with pregnant women, fetuses, and prisoners were issued in 1978. In 1979, a visionary statement, the "Belmont Report," was developed by the President's Commission, outlining three principles governing the ethical conduct of research: (1) respect for persons, promoting the dignity and autonomy of research participants; (2) beneficence, insisting that study design be scientifically rigorous, avoid known harms, and intend to bring about maximal good and minimal injury; and (3) justice, ensuring that vulnerable groups not be taken advantage of and that the segment of the population bearing the burden for research also benefits from itY From 1980 to 1983, a second President's Commission reviewed all federal policies regarding human research to assess their adequacy and implementation. 36 In 1983, federal guidelines for protecting children in research were issued. In 1991, a final common federal policy (the

ETHICAL BASIS OF PSYCHIATRIC RESEARCH

"Common Rule") for human research was adopted by executive order. Throughout these endeavors, ethically acceptable research has been conceptualized in relation to the ideal of informed consent. 3°-37 Informed consent is a philosophical and legal doctrine requiring that individuals truly understand and freely undertake involvement in an experiment or clinical care. 38In human research, informed consent presupposes the integrity and competence of clinical investigators and entails three key elements. 3°31.37-38First, it involves the sharing of information regarding the proposed experimental risks, benefits, and alternatives, and the natural progression of the illness without intervention. Second, the individual must also be capable of communicating, comprehending, rationally assessing, and fully appreciating the nature and consequences of the decision at hand. 3~,38 Finally, informed consent requires the authentic voluntarism of the patient, i.e., expression of the patient's true wishes in the absence of coercion. In retrospect, informed consent also has become a critical measure of human subject vulnerability. Individuals whose consent is compromised due to insufficient information, deficits in fully understanding the meaning of a research decision, or coercion are especially susceptible to exploitation. Vulnerable populations are therefore very diverse and include such groups as the impaired elderly, whose cognitive capacities may diminish over time; terminally, severely, or chronically ill patients, whose hopelessness and despair may influence their judgment; psychiatrically ill patients, whose decisional abilities may fluctuate and/or be reversible; children, fetuses, and developmentally disabled individuals, who have never been fully decisionally capable and cannot advocate for themselves; women of child-bearing age or pregnant women, whose health or other interests may be in conflict with those of the fetus; very poor people, for whom even modest financial incentives may seem irresistible; "captive" or dependent individuals (e.g., prisoners, hospitalized patients, nursing home residents, and veterans), whose sense of autonomy and freedom may be radically limited; and ethnic minorities, who may experience societal and interpersonal powerlessness.l-I 8,21,30-45 By broad consensus, the inclusion of members of vulnerable groups in research is considered defensible only if two conditions are met: if members of

101

less exploitable groups cannot, for scientific reasons, serve as participants in the experimental protocols, and if the protocols have absolutely met criteria for being ethically sound (Table 1).31 Even when experimentation with vulnerable participants is necessary and justifiable, their involvement entails greater attentiveness to ethical issues and strict guidelines (Table 2). 1 IRBs are one mechanism by which adherence to scientific, ethical, and legal guidelines is ensured. 3',35 Every institution that receives federal research funding must have an IRB process to perform prospective review and to monitor ongoing protocols. The IRB is empowered to approve, disapprove, or require modifications in proposed studies. This institutionally based approach allows for greater attention to local or regional concerns related to research conduct. IRB activities across the United States are governed by the Office for Protection for Research Risks of the National Table 1. Guidelines for Ethically Sound Research • The experiment must represent "good science," i.e., the question posed must be valuable and justified and the study design must be appropriate. • The experiment must undergo appropriate prospective peer and institutional review processes given the level of risk undertaken by study participants. • The investigators must possess integrity and competence to ensure that experimental work will be conducted with respect and expertise. • The research institution must have adequate resources and commitment for conduct of the project. • Physical, psychological, social, and economic risks to participants must be assessed and minimized, as should analogous risks to society. • Potential benefits must be optimized for both individuals and populations. • Confidentiality protections must be rigorously upheld. • The consent process itself must include a clear invitation to participate in a study; a statement of the purpose of the study and the basis of subject selection, an explanation of procedures, discomforts, and risks, a plan in case of injury or untoward consequences, potential benefits, alternatives to study participation, and financial considerations; assurances of confidentiality; opportunities for continuing disclosure; and measures to ensure that the subject's decision is understood and voluntary. • Special care must be given to subject selection and sites of recruitment so that disadvantaged individuals or populations are not used without sufficient reason. • The incentives to participate in research must be adequate but not coercive. • The documentation of the scientific process must be accurate, accountable, and thorough. Data from Levine. 31

102 Table 2. Additional Ethics Guidelines for Research Involving Mentally Disabled Subjects • Enhance the autonomy of capable study subjects and accurately identify subjects incapable of independent choice. • Offer capable subjects an opportunity to share decisionmaking with a family member or friend. • Instruct proxy decision-makers on ethical standards for research participation. • Exclude objecting decisionally incapable subjects from research unless an exceptional opportunity for direct benefit exists. • Improve assessment of risks and benefits of research involving decisionally incapable subjects. • Be aware of subjects who are vulnerable to becoming decisionally incapable during a study. • Arrange for meaningful representation of the vulnerable subject population when research is planned and reviewed for conformity to community standards. Data from Dresser?

Institutes of Health and the Food and Drug Administration. Three critical issues now face psychiatric research ethics: (1) insufficient attention to the vulnerabilities of mentally ill research participants, (2) challenges in applying the "new ethic" to psychiatric investigation, and (3) claims of intentional and inadvertent misconduct by psychiatric investigators.

Insufftcient Attention to Vulnerabilities of Mentally Ill Research Participants Patients who are mentally ill are open to harm in human experimentation for several reasons. Psychiattic syndromes are often characterized by symptoms such as ambivalence, apathy, paranoia, isolativeness, lability, cognitive deficits, impaired insight, self-destructiveness, or impulsivity. Such characteristics may prevent psychiatric patients from fully engaging in the consent process, which requires a capacity for gathering information, establishing trust, identifying risks, advocating for oneself, making choices, and adhering to decisions. Furthermore, it is unclear how to apprise patients adequately of the risks that accompany psychotropic drug studies with design features such as randomization, placebo comparisons, washout periods, double-blind methods, and unproven treatments. ]20,3o-44Such experiments present extraordinary consent problems for all human participants but particularly for psychiatric patients, who may have more difficulty appreciating and tolerating the impact of unexpected consequences. Efforts to obtain ad-

LAURA WEISS ROBERTS

vance consent (the "Ulysses Contract") in these situations have met with serious moral and legal problems. 43-44 Psychiatric patients may also be easily coerced into research participation because of their suffering, tremendous need for health care services, institutionalization experiences, and relative powerlessness on interpersonal and societal levels. Finally, because mental illness affects individuals from all cultural and social backgrounds, it is not uncommon for psychiatric patients to possess multiple sources of vulnerability, e.g., a reproductive-age black woman institutionalized due to schizophrenia, a poor elderly man with Alzheimer's dementia, or a depressed and stigmatized patient with AIDS. Research with mentally ill patients thus "tests the mettle" of each element of informed consent and creates special duties for psychiatric investigators to protect their study participants.

Problems in Applying the New Ethic to Psychiatric Research Applying the new ethic to psychiatric research presents great challenges. First, it is well established that optimal scientific design and ethical protections can be at cross-purposes. Scientific uncertainty ("equipoise") is necessary to justify the activity of human experimentation, and yet this same uncertainty creates seemingly insoluble ethical problems when recruiting potential research subjects to participate in studies with unknown benefits and harms. 46 Placebo trials, for instance, might definitively demonstrate the effectiveness of therapeutic interventions, but very prolonged medication-free periods in protocols have been historically considered unethical when proven treatments exist. 12,46-48In addition, selection biases are introduced by the very process of obtaining informed consent from participants. 49 Finally, what constitutes "good-enough" study designs, i.e., those that minimize risk but may yield more ambiguous or less definitive scientific results, in ethically complex situations with psychiatric patients has not yet been determined.31 Second, as a nation, we have tried to develop new approaches to human subject selection as our understanding of justice and other social values has evolved. 35-4°,45,5° The considerable efforts to include more women and ethnic minorities in psychiatric research reflect this change in values. 39'45 In recent years, evidence has emerged that men and

ETHICAL BASIS OF PSYCHIATRIC RESEARCH

women take distinct paths in moral development, experience ethical questions differently, have different sensitivities to power and confidentiality issues in relationships, communicate differently, and possess both shared and unique cognitive strengths. 51-52 As women increasingly serve both as researchers and as participants in research, it is plausible that gender differences may affect scientific findings, as well as ethical dimensions of experimentation such as informed-consent processes. However, little is known about how research approaches might be adapted to obtain optimal and meaningful informed consent from individuals with varying illnesses or attributes (e.g., gender, ethnicity, social support, and socioeconomic factors). Third, protecting confidentiality in psychiatric research is critical because of the tremendous stigma experienced by individuals with mental illness. However, growing demands for access to government-sponsored medical research, widespread changes in computer technologies, and new areas of scientific inquiry (e.g., psychiatric genetics research) have recently raised the question as to whether societal interests may properly supercede the traditional value of individual privacy.3°,53 Fourth, the extreme suffering of patients with illnesses such as AIDS, cancer, and schizophrenia has led to pressure for the accelerated testing and release of investigational medications in the United States. 5° For these patient groups, protocol participation may serve as a means of access to desperately needed health care. Although frequently criticized, the conservatism of federal agencies and regulatory guidelines has been justified historically on the basis of early disturbing experiences with investigational drugs such as Thalidomide. 3°-31This dilemma is not easily resolved, and it is likely to become more complex as regional legal decisions emphasizing human subject protections, rather than the benefits of protocol participation such as access to basic health care and experimental therapies, influence national policies in human research. Finally, role conflicts and related conflicts of interest are inherent to clinical research. 3°-31,34-36 Investigators, for example, must serve both the well-being of their patients and the aims of their scientific inquiries. From the perspective of the patient, the integrity of the clinical investigator who is faithful to "two masters" at once is the source of both trust and mistrust. 34-36 In addition, there are many other parties beyond the investiga-

103

tor and patient who have interests in clinical research endeavors, including the public, the host institution, the sponsor (e.g., governmental agency or pharmaceutical company), and the scientific community. Potentially detrimental conflicts of interest commonly exist between and within each level of this hierarchical system. 15,21For this reason, institutional review processes may not offer sufficient protection to research subjects. 15,21,3°,31,3436,54,55

Claims of Misconduct As described in several s o u r c e s , 1,7,12,13,17,20,22'30 recent claims of unethical research practices with mentally ill participants at a few medical centers have kindled concern and mistrust in our field. The allegations are many and varied: exploitation of institutionalized and cognitively compromised individuals, misuse of placebos, exposure to inordinate risks in nontherapeutic experiments, inappropriate or insufficient consent processes, confidentiality breaches, evidence of coercion, undisclosed conflicts of interest, misappropriations of funding, and others. Such claims have raised doubts about the integrity of some psychiatric investigators and their ability to act respectfully, beneficently, and justly toward their research participants. EMPIRICAL FINDINGS: INFORMED CONSENT, VULNERABILITY, AND RELATED ISSUES

It is curious that the investigation of the ethical aspects of human experimentation, with a few important exceptions, has remained relatively neglected. 8,~7,56Psychiatric research ethics has undergone an elegant line of empirical inquiry primarily through the efforts of several dedicated researchers over the past 20 years, 6,8,11-14,31,37,41,57-59 although many gaps remain in the literature. Early findings in seven related areas nevertheless offer valuable insights into the ethical conduct of mental illness research.

Vulnerability and Informed Consent in Psychiatrically Ill Patients Psychiatric symptoms have repeatedly been found to influence the patients' cognitive ability to provide informed consent. 8,37,23,56,59-66 The 1995 MacArthur Treatment Competence Study was an extensive multicenter project that sought to provide information regarding clinical decision-making of patients with mental illness. 59 The study involved

104

498 individuals from three groups: patients with schizophrenia or major depression, patients with ischemic heart disease, and nonpatient community volunteers. The psychiatric patients in this study, when acutely ill, demonstrated more decisionmaking difficulties on several formal cognitive measures (schizophrenic patients more so than depressed patients) when compared with the medically ill patients and community participants. However, once treatment began, these decisional deficits greatly improved. On a smaller scale, Schachter et al. 61 used the Brief Psychiatric Rating Scale and other measures with 38 patients diagnosed with schizophrenia. Similar to the MacArthur study, conceptual disorganization and unusual thought content were associated with patients' poor performance on the information-based learning tasks related to consent. Not unexpectedly, studies of psychiatric patients with dementia and clear neurological damage have also revealed significant deficits with respect to informed consent and decisional capacity. Stanley et al. 57 compared the reasoning ability of elderly individuals with major depression (n = 45) and dementia (n = 38) with that of elderly control subjects (n -- 20) and found that cognitively impaired dementia patients had the greatest deficits on a number of "functional competency" measures. A recent retrospective report of 41 hospital cases in which medical treatment was refused showed that patients with significant organic deficits were far more likely to be decisionally incapable on multiple clinical measures than the other patients with adjustment, psychotic, affective, personality, and substance use disorders. 62 Prospective studies of the cognitive aspects of competency in dementia have been performed with promising results. Marson et al.63 described the use of a clinical vignette-based instrument to measure the level of competency exhibited by patients with Alzheimer's dementia and by normal individuals in medical treatment decisions. The severity of dementia, defined with strict cognitive measures, correlated with the competency status of the Alzheimer's patients based on a "rational reasons" standard. The same research group considered the predictive validity of neuropsychological tests in determining the decisional competency of Alzheimer's disease patients. 64 For patients in this study, cognitive functions requiring high frontal lobe activity such as word-finding were related to the patients' capac-

LAURA WEISS ROBERTS

ity to formulate rational reasons for treatment choice. However, global scores of dementia severity, such as the Mini-Mental Status Exam, were not predictive of this standard of competency. The emphasis of these two projects is more on the performance of cognitive tasks than on the assessment of patients' true and subtle "decisional" processes. Such studies may nevertheless help to identify key cognitive features presupposed by sound decisional ability. Elements beyond cognitive and informational factors, i.e., subjective factors such as attitudes, values, motivations, feelings, and insights of both patients and investigators, are critical to informed consent and research participation. Grossman and Summers 65 found that 85% of 20 patients diagnosed with schizophrenia were unable to retain information about a hypothetical medication offered during an experimental consent session, and yet 60% had "agreed" to take it for uncertain reasons. In an experiment aimed at enhancing study subjects' understanding of research, Benson et al.66 videotaped consent disclosure sessions with 48 psychiatric patients (24 with schizophrenia and 24 with depression) and investigators involved in psychotropic medication protocols. As with treatment studies, it was found that the factual knowledge and subjective appreciation of the interventions were limited in the patient groups. For example, a majority of patients did not exhibit an awareness of randomization and double-blind procedures, and only two of 48 patients fully understood that being involved in a protocol might narrow their treatment options. Interestingly, it was also found that 42% of the patients with schizophrenia and 71% of the patients with depression "made their decision to participate in the research prior to the formal disclosure.., when first approached... and before they received any detailed information concerning [the] prior research." This crucial study helped to begin the work of characterizing the subjective aspects of the consent- and experiencerelated dimensions of research participation. Finally, empirical investigation of the effects of depression and psychological stresses on end-oflife care preferences also underscores the relevance of subjective "psychological state" factors in the decision-making of vulnerable patients. 67-71 In the clinical literature, severely depressed elderly psychiatric in patients described by Ganzini et al. 7° expressed exaggerated levels of hopelessness and

ETHICAL BASIS OF PSYCHIATRIC RESEARCH

negativism when predicting the relative benefits of various future treatment options--perceptions that reversed and became less distorted once treatment for depression was implemented. Treatment also changed the severely depressed patients' plans for the future and therapy choices. Similar data demonstrating the impact of depression and stress on outlook have also been found with HIV patients. 7~ As nicely summarized by Carl Elliott, 3 "if a person is depressed, he or she may be aware that a protocol carries risks, but simply not care about those risks." Most studies on consent and mental illness focus on clinical treatment decisions. This distinction is important because clinical research decisions differ with respect to the uncertainties and risks encountered by subjects and because efforts to advance science may "trump" the well-being and interests of patients. 5,15,3° Comparison of clinical treatment and research decisions by patients suggests that the settings and relationships in which consent activities occur greatly influence their ethical character. Lidz et al. 6 closely observed consent interactions between doctors, researchers, staff, and patients in three diverse psychiatric settings (evaluation center, research ward, and outpatient clinic) over 4 years. They found that consent practices and decision-making varied significantly according to the context. Patient autonomy, for example, was more evident in outpatient care than in research endeavors. As yet, little is understood about the values, pressures, attitudes, and motivations of psychiatric patients leading to research participation.

Potential Strengths of Psychiatric Patients as Human Participants Psychiatric patients may possess strengths with respect to informed consent and human research participation. Kleinman et al. 6° performed a study with 48 outpatients with schizophrenia and found that a single informational session on the risks and benefits of neuroleptic medications had a positive and persistent impact even after 6 months. In another study comparing 25 schizophrenia patients with 15 medically ill patients, the psychiatric patients understood the specific risks and side effects of their medications more fully than the medical patients. 72 In a third, older study comparing 27 psychiatric and 38 medical inpatients, the mentally ill participants were no more likely to express a willingness to participate in risky human

105

experiments, and the severity of psychopathology was a poor indicator of mentally ill patients' inability to assess research risks. 58 Lastly, a 7-year follow-up study of 76 patients who had been in a placebo-controlled clinical trial found that there were no clear differences "in any measure of clinical or social outcome" between patients who received placebo and those who received active drug.19 These encouraging studies suggest the value of providing information, evaluating patient outcomes over time, and looking for the strengths of psychiatric patients in research activities.

Problems of Proxy Decision-Making in Research Proxy decision-making commonly occurs when patients are unable to make choices for themselves in clinical and research contexts. Two valuable studies have revealed serious problems in proxy consent regarding dementia. Sachs et al. 73 performed structured interviews with 42 early dementia patients, their family decision-makers, and a comparison group of 60 elderly individuals. The dementia patients were more willing to accept the risks of experimental protocols, and for different reasons, than were the family proxies. This suggests that choices surrounding experimentation are distinctly personal and not fully predictable. In contrast, Warren et al. 74 found that 31% of proxy decision-makers would allow nursing home patients to participate in experiments despite their belief that it would be against the patients' wishes. It has thus been argued that study participants, regardless of their competence, should always be included in research decisions. 73-75

Vulnerabilities of Medical Patients and Other Study Participants Informed consent is an ideal that is difficult to achieve with medical populations. In an early interview study with 43 pregnant women who participated in a labor-induction protocol involving experimental medication, 20% of the women were unaware that they were in a protocol, 30% felt they had inadequate information about the experiment, 45% felt they did not understand the protocol, 74% indicated that it "did not occur" to them to refuse participation, and 89% did not know about the experimental drug to be used. 55 However, trust in physicians was found to be a large factor in the patients' willingness to participate. Cassileth et al. 76 found similar results with 200 cancer patients who

106

were surveyed within 1 day of agreeing to--and signing consent forms for---chemotherapy, radiation therapy, or surgery. The majority of patients (75%) felt that the explanations offered during the consent process had helped them to decide about their care. Only 60% understood the nature and purpose of the procedure, and fewer (55%) could correctly identify a single risk of the proposed treatment they had agreed to. This study also echoed others with respect to the observation that consent forms are difficult to read and comprehend. 76-8° Moreover, a written survey of healthy control subjects and breast cancer patients in clinical trials (N = 142) showed that the cancer patients retained less but, importantly, were given less information in consenting to protocols (e.g., written descriptions, expected physical discomforts, and side effects).79 Finally, the Structured Interview Study 3° revealed that 4% of 1,882 patients were unaware that they had been in clinical studies in the past. Studies of research participants drawn from theoretically nonvulnerable groups also reveal problems related to informed consent. For instance, in one study, fewer than 20% of 65 college students who had participated in psychology experiments viewed the consent process as a true decision point; in fact, some students felt that they had entered a "contract" to participate in research by answering recruitment advertisements via telephone. Moreover, the consent form was commonly perceived as a "liability waiver" to protect researchers. 16Philips and Vazquez81 have commented on the vulnerabilities of "normal volunteers" who are discovered unexpectedly and disturbingly to have serious medical abnormalities during participation in the protocols. In addition, elderly individuals, even those without cognitive deficits, have been shown to be less capable of providing informed consent than other adults. 57,78 This body of research suggests that informed consent is precarious and research involvement may always entail some uncertainty and vulnerability.

Motivations for Research Participation Research subjects with severe medical illnesses apparently choose to participate in protocols in the hope that they will be personally helped by experimental therapies. 11,14,31-34,5°,8285 For this reason, research participation has been described as both a "desperate solution" and a "therapeutic misconcep-

LAURA WEISS ROBERTS

tion. ''3,14,84,86 In the Structured Interview Study, 30 patients described feeling helpless and "constrained to participate [in various experiments] because of their medical situation, not by their providers or the research investigators." Parents who consent to have their seriously ill children participate in studies feel similarly pressured. 83 More dramatic are the results of another study that examined the motivations of 27 cancer patients enrolled in phase I clinical trials, or "toxicity" experiments, s4 Although these protocols are not expected to offer the possibility of medical improvement, 85% of patients in this study decided to enroll in the hope that the investigational drugs might help them. The vast majority of patients (93%) felt that they understood most or all of the information they were given about the trials, and yet only one third could state the purpose of their phase I experiment. Most patients (89%) in this study believed that they had no better option than to participate. A third, older study was conducted with 104 cancer patients, 84 cardiology patients, and 107 members of the general public. 85 A large majority (71%) approved of patients serving as research subjects, primarily because they felt that medical research benefits society and increases medical knowledge. Nearly all (97%) felt that helping future patients would be extremely or moderately important to them if they were research participants. However, when asked why they personally would choose to be in research, advancing medical science was a far less common motivation (23 %) than the wish to obtain better personal health care (52%). Such findings are alarming in light of a well-performed study of 156 cancer patients by Cassileth et al. 86 showing a decreased quality of life and no increase in survival rates among those who had been in experimental protocols.

Underreporting of Ethical Aspects of Human Research Documentation of the ethical elements of research has been sadly neglected, an observation with scientific and ethical repercussions. 34,4°,49Edlund et al.49 reviewed publications in three psychiatric journals and discovered no mention of the potential selection bias introduced by including only "consentable" and "consenting" patients. 49 My colleagues and I performed a focused literature review of geriatric research revealing that ethical protections such as confidentiality measures, IRB

ETHICAL BASIS OF PSYCHIATRIC RESEARCH

approval, sites of subject recruitment, and other features are frequently not identified in published reports. 4° More thorough documentation will allow for greater accountability with respect to researchers' scientific methodology and ethical conduct with vulnerable populations in the future.

Adequacy of lnstitutional Review Procedures The ability of institutional review procedures to fulfill their charge of protecting "the rights and welfare" of vulnerable human subjects has been an area of concern since the 1970s),5,9,15,3°,31,35,36Early criticisms were related to establishing the domains of institutional review, whereas more recent questions have arisen regarding the appropriate application of research guidelines (e.g., risk determination, retrospective and prospective review processes, and experiments in facilities without federal research funding) and potential conflicts of interest of IRBs. 15A7 Empirical investigation to clarify the functioning, composition, and effects of IRBs in the United States is virtually nonexistent) 5,54,3° However, in a highly important study by Kodish et al., 54 32 IRB Chairpersons and 53 investigators at national cancer centers revealed that phase I oncology studies "face no more scrutiny or resistance in their institution's IRB process than other research protocols." Small studies of IRB activities at two institutions suggest "low disapproval rates." 15It is difficult to know whether this means that investigators are doing well prospectively in their efforts to protect human participants or if IRBs are doing poorly in their efforts to identify and prevent unethical research. In either case, systematic documentation of the work and impact of institutional review would clarify a number of the recent concerns about IRB effectiveness in protecting study participants while supporting meritorious research. In sum, much excellent work has been performed to clarify the interrelated areas of informed consent and vulnerability in the context of psychiatric and medical illness. It appears that psychiatric patients, particularly those who are highly distressed or neurologically compromised, may have greater difficulty with consent processes than medical patients. Information offered during consent disclosure sessions may be poorly comprehended, retained, and recalled by patients. Efforts to treat patients' symptoms, alleviate their suffering, and improve consent processes may lead to more

107

positive results. In general, patients appear to be motivated by the hope for personal benefit. This is a crucial area for inquiry in mental illness because it taps directly into the poorly understood subjective aspects of decision-making such as voluntarism and autonomy. Consent processes that occur in the context of a lasting physician-patient relationship without outside pressures may be ideal. How this might translate to research settings with inevitable, externally imposed structures and stresses remains uncertain. The strengths of vulnerable subjects and the vulnerabilities experienced by all research participants require further study. How we report what we do in our research endeavors is extraordinarily important to fostering scientific understanding and ensuring the ethical caliber of our conduct. Limited study in the area of IRB activities suggests that more rigorous efforts may be necessary to assess scientific merit and ethical protection in human experimentation. CONCLUSION

Accurate self-observation and careful selfreflection are highly valued tasks in clinical psychiatry and greatly enrich the ethical nature of our work with patients. Emphasizing these same tasks is a distinct contribution that psychiatry can make in the field of research ethics. Past conceptual and empirical work suggests that investigation of mental illness meets ethical standards within biomedical and behavioral sciences only insofar as researchers and institutions are attentive to the ethical aspects of experimentation with vulnerable participants and real efforts are undertaken to enhance psychiatric patients' participation in research decisions. Psychiatric investigators must possess integrity and competence, seek knowledge of both science and ethics, and demonstrate fidelity to the principles of respect for persons, beneficence, and justice and the doctrine of informed consent. Institutions, furthermore, must provide adequate support and structure to foster ethical research practices. While tremendous gaps in the empirical literature remain, we have discovered through such study that all human participants of research may possess both vulnerabilities and strengths. Most importantly, we have seen that there is much we can do to improve the ethical rigor of the research we conduct with human participants who suffer from mental illness. Psychiatric research poses special but perhaps

108

LAURA WEISS ROBERTS

not unique or insurmountable ethical dilemmas. Continuing to explore the moral aspects of human experimentation is itself an ethical imperative for clinical science and psychiatry. Remembering our past experiences, learning to apply what we have come to know about research ethics in new situations, and testing our hypotheses empirically are essential if psychiatric research is to be scientifically meritorious and ethically defensible. Finally,

as a profession, the field of psychiatry must anticipate and respond conscientiously to the next generation of ethical problems that will arrive with scientific and cultural discovery. ACKNOWLEDGMENT The author offers thanks to Drs. Joel Yager, Brian Roberts, Teresita McCarty, and Zach Solomon and to Megan Smithpeter for constructive suggestions on this report.

REFERENCES 1. Dresser R. Mentally disabled research subjects: the enduring policy issues. JAMA 1996;276:67-72. 2. Bonnie RJ. Research with cognitively impaired subjects. Arch Gen Psychiatry 1997;54:105-111. 3. Elliott C. Caring about risks: are severely depressed patients competent to consent to research? Arch Gen Psychiatry 1997;54:113-116. 4. Eichelman B, Wikler D, Hartwig A. Ethics and psychiatric research: problems and justification. Am J Psychiatry 1984;141: 400-405. 5. Kantor JE. Ethical issues in psychiatric research and training. In: Annual Review of Psychiatry, Vol 13. Washington, DC: American Psychiatric Press, 1994. 6. Lidz CW, Meisel A, Zernbavel E, Carter M, Sestak RM, Roth LH. Informed Consent: A Study of Decision Making in Psychiatry. New York, NY: Guilford, 1984. 7. Lieberman JA. Ethical dilemmas in clinical research with human subjects: an investigator's perspective. Psychopharmacol Bull 1996;32:19-25. 8. Meisel A, Roth LH. What we do and do not know about informed consent. JAMA 1981 ;246:2473-2477. 9. Munir K, Earls E Ethical principles governing research in child and adolescent psychiatry. J Am Acad Child Adolesc Psychiatry 1992;31:408-414. 10. US Congress, Office of Technology Assessment. Mental Disorders and Genetics: Bridging the Gap Between Research and Society. Washington, DC: US Government Printing Office (OTA-BP-H- 133), 1994. 11. Appelbaum PS, Roth LH, Lidz CW. The therapeutic misconception: informed consent in psychiatric research. Int J Law Psychiatry 1982;5:319-329. 12. Appelbaum PS. Drug-free research in schizophrenia: an overview of the controversy. Inst Rev Board 1996;18:1-5. 13. Appelbaum PS. Rethinking the conduct of psychiatric research. Arch Gen Psychiatry 1997;54:117-120. 14. Appelbaum PS, Roth LH, Lidz CW, Benson P, Winslade W. False hopes and best data: consent to research and the therapeutic misconception. Hastings Cent Rep 1987; 17: 20-24. 15. Spece RG, Shimm DS, Buchanan AE. Conflicts of Interest in Clinical Practice and Research. New York, NY: Oxford University Press, 1996. 16. Brody J, Gluck JP, Aragon F. Participants' understanding of the process of psychological research: informed consent. Ethics Behav. In press. 17. Roberts LW, Solomon Z, Roberts BB, Keith SJ. Ethics of psychiatric research: resources for faculty development and resident education. Acad Psychiatry. In press.

18. Appelbaum PS, Roth LH. Competency to consent to research: a psychiatric overview. Arch Gen Psychiatry 1982;39: 951-958. 19. Curson DA, Hirsch SR, Platt SD, Bamber RW, Barnes TRE. Does short term placebo treatment of chronic schizophrenia produce long term harm? BMJ 1986;293:726-728. 20. Grinfeld MJ. NY court orders halt to research on psychiatric patients. Psychiatr Times 1996; 13:1-3. 21. Griffith JF, Davis FD, Johns ME, Blake DA. Integrity in scientific research. In: Kelley WN, Osterweis M, Rubin ER (eds): Emerging Policies for Biomedical Research. Washington, DC: Association of Academic Health Centers, 1993. 22. Viorst J. Listening at the keyhole: the Anne Sexton tapes. J Am Acad Psychoanal 1992;20:655-658. 23. Zaubler TS, Viederman M, Fins JJ. Ethical, legal and psychiatric issues in capacity, competency, and informed consent: an annotated bibliography. Gen Hosp Psychiatry 1996;18: 155-172. 24. Blumenthal D, Meyer GS. The future of the academic medical center under health care reform. N Engl J Med 1993;329:1812-1814. 25. Cadman E. The academic physician-investigator: a crisis not to be ignored. Ann Intern Med 1994; 120:401-410. 26. Kessler DA. Drug promotion and scientific exchange. N Engl J Med 1991 ;325:201-203. 27. Moran M. Psychiatrist finds patients' dignitary interests often at odds with 'research-industrial complex.' Psychiatr News 1996;10:17. 28. Pincus HA, Dial TH, Gaviland MG. Research activities of full-time faculty in academic departments of psychiatry. Arch Gen Psychiatry 1993;50:657-664. 29. Health United States and Healthy People 2000 Review, 1992. Bethesda, MD: US Department of Health and Human Services, 1993. 30. The Human Radiation Experiments: Final Report of the President's Advisory Committee. New York, NY: Oxford University Press, 1996. 31. Levine RJ. Ethics and Regulation of Clinical Research. Baltimore, MD: Urban & Schwartzenberg, 1981. 32. Rothman DJ. Ethics and human experimentation. N Engl J Med 1987;317:1195-1199. 33. Schaefer A. The ethics of the randomized clinical trial. N Engl J Med 1982;307:719-725. 34. Beecher HK. Ethics and clinical research. N Engl J Med 1966;274:1354-1361. 35. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research: The Belmont Report: Ethical Principles and Guidelines for the Protection of

ETHICAL BASIS OF PSYCHIATRIC RESEARCH

Human Subjects of Research. Washington, DC: US Government Printing Office, 1979. 36. President's Commission for the Protection of Human Subjects of Biomedical and Behavioral Research: Protecting Human Subjects. Washington, DC: US Government Printing Office, 1981. 37. Lidz CW, Meisel A, Osterweis M, Holden JL, Marx JH, Munetz MR. Barriers to informed consent. Ann Intern Med 1983;99:539-543. 38. Appelbaum PS, Grisso T. Assessing patients' capacities to consent to treatment. N Engl J Med 1988;319:1635-1638. 39. Gostin L. Informed consent, cultural sensitivity, and respect for persons. JAMA 1996;274:844-845. 40. Lane LW, Cassel CK, Bennett W. Ethical aspects of research involving elderly subjects: are we doing more than we say? J Clin Ethics 1990;1:278-284. 41. Roth LH, Appelbaum PS, Lidz CW, Benson E Winslade WJ. Informed consent in psychiatric research. Rutgers Law Rev 1987;39:425-441. 42. American Medical Association. Code of Medical Ethics: Current Opinions With Annotations. Ed. 1994. Chicago, IL: American Medical Association, 1994. 43. Winston ME, Winston SE, Appelbaum PS, Rhoden NK. Commentary: can a subject consent to a 'Ulysses contract'? Hastings Cent Rep 1982; 12:26-28. 44. Dresser RS. Ulysses and the psychiatrists: a legal and political analysis of the voluntary commitment contract. Harvard Civil Rights Civil Liberties Law Rev 1982;16:777-854. 45. Allen M. The dilemma for women of color in clinical trials. JAMWA 1994;9:105-109. 46. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317:141-146. 47. Rothman KJ, Michels KB. The continuing unethical use of placebo controls. N Engl J Med 1994;331:394-398. 48. Orr RD. Guidelines for the use of placebo controls in clinical trials of psychopharmacologic agents. Psychiatr Serv 1996;47:1262-1264. 49. Edlund M J, Craig TH, Richardson MA. Informed consent as a form of volunteer bias. Am J Psychiatry 1985; 142:624627. 50. Thomasma DC, McElhinney TK. Ethical concerns about AIDS. Pharos 1990;7-11. 51. Gilligan C. Concepts of self and morality. In: Gilligan C (ed): A different Voice. Cambridge, MA: Harvard University Press, 1982. 52. Roberts LW, McCarty T, Lyketsos C, Hardee JT, Jacobson J, Walker R, et al. What and how psychiatry residents at ten training programs wish to learn about ethics. Acad Psychiatry 1996;20:127-139. 53. Yolles B, Connors JD, Grufferman S. Sounding board: obtaining access to data from government-sponsored medical research. N Engl J Med 1986;315:1669-1672. 54. Kodish E, Stocking C, Ratain MJ, Kohrman A, Siegler M. Ethical issues in phase I oncology research: a comparison of investigators and institutional review board chairpersons. J Clin Oncol 1992;10:1810-1816. 55. Gray BH. Human Subjects in Medical Experimentation. New York, NY: Wiley, 1975. 56. Stanley B, Sieber JE, Melton GB. Empirical studies of ethical issues in research. Am Psychol 1987;42:735-741. 57. Stanley B, Stanley M, Guido J, Garvin L. The functional

109

competency of elderly at risk. Gerontologist 1988;28(Suppl):5358. 58. Stanley B, Stanley M, Lautin A, Kanw J, Schwartz N. Preliminary findings on psychiatric patients as research subjects: a population at risk? Am J Psychiatry 1981 ;38:669-671. 59. Appelbanm PS, Grisso T. The MacArthur Treatment Competence Study I, If, III. Law Hum Behav 1995;19:105-174. 60. Kleinman I, Schachter D, Jeffries J, Goldhamer E Effectiveness of two methods for informing schizophrenic patients about neuroleptic medication. Hosp Community Psychiatry 1993;44:1189-1191. 61. Schacter D, Kleinman I, Prendergast E Remington G, Schertzer S. The effect of psychopathology on the ability of schizophrenic patients to give informed consent. J Nerv Ment Dis 1994;182:360-362. 62. Katz M, Abbey S, Rydall A, Lowy E Psychiatric consultation for competency to refuse medical treatment. Psychosom Bull 1995;36:33-41. 63. Marson D, Cody HA, Ingram KK, Harrell LE. Neuropsychological predictors of competency in Alzheimer's disease using a rational reasons legal standard. Arch Neurol 1995;52:955959. 64. Marson D, Ingrain KK, Cody HA, Harrell LE. Assessing competency of patients with Alzheimer's disease under different legal standards. Arch Neurol 1995;52:949-954. 65. Grossman L, Summers E A study of the capacity of schizophrenic patients to give informed consent. Hosp Community Psychiatry 1980;31:205-206. 66. Benson PR, Roth LH, Winslade WJ. Informed consent in psychiatric research: preliminary findings from an ongoing investigation. Soc Sci Med 1985;20:1331-1341. 67. Block SD, Billings JA. Patient requests for euthanasia and assisted suicide in terminal illness: the role of the psychiatrist. Psychosomatics 1995;36:445-457. 68. Brown JH, Henteleff P, Barakat S, Rowe CJ. Is it normal for terminally ill patients to desire death? Am J Psychiatry 1986;143:208-211. 69. Cavanaugh S. Depression in the medically ill: critical issues in diagnostic assessment. Psychosom Bull 1995;36:4859. 70. Ganzini L, Lee MA, Heintz RT, Bloom JD, Fenn DS. The effect of depression treatment on elderly patients' preferences for life-sustaining medical therapy. Am J Psychiatry 1994; 151: 1631-1636. 71. Breitbart W, Rosenfeld BD, Passik SD. Interest in physician-assisted suicide among ambulatory HIV infected patients. Am J Psychiatry 1996;153:238-242. 72. Soskis DA. Schizophrenic and medical inpatients as informed drug consumers. Arch Gen Psychiatry 1978;35:645647. 73. Sachs GA, Stocking CB, Stern R, Cox DM, Hougham G, Sachs RS. Ethical aspects of dementia research: informed consent and proxy consent. Clin Res 1995;403-412. 74. Warren JW, Sobal J, Tenney JH, Hoopes JM, Damron D, Levenson S, et al. Informed consent by proxy: an issue in research with elderly patients. N Engl J Meal 1986;315:11241128. 75. High DM. Research with Alzheimer's disease subjects: informed consent and proxy decision making. J Am Geriatr Snc 1992;40:950-957. 76. Cassileth BR, Zupkis RV, Sutton-Smith K, March V.

110

Informed consent: why are its goals imperfectly realized? N Engl J Med 1980;302:896-900. 77. Grossman SA, Piantadosi S, Covahey C. Are informed consent forms that describe clinical oncology research protocols readable by most patients and their families? J Clin Oncol 1994;12:2211-2215. 78. Taub HA, Baker MT, Stuff JE Informed consent for research: effects of readability, patient age, and education. J Am Geriatr Soc 1986;34:601-606. 79. Tomamichel M, Sessa C, Herzig S, de Jong J, Pagani O, Willems Y, et al. Informed consent for phase I studies: evaluation of quantity and quality of information provided to patients. Ann Oncol 1995;6:363-369. 80. Maslin A. A survey of the opinions on 'informed consenr of women currently involved in clinical trials within a breast unit. Eur J Cancer Care 1994;3:153-162. 81. Philips M, Vazquez AJ. Abnormal findings in 'normal' research volunteers. Control Clin Trials 1987;8:338-342.

LAURA WEISS ROBERTS

82. Logue G, Wear S. A desperate solution: individual autonomy and the double-blind controlled experiment. J Med Philos 1995;20:57-64. 83. Harth SC, Thong YH. Parental perceptions and attitudes about informed consent in clinical research involving children. Soc Sci Med 1995;40:1573-1577. 84. Daugherty C, Ratain MJ, Grochowski E, Stocking C, Kodish E, Mick R, et al. Perceptions of cancer patients and their physicians involved in phase I trials. J Clin Oncol 1995;13:10621072. 85. Cassileth BR, Lusk EJ, Miller DS, Hurwitz S. Attitudes toward clinical trials among patients and the public. JAMA 1982;248:968-970. 86. Cassileth BR, Lusk EJ, Guerry D, Blake AD, Walsh WP, Kascius L, et al. Survival and quality of life among patients receiving unproven as compared with conventional cancer therapy. N Engl J Med 1991;324:1180-1185.